European Competition Law Annual 2000: The Modernisation of EU Competition Law

EUROPEAN COMPETITION LAW ANNUAL 2000: The Modernisation of EC Antitrust Policy

EUROPEAN COMPETITION LAW ANNUAL: 2000 The Modernisation of EC Antitrust Policy

Edited by

Claus-Dieter Ehlermann and Isabela Atanasiu

•HARTPUBLISHING

OXFORD - PORTLAND OREGON 2001

Hart Publishing Oxford and Portland, Oregon Published in North America (US and Canada) by Hart Publishing c/o International Specialized Book Services 5804 NE Hassalo Street Portland, Oregon 97213-3644 USA Distributed in the Netherlands, Belgium and Luxembourg by Intersentia, Churchillaan 108 B2900 Schoten Antwerpen Belgium © The contributors severally, 2001 The contributors severally have asserted their rights under the Copyright, Designs and Patents Act 1988, to be identified as the authors of this work Hart Publishing is a specialist legal publisher based in Oxford, England. To order further copies of this book or to request a list of other publications please write to: Hart Publishing, Salter's Boatyard, Folly Bridge, Abingdon Road, Oxford OX1 4LB Telephone: +44 (0)1865 245533 or Fax: +44 (0)1865 794882 e-mail: [email protected] WEBSITE: http//www.hartpub.co.uk British Library Cataloguing in Publication Data Data Available ISBN 1-84113-242-X (hardback)

Typeset by Hope Services (Abingdon) Ltd. Printed and bound in Great Britain on acid-free paper by Page Bros, Norwich

CONTENTS

List of Sponsors Table of Cases

.

INTRODUCTION

vii ix XV

PANEL ONE: COMPATIBILITY, EFFICIENCY, LEGAL SECURITY

1 Panel Discussion Introductory Statement by Mario Monti Discussion 2 Working Papers I Simon Bishop II Ulf Boge III Ian Forrester IV Eleanor Fox V Barry Hawk and Nathalie Denaeijer VI Giuliano Marenco VII Mitsuo Matsushita VIII Petros Mavroidis and Damien Neven IX Ernst-Joachim Mestmacker X Alexander Schaub XI Giuseppe Tesauro

3 13 55 66 75 123 129 145 197 207 223 241 259

PANEL TWO: COHERENCE

1 Panel Discussion

271

2 Working Papers I Christopher Bellamy II Jacques Bourgeois III Laurence Idot IV Ulrich Immenga V Frederic Jenny VI Anne Willem Kist and Maria Luisa Tierno Centella VII Gheorghe Oprescu VIII Emil Paulis IX James Rill X Mario Siragusa XI James Venit

315 323 335 353 360 369 387 399 429 443 457

vi

Contents

PANEL THREE: COURTS AND JUDGES

1 Panel Discussion

477

2 Working Papers I Carl Baudenbacher II Jochen Burrichter III John D. Cooke IV David Edward - V DanGoyder VI Jochem Groning VII Santiago Martinez Lage and Helmut Brokelmann VIII Anne Spiritus-Dassesse IX Sarah S. Vance X Diane P. Wood

521 539 551 564 571 579 593 613 617 627

BIBLIOGRAPHY

639

INDEX

645

LIST OF SPONSORS

Akin, Gump, Strauss, Hauer & Feld L.L.P. Website: http://www.akingump.com Contact: Brussels Office Avenue Louise 65 P.B. No. 7 B-1050 Brussels Tel: (32 2) 535 29 11 Fax: (32 2) 535 29 00 E-mail: [email protected] Cleary, Gottlieb, Steen & HamUton Website: http://www.cgsh.com Contact: Brussels Office Rue de la Loi 23 B-1040 Brussels Tel: (32 2) 287 20 00 Fax: (32 2) 231 16 61 Howrey Simon Arnold & White Contact: James F. Rill, esq 1299 Pennsylvania Ave., NW Washington, DC 20004 Tel: (001 202) 383 65 62 E-mail: [email protected] Fax: (001 202) 383 66 10 Martinez Lage & Asociados Contact: Santiago Martinez Lage Madrid Office Calle Serrano 25 28001 Madrid Tel: (34 91) 426 44 70 Fax:(34 91)577 37 74 E-mail: [email protected]

viii

List of Sponsors

Skadden, Arps, Slate, Meagher & Flom L.L.P. Website:

http://www.skadden.com

Contact: Barry Hawk/Henry Huser Brussels Office 523 Avenue Louise B-1050 Brussels Tel: (32 2) 639 03 00 Fax: (32 2) 639 03 39 White & Case, L.L.P. Website:

http://www.whitecase.com

Contact: Brussels Office 1 Place Madou Box 34 B-1210 Brussels Tel: (32 2) 219 16 20 Fax:(32 2)219 16 26 Wilmer, Cutler & Pickering Website: http://www.wilmer.com Contact: Brussels Office Rue dela Loi 15 B-1040 Brussels Tel: (32 2) 285 49 00 Fax: (32 2) 285 49 49

TABLE OF CASES

EC Cases Ahlstrom and Others v Commission, Joined Cases 89, 104, 114, 116, 117 and 125-129/85 [1992] ECR 1-1037. Ahmed Saeed, Case 66/86 [1989] ECR 803. Albany International v Stischting Bedrijfspensioenfonds Textielindusirie, Case C-67/96 [1999] ECR 1-5751. Alcatel/EspacelANT Nachrichtentechnik, Commission Decision, OJ L 31/19 1990. Anic Partecipazioni, Case C-49/92 [1999] ECR 1-4125 Asjes, Joined Cases 209 to 213/84 [1986] ECR 1425. Automec v Commission (Automec II), Case T-24/90 [1992] ECR 11-2223. Banks, Case 313/93 [1994] ECR 1-1279. Banks v British Coal, Case C-128/89 [1994] ECR 1-1209. BATv Reynolds, Joined Cases 142/84 and 156/84 [1987] ECR 4487. Belasco and Others v Commission, Case 246/86 [1989] ECR 2117. Belgische Radio en Televisie et Societe beige des auteurs, compositeurs et editeurs v SVSABAMet NVFonior (BRT-I), Case 127/73 [1974] ECR 51. Bier v Mines de Potasse d'Alsace, Case 21/76 [1976] ECR. BNICv Clair, Case 123/83 [1985] ECR 391. BoeinglMcDonnell Douglas, Commission Decision M877, OJ L 336/16 of 8.12.1997. Boussoisllnterpane, Commission Decision of 15.12.1986 in Case IV/31.302, OJ L 50/30 1987. Brasserie de Haecht v Wilkin (Brasserie de Haecht II), Case 48/72 [1973] ECR 77. British Leylandv Commission, Case 226/84 [1986] ECR 3263. BRT\ Sabam, Case 127/73 [1974] ECR 51. Carbon Gas Technologie, Commission Decision of 8.12.1983 in case IV/29.955, O J L 376/17 1983. Carlo Bagnasco and Others v Banca Popolare di Novara s.c.a., Cassa di Risparmio di Genova e Imperia SpA (Carige), Joined Cases C-215/96 and C-216/96 [1999] ECR 1-179. CILFIT, Case 283/81 [1982] ECR 3415. Coditel SA, Compagnie General pour la Diffusion de la Television and Others v Cine-Vog Films SA and Others, Case 262/81 [1982] ECR 3381. Commission v Anic Partecipazioni, Case C-49/92 P [1999] ECR 1-4125. Commission v France, Case C-159/94 [1997] ECR 1-5815. Commission v Germany, Case C-217/88 [1990] ECR 1-2879.

Table of Cases

Consten SARL & Grundig- Verkaufs-GmbH v Commission, Joined Cases 56/64 and 58/64 [1966] ECR 299. Corbeau, Case C-320/91 [1993] ECR 1-2533. Corte Ingles v Cristina Bldzquez, Case C-192/94 [1996] ECR 1-1281. De Bloos v Bouyer, Case 14/76 [1977] ECR 2359. De Geus v Bosch, Case 13/61 [1962] ECR 89. Delimitis v Henninger Brau, Case 234/89 [1991] ECR 1-935. Direction General de la Defensa de la Competencia v AEBP, case C-67/91 [ 1992] Ecr1-1475. DM Transport SA, Case C-256/97 [1999] ECR 1-3926. Drouot, Case C-357/96 [1998] ECR 1-3075. Eco Swiss China Time v Benetton International, Case C-126/97 [1999] ECR 1-3055. ElopaklMetal Box Odin, Commission Decision, OJ L 209/15 1990. Emmot v Minister for Social Welfare, Case C-208/9 [1991] ECR 1-4269. European Night Services and Others v Commission, Joined Cases T-374/94, T-375/94, T-384/94 and T-388/94 [1998] ECR 11-3141. FacciniDori, Case C-91/92 [1994] ECR 1-3325. Factortime, Case C-213/89 [1992] ECR 1-2433. Francovich v Italian Republic, Joined Cases C-6/90 and C-9/90 [1991] ECR 1-5357. Ford Agricultural Tractors, Commission Decision, OJ L 68/19 1992. Ford-Volkswagen, Commission Decision 93/49/EEC of 23.12.1992, OJ L 20/14 1993. Fotofrost, Case 314/85 [1987] ECR 4199. Frubo v Commission, Case 71/74 [1975] ECR 563. Giry and Guerlain, Joined Cases 253/78 and 1 to 3/79 [1980] ECR 2327. Gottrup Klim Grovvareforeninger v Dansk Landbrugs Grovvareslskab AmbA, Case C-250/92 [1994] ECR 1-5641. Groupement des cartes bancaires, [1994] ECR 11-49. Hoffman La Roche, Case 85/76 [1979] ECR-461. Hugin, Case 22/78 [1979] ECR 1898. John Deere v Commission, Case C-7/95 P [1998] ECR 1-3111. Kali und Salz, [1975] ECR 499. Keck and Mithouard, case C-268/91 [1993] ECR 1-6097. Kledingverkoopbedrijf de Geus en Uitdenbogert v Robert Bosch GmbH and Maatschapppij tot voortzetting van den zaken der Firma Willem van Rijn, Case 13/61 [1962] ECR 89 (see also Opinion of Advocate General Lagrange [1962] ECR 309). KO v de Agostini and TV Shop, Joined Cases C-34/95 and C-36/95 [1997] ECR 1-3843. Konsortium ECR 900, Commission Decision, OJ L 228/31 1990. Kruidvat v Commission, Case T-87/89 [1996] ECR 11-1931, on appeal Case 70/97 P, (judgment of 17.10.1998). Langnese Iglo GmbHv Commission, Case T-7/93 [1995] ECR 11-1533.

Table of Cases

xi

L. C. Nungesser KG and Kurt Eisele v Commission of the European Communities, Case 258/78 [1982] ECR 2015. Louis Erauw-Jacquery SPRL v La Hesbignonne SC, Case 27/87 [1988] ECR 1919. Marshall I, Case 152/84 [1986] ECR 723. Marshall II, Case C-271/91 [1993] ECR 1-4367. Marty v Lauder, Case 127/73 [1980] ECR 2481. Masterfoods, Case C-344/98,(judgment of 14.12.2000, not yet reported). Matra Hachette v Commission, Case T-17/93 [1994] ECR 11-595. Metro v Commission (Metro I), Case 26/76 [1977] ECR 1875. Metro v Commission (Metro II), Case 75/84[1986] ECR 3021. Metropole Television, Case T-528/93 [1996] ECR 11-649. Michelin v Commission, Case 322/81 [1983] ECR 3461. Miller v Commission, Case 19/77 [1977] ECR 131. Nederlandse Vereniging van Banken (1991GSA Agreement), Commission Decision of 8.9.1999 in Cases IV/34.010, IV/33.793, IV/34.234 and IV/34.888, OJ L 271/28 1999. Nordsee v Reedrei Mond, Case 102/81 [1982] ECR 1095. Opel Austria, Case T-l 15/94 [1997] ECR 11-39. Oscar Bronner, Case C-7/97 [1998] ECR 1-7791. Oude Luttikhuis and Others v Verening de Cooperatieve Melkindutries, Case C-399/93 [1995] ECR 1-4515. Papiers Feints de Belgique, Case 73/74 [1975] ECR 1491. Procureur du Roi v Dassonville, Case 8/76 [1974] ECR 837. Pronuptia de Paris GmbH v Pronuptia de Paris Irmgard Schillgalis, Case 161/84 [1986] ECR 353. Rechberger, Case C-140/97 [1999] ECR 1-3499. Remia BVand Others v Commission, Case 42/84 [1985] ECR 2566. Rewe-Zentral v Bundesmonopolverwaltung fur Branntwein (Cassis de Dijon), Case 120/78 [1979] ECR 649. Rurhgas and Thyssengas v Bundeskartellamt, Case C-365/96, cancelled from the Court register by order of 26.3.1998. REWEv Hauptzollamt Kiel, Case 158/80 [1981] ECR 1805. RWE and Stadt Nordhorn v Bundeskartellamt, Case C-34/97, cancelled from the Court register by order of 13.10.1998. Sacchi, Case 155/73 [1974] ECR 409. SA Lancome and Co sparfranee Nederland BV v Etos BV and Albert Heyn Supermarket BV, Case 99/79 [1980] ECR 2511. S.A. Portelange v SA Smith Corona Marchant International and Others, Case 10/69 [1969] ECR 309. Schoeller, Case T-9/93 [1995] ECR 11-1611. Simmenthal, Case 106/77 [1978] ECR 629. Sirena Sri v Eda Sri and Others, Case 40/70 [1971] ECR 69. Societe Technique Miniere v Maschinenbau Ulm, Case 56/65 [1966] ECR 337. Sole distribution agreements for whisky and gin, Commission Decision of 13.12.1985 in Case IV/30.570, OJ L 369/19.

xii

Table of Cases

SPO and Others v Commission, Case T - 29/92 [1995] ECR 11-289 (upheld on appeal. Case C-137/95 P [1996] ECR 1-1613). Stork, Case 1/58 [1959] ECR 17. Suzen, Case C-13/95 [1997] ECR 1-1259. Stork Amstedam, Case T-241/97 [2000] ECR 11-30. Sydhanvnens sten og grus, Case C-209/98 (judgment of 23.5.2000, not yet reported). Syndacat national des fabricants raffineurs d'huile de graissage, Case 172/83 [1983] ECR 555. Tetra Pak v Commission, Case T-51/89 [1990] ECR 11-309. Tipp-Ex v Commission, Case 279/87 [1990] ECR 1-361. Trefilunion v Commission, Case T-148/89 [1995] ECR 11-1063. TWD, Case C-188/92 [1994] ECR 1-0833. Ufex and Others v Commission, Case C-l 19/97 P [1999] ECR 1-1341. Union Royale Beige des Societes de Football Association v. Bosman and Others, Case C-415/93 [1995] ECR 1^921. Van den Bergh Foods Limited, Commission Decision in Cases IV/34.073, IV/34.395 and IV/35.35.436, OJ L 246/1 1998. Van den Bergh Foods Limited, Case T-65/98. Van Gend&Loos, Case 26/62 [1963] ECR 5. Van Landewick (Fedetab), [1980] ECR 3125. Van Munsten, Case C-l65/91 [1994] ECR 1—4661. Van Schjindel, Joined Cases C-430/91 and C-431/91 [1995] ECR 1-4705. VBVB and VBBB, Joined Cases 43/82 and 63/82 [1984] ECR 19. VBVB v VBBB, Case T-17/93 [1994] ERC 11-595. Verband der Sachversicherer, Case 45/85 [1987] ECR 405. Vereniging van Samenwerkende Prijregelende Organisatie in de Bouwnijverheidv Commission, Case T-29/92 [1995] ECR 11-289. Vereniging van Cementhandelaren, Case 8/72 [1972] ECR 977. VGB and Others v Commission, Case T-77/94 [1995] ECR 11-759. Walt Wilhelm v Bundeskartellamt, Case 14/68 [1969] ECR 1. WEA-Filipacchi, Commission Decision 72/480/CEE, OJ L 303/52 1972. XlOpem Group, Commission Decision, OJ L 35/36 1987. Zuchner v Bayerische Vereinsbank, Case 172/80 [1981] ECR 2021.

US Cases Alaska v Safeway, Inc. and Carr-Gottstein Foods Co., Case No. 3AN-99-4371 (Sup. Ct. Alaska 1999). Brown v Allen, 344 U.S. 443 (1953). California Dental Associations v FTC, 526 U.S. 756, 119 S. Ct. 1604 (1999). Chevron U.S.A. Inc. v Natural Resources Defense Council, Inc., 461 U.S. 837 (1984).

Table of Cases

xiii

Commonwealth of Pennsylvania v Russell Stover Candies, Inc., 1993-1 Trade Cas. (CCH) (E.D. Pa. 1993). Continental T. V, Inc. v GTE Sylvania, Inc., 433 U.S. 36 (1977). Elliot v Commodities Futures Trading Comm'n, 202 F. Supp. 3d 926 (7th Cir. 2000). Food & Drug Administration v Brown & Williamson Tobacco Corp., 120 S. Ct. 1291 (2000). FTC vBP Amoco, pic, Case No. 000415SI (N.D. Cal. Feb. 4, 2000) FTCvButterworth Health Corp., 1997-2 Trade Cas. (CCH) (6th Cir. 1997). FTCv Cardinal Health, Inc., 12 F. Supp. 2nd 34 (D.D.C. 1998). FTC\ Staples, Inc., 970 F. Supp. 1066 (D.D.C. 1997). FTCv Tenet Health Care Corp., 117 F. Supp. 2d 937 (E.D. Mo. 1998). HardfordFire Ins. Co. v California, 509 U.S. 764 (1993). Jefferson Parish Hosp. Dist. No. 2 v Hyde, 104 S. Ct. 1551 (1984). Loeb v Eastman Kodak Co., 183 F 704 (3rd Cir. 1910). McLain v Real Estate Board, 444 U.S. 322. New York v Kraft General Foods, Inc. et al, 926 F. Supp. 321 (S.D.N.Y. 1995). New York v Primestar Partners, L.P., 1993-2 Trade Cas. (CCH) (S.D.N.Y. 1993). Spectrum Sports v McQuillan, 506 U.S. 447. Summit Health v Pinhas, 500 U.S. 322. Toys 'R' US v FTC, No. 98-4107 (7th Cir. 1999). United States v Allied Waste Indus., Inc., 1992-2 Trade Cas. (CCH) (D.D.C. 1999). United States v Cargill, Inc., 1997-2 Trade Cas. (CCH) (W.D.N.Y. 1997). United States v Long Island Jewish Med. Ctr, 983 F. Supp. 121 (E.D.N.Y. 1997). United States v Microsoft Corp., 84 F. Supp. 2d 9 (D.D.C. 1999) (findings of fact); 87 F. Supp. 2d 30 (D.D.C. 2000) (conclusions of law). United States v Primestar Partners, L.P., 1994-1 Trade Cas. (CCH) (S.D.N.Y. 1994). United States v Topco Assocs., Inc., 405 U.S. 596, 629-12 (1972).

INTRODUCTION

On 2-3 June 2000 the European University Institute (EUI) hosted thefifthedition of the Annual EC Competition Law and Policy Workshop. This program, started in 1996 by law professors Giuliano Amato and Claus-Dieter Ehlermann at the Robert Schuman Centre of the EUI, brings together every year top-level policy-makers, academics and legal specialists to discuss critical issues of EC competition policy. The first edition of the Workshop (1996) focused on the problems of implementing competition policy in a federal context.' The second (1997) debated the objectives of competition law and policy.2 The third (1998) analysed how to ensure effective competition in the rapidly evolving market of communications.3 The fourth (1999) examined selected problems of State aid control.4 The fifth edition of the Workshop was devoted to the ongoing debate about the modernisation of EU antitrust law.5 This publication includes a transcript of the discussions and a collection of the writings presented by the participants at thefifthEU Competition Law and Policy Workshop. The event took place one year after the European Commission published the White Paper containing its suggestions for the reform of EC antitrust law enforcement,6 at a stage when consultations on the suggestions were already fairly advanced. The Workshop concentrated on those issues which the White Paper and the ensuing discussions had identified as particularly problematical and controversial. Proceedings were divided into three sessions. Panel One was devoted to the following three aspects of the reform: its compatibility with the EC Treaty; its effects on the efficiency of EC anti-cartel enforcement, and its consequences for undertakings interested in legal security. Panel Two focused on the risks that the reform entails for the coherent application of Article 81 EC Treaty. Panel Three 1

Ehlermann C.-D. and Laudati. L. L., eds. (1997): The Robert Schuman Centre Annual on European Competition Law 1996, The Hague, Kluwer Law International. 2 Ehlermann C.-D. and Laudati. L. L., eds. (1998): European Competition Law Annual 1997—Objectives of Competition Policy, Oxford, Hart Publishing. 3 Ehlermann C.-D. and Gosling L., eds. (2000): European Competition Law Annual 1998—Regulating Communications, Oxford, Hart Publishing. 4 Ehlermann C.-D. and Everson M., eds: European Competition Law Annual 1999: State Aid Control in the European Union—Selected Problems, Oxford, Hart Publishing (forthcoming during Winter 2000/2001). 5 The term 'antitrust' is used to cover in particular Article 81 EC Treaty. Issues concerning the implementation of Article 82 EC Treaty were not discussed during the Workshop. 6 European Commission: White Paper on modernisation of the rules implementing Articles 85 and 86 of the EC Treaty, OJ C 132 of 12.5.1999.

xvi

Introduction

concentrated on the specific problems that the reform will generate for national judges, who will become responsible for the interpretation and application of Article 81 (3) EC Treaty. The Workshop revealed broad consensus on the main thrust of the Commission's reform initiative. Yet it also showed that a series of issues have to be considered further and addressed in a formal proposal from the Commission to the Council. Among these issues are: the scope of supremacy of EC competition law with respect to national law; the legal effects of positive (or 'noninfringement') Commission decisions based on Article 81 (3); the legal effects of decisions taken by national competition authorities and by national courts in other Member States; the burden of proof with respect to Article 81 (3); the distribution of cases between the Commission and national competition authorities, as well as among the national authorities themselves. Background. The White Paper suggested a radical departure from the existing EC antitrust law enforcement system. Council Regulation 17/62 reserves to the Commission the power to exempt, according to Article 81 (3) EC Treaty, agreements that fall under Article 81(1) EC Treaty (and that are, therefore, prohibited), provided they have been properly notified. Because of the Commission's exemption monopoly, Article 81 (3) has no direct effect. The White Paper proposes to abolish the system of notifications and the Commission's exemption monopoly. It suggests instead that Article 81 (3) should become directly effective, so that it can be also be applied by national competition authorities (NCAs) and by national courts. The centralisation of the exemption power under Article 81 (3) in the hands of the Commission, and the corresponding notification requirement, have given rise to enormous problems in enforcement. In the early sixties, the Commission was swamped by notifications. In reaction, the Commission adopted so-called 'group exemptions' (i.e. regulations declaring Article 81 (1) inapplicable to whole categories of agreements), and developed the instrument of 'comfort letters', giving a sort of green light to individual agreements. Individual exemption decisions remained extremely rare; their average number did not exceed 5 during the last 5 years. Group exemptions were criticised as anti-economic and over-bureaucratic. Comfort letters were considered to be poor substitutes for formal decisions, because they do not have legal effects, and are, therefore, not able to eliminate the invalidity resulting from Article 81 (1) and (2). In addition, the Commission was not even able to eliminate totally the backlog of pending requests for exemptions. Thus, the Commission never succeeded to manage its exemption monopoly in a satisfactory way. Over the last decade, criticism of the existing antitrust enforcement system intensified. Comparisons were made with the successful application of the 1989 Merger Regulation. The debate on subsidiarity stimulated requests for the abolition of the Commission's exemption monopoly in favour of the NCAs. Until recently, the Commission responded to this largely justified criticism by trying to improve the application of Article 81 within the existing legal framework, e.g. by favouring the decentralised application of Article 81 (1) and (2) by

Introduction

xvii

NCAs and national courts, while maintaining the monopoly to adopt exemption decisions under Article 81 (3). In this perspective, the reform suggestions of the White Paper represented not only a legal, but also a 'cultural' revolution. Previous EC Competition Law and Policy Workshops held at the EUI might have contributed to the Commission's change of mind. During the 1996 and 1997 meetings, as one of the participants at the June 2000 Workshop put it, '. . . the faces of Commission officials flushed and blushed to hear severe criticism: the European system . . . did not work, could not work, had to change.' In spite of the traditional criticism of the existing situation, the immediate reactions of the legal and business community to the Commission's reform proposal were sometimes rather sceptical, particularly in Germany, though in the past German voices have been most critical about the centralising effects of Regulation 17/62. During the consultations following the publication of the White Paper, however, the initial scepticism has somewhat diminished. Discussions at the June 2000 Workshop revealed that, in the meantime, a broad consensus seems to have emerged around the Commission's proposals for the decentralisation of EC antitrust enforcement. However, there were also warning views. In addition to the well-known German critics, some participants expressed the fear that the Commission might go too far in its enthusiasm for decentralisation, and that it might jeopardise its essential role in determining and guiding competition law enforcement in the EU. Everybody agreed, nevertheless, on the need for careful preparation of the reform, and in particular on the need to give precise, clear and convincing answers to the issues identified in the debate. The reform was generally considered to be too important for the future of the EU's competition policy to be rushed through the EU's legislative process, thus leaving well-known questions unresolved, reserving them to later administrative decisions of competition authorities and judgements of the courts. The Workshop Proceedings. The annual EC Competition Law and Policy Workshops hosted by the EUI in Florence provide an ideal forum for the informal debate of critical issues in EU competition policy among policy-makers and actors involved in EU competition law implementation. Thefifthedition of the Workshop brought together a group of thirty-three top-level participants, including: high-ranking Commission officials directly involved in the preparation of the reform; judges from the EU and Member State courts, as well as from the US and EFTA; representatives of competition authorities in the Member States and one of the countries candidate to EU membership (Romania); academics; and outstanding practitioners in the field of antitrust. The event was co-chaired by Claus-Dieter Ehlermann and Karel van Miert, former Member of the European Commission responsible for Competition. The proceedings of the Workshop were opened by Professor Mario Monti, Karel van Miert's successor in the Prodi Commission. Professor Monti explained the reasons requiring a fundamental reform of EU antitrust enforcement. He underlined the need for the intensified application of EU competition law both by EU and by NCAs and national courts. This would be achieved by

xviii

Introduction

establishing a network of competition authorities, based on vertical and horizontal co-operation and mutual assistance, and by giving the national courts their natural role as protectors of private rights derived from the EC Treaty. Professor Monti recognised the risk of divergence that decentralisation will inevitably entail. He expressed, however, his determination to clarify existing rules, thus bringing about in the end more convergence than ever before. Of particular interest is Professor Monti's view that the modernisation of EU antitrust law enforcement through increased decentralisation is a paradigm for the reform of other activities of the European Commission. Panel 1 of the Workshop was devoted to discussing the following issues: (a) the compatibility of the reform envisaged by the Commission in the White Paper with the EC Treaty, (b) whether the reform will enhance the efficiency of EC antitrust policy, (c) whether it will bring about the desired simplification of EC antitrust procedures, but nevertheless ensure sufficient legal certainty for the undertakings concerned. The issue of the compatibility of the Commission's reform proposal with the EC Treaty is discussed in several of the papers prepared for the Workshop. One of the participants in particular elaborated the three main arguments which, in his view, plead against the reform's compatibility with the Treaty. Thefirstargument is that the legal exemption system envisaged by the Commission embodies de facto a 'control of abuse' approach in enforcement, rather than the prohibition approach established by the Treaty. Furthermore, the drafting of Article 81 EC Treaty, and in particular the introductory formula of paragraph 3 ('. . . the provisions of paragraph 1 may however be declared inapplicable ...'—emphasis added), as well as the wording of Article 83 EC Treaty, indicate that the founders of the Treaty intended to establish an administrative authorisation system. Finally, the jurisprudence of the EC Court of Justice (ECJ), and in particular its recognition of the Commission's discretion in applying Article 81 (3), is proof of the fact that Article 81 (3) is incapable of having direct effect. Assigning its implementation to national courts would, de facto, deprive paragraph 1 and 2 of Article 81 of their direct effect. According to this argumentation, the system of administrative authorisation has to be maintained. While the Commission's exemption monopoly may be shared with national competition authorities, Article 81 (3) can not be applied directly by national courts. Though some participants expressed doubts as to whether national judges were capable of applying directly Article 81 (3), or whether it was appropriate and wise, particularly at this stage, to entrust direct application of this provision to them, nobody shared the fundamental legal objections against the envisaged reform. Noteworthy is, in particular, the positive position of present and former members of the Courts in Luxembourg who participated in the Workshop. Defenders of the reform repeated and elaborated the already well-known arguments about: the respect and reinforcement of the prohibition principle established in Article 81(1) and (2); the deliberately ambiguous form of Article

Introduction

xix

81 (3) as compared to the clear wording of Article 65 ECSC; the wide scope of Article 83; the similarities between Article 81 (including its paragraph 3) and other Treaty provisions for which the ECJ has recognised direct effect. After all, direct effect is the rule, and not the exception, in EC law. Of particular interest are the findings of two participants who investigated in depth the history of the negotiations leading to the wording of Article 85 EEC Treaty (now 81 EC Treaty). While one of them qualifies the discussions as not focussed and amicable explorations of concepts unfamiliar to negotiators, the other comes to the conclusion that the White Paper is right in its assertion that the divergence of views between the legal exemption versus the administrative authorisation approach was deliberately left open and reserved for future Council legislation. In practice, the absence of direct effect of Article 81 (3) under the actual Regulation 17/62 might not have had as drastic effects as a strictly legal analysis might suggest. One participant underlined that even under the existing authorisation system, national judges have found ways to take into account the outcome of an analysis of Article 81 (3), though this analysis is formally reserved to the Commission. Several participants noted that the switch from the Commission's exemption monopoly to direct effect of Article 81 (3) will have repercussions on the interpretation and application of this provision by the ECJ: the margin of discretion granted to the Commission by the ECJ will disappear. The categories of considerations used explicitly or implicitly in interpreting and applying Article 81 (3) may also change. The 'judicialisation' of Article 81 in its totality may even have repercussions on the Commission's administrative procedures under Regulation 17/62. The Workshop showed consensus that it is urgent to clarify the scope of Article 8(3). It is obvious that this scope depends on the interpretation of Article 81 (1): the broader the reach of Article 81 (1), the greater the need for recourse to Article 81 (3). On the contrary, if Article 81 (1) is interpreted more restrictively, the field of application of Article 81 (3) will shrink correspondingly. The discussion of this issue showed a wide spectrum of opinions. One participant argued in favour of a restrictive interpretation of the requirement of effect on trade between Member States. Others suggested to situate the totality of strictly competition-oriented arguments in Article 81 (1), leaving to Article 81 (3) only non-competition oriented considerations. In such a perspective, the Commission's reform proposal would be misplaced: the Commission should maintain its exemption monopoly under Article 81 (3). The majority of participants seemed, however, to follow the White Paper's 'middle of the road' approach: the notion of 'restriction of competition' has to be interpreted more realistically, in particular taking into account the position of the undertakings concerned on the relevant market, but without situating all the weighting and balancing of economic factors in Article 81 (1). The problem of the borderline between Article 81 (1) and (3) therefore remains, as much as the debate on the legality of taking into account non-competition oriented considerations in Article 81 (3). Recent EC Treaty reforms have increased the number and importance of such considerations, provided for in other parts of

xx

Introduction

the Treaty. The need for Commission guidelines on the interpretation of Article 81 (3), recognised by the Director General of DG Competition, has grown correspondingly. Opponents of the Commission's reform suggestions have shown scepticism as to whether the reform will simplify and render more efficient the enforcement of EC antitrust law. German critics point to the fact that the notification and administrative authorisation system has proved to be effective in the context of German antitrust enforcement. One of its advantages is that it enabled the Bundeskartellamt to remain informed on market developments. In line with this experience, the Head of the Bundeskartellamt repeated his recent proposal to oblige undertakings to inform NCAs about their restrictive agreements and to make this information available on the Internet. He noted, however, that other positive aspects of the German experience, like the preventive and deterrent effect of the administrative authorisation system, are generally considered not to exist at EU level. He was joined by others in observing that 'exemptions allow negotiation, and fine tuning' (like fixing a time limit for the validity of an exemption) which will disappear under a legal exemption regime. The great majority of participants welcomed the initiative to abandon the notification and administrative authorisation system. They shared the view expressed in the White Paper and by Commission representatives at the Workshop that requests for exemptions detract the Commission's rare resources from more important work, i.e. dealing with complaints and detecting horizontal, hard core cartels. They also greeted the elimination of the official separation of Article 81(1) and (3). Not everybody was convinced that the elimination of the Commission's exemption monopoly would lead to the desired results, i.e. that NCAs and national courts will use more often Article 81. NCAs might continue to prefer to apply national competition law. National courts depend on private action, i.e. the initiative of undertakings to have recourse to juridical proceedings. Where are the incentives (in addition to the elimination of the exemption monopoly obstacle) to use Article 81 more frequently than in the past? In addition, adequate resources are lacking not only at Commission level. They are also needed at the level of NCAs and national courts. The new approach to Article 81(1), emphasising economic instead of purely legal factors, was generally welcome, though again, German voices warned against an excessive softening of the EC antitrust discipline. Participants noted that this approach would have consequences in terms of the predictability of outcome of the analysis under Article 81 (1). These consequences are not dependent on the choice of systems for the implementation of Article 81 (3). Under the new 'economic' approach, it will indeed be more difficult to foresee whether Article 81 (1) applies than under the traditional, more legalistic interpretation. Part of the old (Commission and ECJ) decisions will not provide guidance for the future. Consequently, the need for Commission guidelines on the interpretation of Article 81 (1) will increase. These guidelines have to be clear and should regularly be updated.

Introduction

xxi

Whether and to what extent the abolition of individual exemption decisions will add to existing and future uncertainties remained a matter of controversy. The White Paper expresses the view that the legal exemption system will eliminate the major cause of legal insecurity, because the direct effect of Article 81 (3) will dispense of the adoption of individual exemption decisions which, in any case, were rarely taken in the past. Participants from the USA went even further. They considered European preoccupations about the validity of contracts to be exaggerated. For them, legal certainty is not a serious problem. Doesn't business prefer a rule of reason approach to per se prohibitions? Participants reflecting the positions of European business took a different view. Considering the Commission's adamant opposition to any system of voluntary notification (underlined by Prof. Monti in his opening statement), requests for such a system were hardly voiced any more. But the need for individual positive decisions and instruments equivalent to comfort letters continued to be emphasised. Discussions at the Workshop showed that insistent request for Commission guidance in individual cases has had a certain effect on DG Competition. According to the Director General, his DG is disposed to adopt and publish, within reasonable deadlines, reasoned opinions on specific genuinely new issues on the basis of available information, provided the issues are not pending before a Court. On the contrary, the Commission remains hostile to enlarging the narrow limits for positive decisions, now called 'non-infringement' decisions. Such decisions, and their legal nature, remained very controversial. While many practitioners insisted that they were necessary or, at least, extremely useful, others expressed doubts about—or even opposition to—such decisions, which can be considered to be contrary to the logic of the new system. The debate about non-infringement decisions is influenced by two other issues. Thefirstconcerns the NCAs: should they be entitled to adopt such decisions? The Commission remains totally opposed to this perspective, as it apprehends that NCAs might retain divergent interpretations of Article 81 (3). Insofar, the Commission shares the fear that EC competition policy might be re-nationalised. Others, however, consider it to be normal that the process of decentralisation does not stop at non-infringement decisions. The second major problem raised by non-infringement decisions concerns their legal nature. For some, they can only be declaratory in nature, not having any binding effect, at least on national courts. Others considered that they should be binding even under the legal exemption regime. In support of this view, reference was made to general EC law principles, like the duty of cooperation and mutual respect (Article 10 EC Treaty), and the supremacy of EC law. Even representatives of the Commission seemed to be divided on this issue, which was said to be still open among Commission departments. Participants agreed, however, that the formal amendments to Regulation 17/62 should provide an answer to this question.

xxii

Introduction

Panel 2 was devoted to the risks which result from the radical decentralisation of the enforcement of Article 81 for the coherence of EC antitrust law, and the mechanisms envisaged in the White Paper to minimise these risks. It is inevitable that an increase in the number of agents entrusted with the implementation of Article 81 (3) will lead to divergences in the interpretation and application of this provision. Opinions were divided as to the importance of these divergences. Participants from the US seemed least concerned, as they are used to a multitude of public and private actors in the field of antitrust enforcement. They, but also others, considered diversity of solutions to a given problem as an opportunity to arrive at the best result. For them, a 'bottom up' approach in which 'the law bubbles up' is a familiar phenomenon which is one of the great merits of a federal structure. After all, nobody guarantees that a central agency like the Commission has the monopoly of insight and wisdom in enforcing Article 81. Others considered that, in the light of decisions taken so far by national courts under the already directly applicable Articles 81(1) and 82, the risk of divergent decisions will be relatively small. This view is shared by those who do not expect a significant increase in the use of Article 81 by NCAs and national courts. However, there are also those who fear divergent decisions because of the overlapping responsibilities of NCAs, the limited territorial effect of NCA decisions, multiple jurisdictions and forum shopping by private parties. Because of the lack of practical experience with any decentralised application of Article 81 (3), it is difficult to appreciate the importance of these risks. The business community, trying to minimise costs and maximise the benefits of the national market, will probably evaluate them to be higher than political defenders of the subsidiary principle. We have already noted the Commission's hostility to decentralise in favour of NCAs the possibility to adopt non-infringement decisions under Article 81 (3), because it fears a re-nationalisation of the interpretation and application of this provision which would lead to too generous results. It is noteworthy that one of the papers prepared for the Workshop points to the risk that decentralisation might produce the opposite results, i.e. a too strict interpretation and application of Article 81 (3), as NCAs might not take into account the positive effect of certain agreements if these effects present themselves outside the territory for which the NCA is responsible. Consequently, this paper proposes to impose on NCAs a positive comity obligation. There is no doubt that the Commission is determined to maintain its central role as policy-making agency, defining and implementing the future course of EC antitrust policy. Nobody contested the need for such a role, which only the Commission can perform. Participants expressed, however, more or less doubts as to whether the mechanisms envisaged in the White Paper will be sufficient to preserve this indispensable function. The Commission suggests relying essentially on two broad principles in order to minimise the risks resulting from the radical decentralisation of the application of Article 81 (3). The first is co-operation within the network of enforce-

Introduction

xxiii

ment agencies formed by the Commission and the NCAs. The second is the principle of supremacy of EC law over national law. Participants expressed broad support for the main lines of the Commission's approach. They expressed, however, also some concerns, and asked for clarifications or suggested additional mechanisms. The spectrum of opinions was large. It spread from rather general criticism (like 'weaker part of the White Paper', 'lack of detail', insufficient distinction between the problems related to NCAs as opposed to those concerning national courts) to very detailed proposals on how to avoid conflicts and to restore coherence in case of diverging decisions. As the problems of national courts were the subject of a special, third session, the following passages will concentrate on issues pertaining specifically to NCAs. The risk of divergent decisions would be reduced if cases were attributed, within the network of enforcement agencies, according to precise, clear and transparent criteria. As one participant observed, such rules should address not only situations in which several NCAs claim to be competent, but also the opposite problem, i.e. that none of the NCAs wants to deal with a complaint. Whether the criteria envisaged so far are satisfactory remained controversial. They are certainly the subject of ongoing in-depth discussions between the Commission and NCAs. Information about these discussions might have influenced the position of those participants who expressed dissatisfaction with the Commission's approach. While participants certainly agreed in general with the position taken by the Head of the Dutch competition authority, who pleaded in favour of preventive mechanisms for conflict avoidance (information and consultation, instead of approval), others emphasised the continuing need for the Commission to exercise its right under Article 9 of Regulation 17/62 to take over cases, thus assuming responsibility for their outcome. Reference was made in this respect to the functions exercised by the US Supreme Court in settling divergences of opinion among US Circuit Courts of Appeal. Some participants floated even the idea to envisage a right of appeal of the Commission against decisions taken by NCAs. The suggestion to allow such an appeal before the Court of first Instance (CFI) would of course affect the architecture of judicial review under the EC Treaty, and therefore would require an amendment of the Treaty. More generally, it should be noted that Commission interventions in cases pending before a NCA raise problems under the principle of subsidiary. In addition, 'hard' as opposed to 'soft' co-ordination could very well become a disincentive for NCAs to apply EC competition law, instead of using, by preference, national competition law, as they seemed to do until now. The White Paper concentrates on vertical co-operation between the Commission and NCAs. It hardly discusses horizontal co-operation among the NCAs themselves. However, such horizontal co-operation is indispensable for the efficient functioning of the network of administrative enforcement agencies within the EU. During the Workshop, the White Paper's weakness in this respect was observed, and reference was made in particular to: the transfer of

xxiv

Introduction

cases between NCAs; the exchange of confidential information, complicated through confidentiality requirements (which might be particularly acute in the case of information provided in the framework of national leniency programs); the already mentioned positive comity principle; and even the possibility that NCAs might collect evidence in other Member States. The White Paper assumes, without further discussion, that the effect of NCA decisions will continue to be limited to the national territory. Several participants observed that such limitation is contrary to the principles of the internal market. One of them suggested to install a 'conditional' EU-wide effect, the condition being the absence of objections raised by the Commission or another NCA. However, specific requirements for the recognition of EC-wide effect were not discussed. In general, most participants seemed to accept that, at least for the time being, the approach of the Commission is appropriate. The White Paper contains no suggestions for minimum procedural standards that NCAs have to fulfil before decentralisation of the application of Article 81 (3) takes place. That all of them have to be entrusted with the possibility to apply Article 81 (in its entirety) is obvious. The new Regulation replacing Regulation 17/62 should make this absolutely clear. But should they go further and provide for the approximation of minimum requirements for NCAs, like the status of NCAs, their procedures, their instruments of action (such as the possibilities of settlement, the nature and size of sanctions)? A few participants considered that some approximation of national legislation regulating the status, the rights and obligations of NCAs was necessary or, at least, highly desirable. Most participants, however, seemed to agree with the Commission's position that, at this stage, it would not be wise to take such an initiative. It is not surprising that the White Paper does not refer to the substance of national competition law, as the reform deals exclusively with the enforcement of existing EC antitrust law. The substance of national competition law remained also outside the discussions of the Workshop. However, the observation was made that differences between Article 81 (3) and similar, but either narrower or broader national provisions could have an influence on the interpretation and application of Article 81 (3) by the NCAs. The co-existence of EC and national rules, and the consequent 'spillover' from national to EC law could, be one factor leading to conflicting interpretations. We have already mentioned that the risk of such conflicts motivates the White Paper's opposition to the decentralisation of non-infringement decisions to NCAs. The decentralisation of Article 81 (3) is already a challenge for the NCAs of the existing Member States. It is even more such a challenge for the NCAs of acceding new Member States. The so-called Europe Agreements have obliged Central and Eastern European candidates to EU membership to apply the criteria of Article 81 and 82, and to harmonise their national competition law with existing EC competition law. As a result, the existent notification and prior administrative authorisation system was taken over in the national laws of the candidate countries. The reform of EC law should logically lead to a corresponding change of national law in Central and Eastern Europe. Whether

Introduction

xxv

administrators (and judges) in these countries are sufficiently equipped and qualified to follow the change of systems in the EU remains to be seen. Transitional arrangements (not discussed in detail at the Workshop) might be the appropriate answer. Contrary to the situation under the Europe Agreements, no corresponding harmonisation obligation exists for Member States of the EU. Member States are, therefore, perfectly entitled to maintain their notification and prior administrative authorisation systems, often inspired by the existing EC regime. Participants who specifically addressed the issue recognised, however, that the reform of EC competition law would probably have repercussions on corresponding national law. It is indeed not logical to maintain notification and authorisation requirements taken over from EC competition law at national level, if such requirements are eliminated in EC competition law. It is, however, remarkable that the likely pressure to amend national law was not invoked by anybody as an argument against the reform of the EC antitrust enforcement system. The reader will remember that the principle of supremacy of EC law is for the Commission one of the two main defences against the risk of divergent decisions in a totally decentralised enforcement system. The discussion on the legal effects of non-infringement decisions with respect to NCAs and national courts is—at least in part—a debate about the scope of supremacy of EC antitrust law in the new context of a legal exception system. In addition, the extent to which Article 81 pre-empts the application of national antitrust law is of fundamental importance for the possible reach of such national law. The Workshop demonstrated wide differences of opinion on this question. While representatives of the Commission considered that the reform will enhance the area pre-empted by Article 81, others took the opposite view, i. e., that under the new system, the logic of existent ECJ jurisprudence would leave more scope for national antitrust law. A particularly important point was made by the Head of the Dutch competition authority: if the Commission representatives were right, and the simple fulfilment of the conditions of Article 81 (3) by a given agreement were to preclude the application of national law, the burden of proof with respect to Article 81 (3) might shift to NCAs; such a result would be unacceptable in view of the general consensus that the existing situation with regard to the burden of proof should not be affected by the reform. Participants agreed that, in view of the uncertainties surrounding the supremacy question, it was urgent to give it a clear response in the new Regulation. Another essential point emerging from the debate referred to the Delimitis principle of 'deference' of national courts vis-a-vis the Commission. In particular, some of the participants questioned whether the deference principle remained in place once the Commission's monopoly over the application of Article 81 (3) was abandoned. It was suggested that this issue, closely related to the need for ensuring a coherent application of EC antitrust law in a system of parallel competencies, needed to be clarified in the new Regulation.

xxvi

Introduction

Panel 3 examined the particular problems arising from the envisaged reform for courts and judges. We have already discussed the first and most important issue for national courts, i.e. the question whether Article 81 (3) is capable of having direct effect. The reader will remember that the majority of participants considered that this question had to be answered in the affirmative. Judges are perfectly able to apply Article 81 (3), which is neither fundamentally different from other provisions of the EC Treaty that have direct effect, nor from national law requiring the appreciation of complex economic situations and the balancing of competing economic and non-economic interests and values. We have also mentioned the opinion of some participants that the Commission might overestimate the role that national courts will play under the new system of decentralised enforcement of Article 81. NCAs and national courts have different perspectives and functions. Taking into account European traditions and the lack of US-type incentives for private action under EC competition law, administrative antitrust law enforcement will remain considerably more important than court decisions obtained by private plaintiffs or defendants. The application of competition law is fact-intensive. Under the new, more economically oriented interpretation of Article 81 (1), even more factual evidence will need to be considered. Procedural rules for court proceedings differ widely within the EU. They converge, however, on one important point. In civil (as opposed to administrative) law proceedings, the facts have to be alleged and proved by the parties. They are not investigated ex officio by the judge. For powerful players with deep pockets this is much less of a problem than for their financially weaker adversaries, in particular if financial strength is combined with otherwise easier access to pertinent information, like for the members of a cartel. For opponents of the Commission's reform suggestions, these are important arguments against the decentralisation of the application of Article 81 (3) to national courts. Supporters of the reform recognised this difficulty, but stressed that judges may need help and assistance from the Commission or NCAs. In addition, the natural limits of judicial procedures, particularly for weaker parties, militate in favour of the continued involvement of the Commission (and NCAs). One of the major objections of those who are opposed to the reform is the possible effect on the distribution of the burden of proof. According to the settled case law of the ECJ, under the existing system the party invoking Article 81 (3) has to prove that the conditions of this provision are met. Opponents to the reform assert that under the new system the burden of proof would shift to the party who alleges that the agreement is prohibited, and, therefore, illegal and void. This party would not only have to prove that the conditions of Article 81 (1) are fulfilled, but also that those of Article 81 (3) are not met. A shift of the burden of proof for Article 81 (3) would have major consequences for the Commission, NCAs, and in particular for private actors invoking the illegality of an agreement, who are often the weaker, less well informed parties to a dispute. Participants agreed that a shift of the burden of proof was neither

Introduction

xxvii

intended nor desirable. It was generally recognised that the formal proposal for the new Regulation should clarify this point. The White Paper suggests to consolidate the traditional—but until know rarely used—practice of the Commission assisting national courts, at their request, with relevant information. The White Paper envisages also that the Commission might act in certain court proceedings as amicus curiae. Participants echoed the well-known criticism that the process of 'information' of judges by the Commission might conflict with the requirements of a due process. A certain preference was therefore voiced for the status of amicus curiae, widely practised in the US and praised by the transatlantic participants. However, continental European legal systems do not provide for such a status. Participation as 'intervener' might be more appropriate. In addition, it was observed that the exercise of such procedural rights requires a good knowledge of national judicial proceedings. It was, therefore, suggested to assign this function to NCAs rather than to the Commission, which has to spare in any case its rare resources. That the amicus curiae suggestion is, once more, one of those that need to be elaborated further, is obvious. Reference was also made to a suggestion advanced in the literature, namely to establish within each Member State an EC law Advocate General to advise national courts on questions of EC law in general, and EC antitrust law in particular. Participants welcomed in general the suggestion to concentrate competition law issues in certain, specialised jurisdictions. Such specialised jurisdictions exist in several Member States (but not in the country with the oldest and richest antitrust tradition, where the law is exclusively enforced through courts, i.e. the US). However, attention was drawn to the fact that competition law issues arise, more or less incidentally, in a wide variety of civil proceedings for which concentration in specialised courts might not be desirable. Instead, the specialisation of individual judges, at the appeal court level in particular, might be preferable. Training of judges was of course welcomed. Even better would be the exchange of judges between Member States—but these proposals go far beyond the limited area of EC antitrust law. We have already mentioned the problems that the decentralised application of Article 81 (3) will raise for NCAs in accession countries. Such problems will, of course, also arise with respect to judges of these countries. Transitional arrangements might, again, be the appropriate solution to attenuate these problems. In relation to the problems of overlapping jurisdiction, participants expressed in general the view that the provisions of the Brussels Convention are sufficient to deal appropriately with the problems of forum shopping. But not everybody agreed. Forum shopping might not only be influenced by differences in substantive and procedural national laws, as well as in interpretation and application of Article 81, but also by purely administrative factors, like speed and language. Most courts are notoriously overburdened; the language issue is an even greater problem for the national judiciaries than for national administrations.

xxviii

Introduction

It is obvious that the risk of diverging decisions does not only exist at administrative level. It is even greater for the much more numerous national courts. According to existing Treaty rules, preliminary rulings by the ECJ under Article 234 should avoid or, at least, reduce this risk. Transatlantic participants explained the much more efficient mechanisms provided by US law for coping with problems of 'judicial federalism'. Establishing similar mechanisms in the EU would require far-reaching modifications of national and EU constitutional law. We have already mentioned one suggestion, made with respect to the judicial control of NCA decisions by the CFI. Other suggestions were advanced with respect to preliminary rulings and national courts. One of them concerns the responsibility for giving preliminary rulings. In line with reform ideas advanced by the ECJ, the responsibility for giving preliminary rulings in antitrust matters might be attributed to (or shared with) the CFI. However, preliminary rulings have to be more or less abstract answers to strict legal questions. Such answers were generally considered not to be necessarily helpful in fact-intensive competition cases. This aspect might explain, at least in part, the rather bold suggestion to submit national court judgements to the control of the CFI (and the ECJ). Critical voices considered that the Commission has overlooked the repercussions of the envisaged reform on the judicial architecture of the EC Treaty. It is, however, more likely that the Commission did not want to add to the already numerous and delicate problems mentioned in the White Paper. This consideration explains probably also why the White Paper does not mention differences in Member States' procedural laws, and why it does not even hint to any initiative that would lead to the approximation of such laws. The Workshop seemed to agree on this cautious approach. Participants drew however attention to an issue which, in their view, has to be addressed, i.e. the growing importance of arbitration proceedings. EC competition law is considered to be arbitrable, but belongs to the group of rules that are in the nature of ordre public provisions. Arbitrators are obliged to take into account EC antitrust law issues, because otherwise the party losing the dispute might challenge the validity of the arbitral award in a national court. However, competition law questions coming up in arbitration proceedings cannot be submitted to the ECJ for preliminary rulings, as arbitration tribunals are not courts in the meaning of Article 234. Should arbitrators be given the right to address requests for preliminary rulings to the ECJ? Should they be given the right to request opinions from the Commission or from NCAs? These questions were raised, but not answered. It was, however, considered that the Commission should consider them carefully before it puts forward its formal proposal for the reform of the EC antitrust enforcement system. Before we leave the specific problems of courts and judges, a final point has to be mentioned and which was hotly debated. The White Paper considers that even after afinaljudgement given by a national Court, the Commission should be able to take up the case and decide differently (e.g., denying the applicability of Article 81 (3), provided that the resjudicata effects between the parties of the

Introduction

xxix

court judgement remain unaffected). A considerable number of participants expressed concern about and even opposition to such an initiative, which seemed to them to be incompatible with constitutional principles such as the separation of powers. The preceding remarks may have raised doubts about our initial observation that a broad consensus seems to have emerged around the Commission's reform suggestion. We would like, however, to confirm our initial statement. The great majority of participants expressed their overall support for the Commission's initiative. Their critical remarks concerned—some times very important— details, but rarely the main thrust of the reform. They warned against a premature and hasty conclusion of the ongoing debate, leading to an insufficiently prepared and thought out proposal that would leave open crucial questions. The reform was generally considered to be too important to be jeopardised by unsatisfactory preparatory work. The Workshop was thus a contribution on the long march to a new, modernised and more efficient system of enforcement of EC antitrust rules. Claus Dieter Ehlermann Isabela Atanasiu EUI Florence, Italy 30 August 2000

PANEL ONE COMPATIBILITY, EFFICIENCY, LEGAL SECURITY

1 OPENING STATEMENT

Opening Speech

THE MODERNISATION OF EC ANTITRUST POLICY Mario Monti Ladies and gentlemen, I am honoured to be here today to give the opening speech at this EC competition workshop on the modernisation of EC antitrust policy, which has brought together such an impressive number of distinguished scholars, practitioners and policy-makers. I am confident that the contributions to, and the discussions of, this workshop will deepen our understanding of the many complex issues raised by the White Paper on modernisation of the enforcement system set out in Regulation 17/62. This will lead us towards our final objective, namely, to replace the present system with a new system of enforcement that will ensure the effective and coherent enforcement of EC competition rules in the coming years. In my introduction this morning I will briefly explain the main aims of the reform proposed by the Commission in the White Paper, and I will give you my thoughts on some of the concerns that have been voiced in respect of the proposal to fundamentally change the present Regulation 17, which was adopted in 1962.

I. Regulation 17/62 and the present environment The enforcement system established by Regulation 17/62 has certainly served us well for many years. In particular, it has allowed the Commission to develop a coherent competition policy for the whole of the Community, contributing greatly to the creation of the internal market. Over the years we have come to think of Regulation 17/62 as a stable and almost sacred companion in an otherwise turbulent world. To change it was almost unthinkable. However, the fact that the Commission is now thinking the unthinkable is a clear indication of the crucial importance that the Commission attaches to this matter. We must adapt Regulation 17/62 to meet the challenges of today and tomorrow. As we all know, the world has changed dramatically since 1962. Particularly in recent years the changes have exceeded even the most far-reaching predictions. Globalisation is increasing and markets evolve at great speeds. The Internet is revolutionising the way that we do business. The introduction of the Euro is creating more transparency and is facilitating capital movements and cross-border transactions. Enlargement will not only considerably expand the Community, but it will also present great challenges in terms of adapting the economies of the applicant countries.

Opening Speech

All these changes and challenges create demands concerning the development of competition policy. Competition law and policy are of fundamental importance to the European Union, European consumers and European industry. The very foundation of the European Union is an open market economy. As is now generally accepted, a competitive economy not only creates benefits for consumers and our societies as a whole, it also ensures the optimal functioning of the Internal Market and the competitiveness of European industry in an increasingly world-wide arena.

II. The overall reform of EC competition policy Considering the importance of competition policy, it is essential that we continuously monitor its effectiveness. The Commission believes that by adapting our rules we can considerably enhance their effectiveness for the benefit of European consumers. That is why the Commission has launched a comprehensive reform of our overall competition policy, reaching from the substantive to the procedural rules applying in thefieldof antitrust, mergers and State aid. The modernisation of Regulation 17/62, although highly important in itself, should not be seen in isolation. It is rather a part of this much larger reform. These reforms pursue the same fundamental objectives, namely: the more efficient enforcement of EC competition rules, less bureaucracy for companies, and the creation of a more level playingfieldfor companies through the development of a common competition culture and more application of Community law.

III. The problems related to the current enforcement system Before dealing with some of the main elements of the White Paper, such as efficiency, coherent application and legal certainty, it is perhaps useful to briefly recall the existing enforcement system and its main perceived deficiencies. Under Regulation 17/62, the prohibition rules of Article 81 (1) and Article 82 have direct effect and can therefore be invoked before national courts. National competition authorities similarly have the power to apply these provisions. On the other hand, the Commission has the exclusive power to grant exemptions under Article 81 (3). An agreement can only benefit from the exemption rule if the Commission has been notified of it and it has received the benefit of an exemption decision. For a long time the Commission has actively sought to encourage the application of EC competition rules by national courts and competition authorities. However, in practice these efforts have been largely in vain. The Commission's

1 - Compatibility, Efficiency, Legal Security

monopoly under Article 81 (3) is a significant obstacle to the enforcement activities of national competition authorities and courts. This obstacle follows from the fact that companies alleged to be in breach of Article 81 (1) quite naturally claim that their agreements meet the conditions of Article 81 (3) and are therefore legal. As the national bodies have no power to apply Article 81 (3), the continuation of their enforcement action is made very difficult. In most cases, a national action must be suspended pending the Commission's decision. This leads to unnecessary delays and creates a clear disincentive to apply EC competition law. The present system has led to a situation where the Commission is virtually the sole enforcer of Community law in an enlarging Union, and where its resources in the area of antitrust are substantially tied up in the handling of notifications. Of course the latter fact would not be a problem if notifications revealed the cases that merit investigation. However, that is not the case. The clarity of the law has increased substantially since 1962, when Regulation 17 was adopted. Companies and their lawyers therefore generally know what is prohibited and what is likely to be exempted. Only the latter cases are 'notified'. Unfortunately, we have good reason to believe that some companies - too many in fact - also engage in practices that are prohibited by the competition rules. For obvious reasons the Commission is never notified of these practices; they are hidden ever more ingeniously from the eyes of the competition authorities. They are therefore resource-intensive to uncover and prosecute. Naturally, the Commission tries both to cope with the notifications and to investigate cartels and other prohibited practices. However, the use we make of our resources is inefficient in terms of protecting competition. The current system therefore no longer ensures effective enforcement of the competition rules. This is a fact that, in my view, clearly justifies fundamental reform.

IV. Efficient enforcement To promote effective protection of competition the new system must enable the Commission to concentrate its resources on the cases that present a real threat to competition to the detriment of consumers. The Commission must step up its prosecution of cartels; of other hard core restrictions such as resale price maintenance, market sharing, restrictions on parallel trade, and exclusive dealings leading to market foreclosure; of state measures that distort competition and of abuses of dominance such as tying, refusal to supply andfidelityrebates. However, in a territory the size of the Community, it is not possible for a single authority to alone enforce the rules effectively. It is also necessary to involve both the national competition authorities and the national courts in the enforcement of EC competition law. In order to do so it is necessary to abolish

Opening Speech

the notification system and the Commission's monopoly over Article 81 (3). This will allow each body forming part of the enforcement system to effectively apply the rules and to concentrate on what it does best. Each authority will have the power to act on the basis of the Community competition rules in respect of agreements or practices affecting trade between Member States. However, the aim is to establish a network of competition authorities with rules on the allocation of cases so as to avoid multiple control, duplication of work or inefficient actions. Generally speaking, parallel action by several competition authorities is not resource efficient since it implies parallel administrative procedures. Furthermore, it could impose significant additional burdens on companies. The basic principle should therefore be that of 'single action'. Effective enforcement by national competition authorities could often require co-operation with, and assistance from, other members of the network. We must therefore establish the necessary basis for horizontal co-operation between the national competition authorities, including the exchange of information and mutual assistance. However, it is also important to set out proper and effective rules for the protection of the rights of undertakings, including rules on confidentiality. Undertakings have a legitimate interest in the protection of the confidentiality of any information exchanged. Both the Commission and national competition authorities should therefore be required not to disclose information collected or received that contains business secrets or is otherwise of a confidential nature. Other safeguards might also be necessary to protect the legitimate interests of companies. Let me now turn to the national courts and their function in the envisaged new system. National courts rule on disputes between two or more parties. Unlike competition authorities, which act in the public interest, the central function of courts is to safeguard the rights of private individuals. Private enforcement by national courts is an essential complement to public enforcement. To realise this one needs to look no further than other areas of Community law. Enforcement of Community law by national courts is the rule and not the exception. National courts have played an essential role in shaping our Community into what it is today: a unique legal order that grants important rights to the citizens. This is important for all of us to bear in mind. Due to the Commission's monopoly on Article 81 (3) we have become accustomed to the courts playing a rather limited role in the antitrust field. Habit, however, should not stand in the way of progress. In the European Union of today I see no valid justification for excluding national courts from the full application of Article 81. This is even more so in the context of the more economic approach adopted by the Commission. Under an economic approach based on the effects of the practice on the market rather than its legal form, the current division of competence under Article 81 is wholly artificial. At present national courts are empowered to assess the negative effects of an agreement due to the direct effect

1 - Compatibility, Efficiency, Legal Security

of Article 81(1), whereas another body - namely the Commission - must assess the positive effects of the very same agreement. In practice these issues are closely related and one is not more complex than the other. Consequently, they should not be separated.

V. Coherent application The Commission is aware that empowering national competition authorities and national courts to apply Article 81 (3) involves a certain risk of inconsistent application. However, this risk can be addressed and diminished by appropriate measures. I must emphasise that associating national authorities and courts with the enforcement of Community rules is not a re-nationalisation of European Competition law. Far from it: over time, the Commission's proposal will lead to the creation of a common policy and a common approach based on Community law, reducing the scope for inconsistencies. Consistent application requires that the rules are sufficiently clear and that the approaches followed in their application are sufficiently similar. Clarification of the antitrust rules will be an important function of the Commission in the new system. In addition to facing up to this responsibility to develop clear rules and to make public its enforcement policy by means of guidelines and other appropriate instruments, the Commission is envisaging a number of other specific instruments that aim to preserve consistency throughout the system. With regard to the national competition authorities, the aim is to establish a network of competition authorities within which the members increasingly apply the same body of rules, instead of the current situation in which 16 authorities apply 16 (sometimes different) sets of rules. The aim is to widen the application of EC competition law. The application of a common set of rules will greatly promote consistency and the creation of a level playing field throughout the Community. The Commission will remain at the centre of the network and will play a leading role in defining policy and ensuring consistent application. Importantly, the Commission also proposes to retain the current Article 9 (3) Regulation 17/62, according to which the Commission can withdraw a case from a national authority. The network of competition authorities will be a forum for discussion about the common policy and for coordination about cases and issues of common interest. In particular, it is important to ensure that there is prior consultation on prohibition decisions, decisions accepting commitments and decisions withdrawing the benefit of a block exemption. These types of decision merit special attention because they can only be reversed with difficulty once they have been adopted and enforced. Consultation on these types of decisions is

Opening Speech

also important in order to maintain a coherent competition policy, since policy in the new system will, to a significant extent, be developed by means of prohibition decisions and decisions with commitments. Let me now say a few words about consistency as regards national courts. As already mentioned, our proposal essentially means that Article 81 will catch up with the rest of Community law, where national courts already play an essential role. In all these areas, the Court of Justice has been able to maintain coherence and consistency via the Article 234 preliminary reference procedure. The Court of Justice will play the same essential role in the application of Article 81 as a whole. However the Commission is proposing additional instruments. In the 1993 notice on co-operation with national courts, the Commission provides assistance to national courts upon request. This instrument will be maintained and made more efficient by the introduction of a deadline for reply. We expect that the incentive for national courts to obtain input and opinions from the Commission will grow in the future system as they will get the power to apply Article 81. The application of EC competition law is frequently a matter of law and fact. In such cases, the Commission might be able to draw on its own administrative practice to provide valuable assistance to national courts. The Commission is also envisaging a mechanism whereby it could intervene more actively in court proceedings, allowing it to defend the common competition policy before national courts. One such instrument, which is mentioned in the White Paper, could be the right to intervene as amicus curiae. The Commission would then, on its own initiative, submit written observations and/or oral observations to the court hearing the case. Finally, let me turn to the issue of forum shopping and multiple control. Multiple control and forum shopping before national courts should of course be avoided. However, in this respect we are fortunate to be able to rely on the rules of the Brussels Convention. The Brussels Convention establishes rules about how to determine the proper forum in commercial and civil cases. The purpose of the Convention is to prevent parallel proceedings before the courts of different Member States and to avoid conflicts between decisions that might result therefrom. The Convention confers jurisdiction on a limited number of courts that are territorially well placed to deal with the dispute. Once an action has been brought before one national court, all other courts must decline jurisdiction provided that the proceedings concern the same parties, the same cause of action and the same subject matter.

VI. Legal certainty In an open market economy, companies should be encouraged to enter into pro-competitive transactions. Competition law should not only discourage

/ - Compatibility, Efficiency, Legal Security

anti-competitive practices, it should also encourage those practices that promote competition and consumer welfare. An adequate level of legal certainty is therefore a legitimate and necessary objective. The Commission's White Paper proposal, while aiming to restore the primary task of effective enforcement, fully recognises the need to provide companies with an appropriate degree of legal certainty. That said, I must however emphasise that I see no justification for introducing a voluntary notification system, as is sometimes called for by industry. A voluntary notification system would in effect reinstate the present system and would undermine the principal objective of the reform, namely effective enforcement. It would prevent the Commission from focussing on the cases that present a real threat to competition and would also perpetuate the blocking effect of the current system on the application of EC rules by national courts. Companies must accept responsibility for assessing their own agreements. In fact, most companies already do that, since the Commission is not notified of most agreements. What companies can reasonably expect, however, is an adequate level of predictability and consistent application of the rules that allows them to properly assess how the rules will be applied. In the future, the Commission will put more emphasis on clarifying the rules by focussing on what is prohibited. In fact, the Commission has already started to overhaul the legal framework and to further clarify the rules, thereby facilitating self-assessment. The Commission will also apply a more economic approach that, for many companies, will reduce the risk of being caught by Article 81 (1). Block exemption regulations will be maintained. However, as is clear from the new block exemption regulation on vertical restraints, the Commission is abandoning the present form-based regulations in favour of regulations based on economic effect. This allows the Commission and the national authorities to concentrate on the important cases where the parties have market power, meanwhile creating legal certainty for the vast majority of companies. The Commission will also issue a number of guidelines and this general body of rules will be further clarified by the decisions in individual cases. The Commission will continue its current policy of imposing fines only in cases where it is clearly established that certain behaviour constitutes an infringement. Consequently, there is no risk for undertakings to be fined without having been able to know that they were infringing the law. One should also keep in mind that the direct applicability of Article 81 (3) will, in itself, promote legal certainty, since it massively legalises agreements which fulfil the conditions of Article 81 (3) without the need for a prior Commission decision. In addition, it provides companies with a new defence against claims based on Article 81 (1). At present, only an agreement of which the Commission has been notified, and which it has exempted by a formal decision, can benefit from Article 81 (3). On balance therefore, civil enforceability of agreements will be improved in comparison with the existing system. In addition, the Commission considers that it would be appropriate to introduce a new instrument that would allow it to provide guidance to undertakings

10

Opening Speech

in cases where there is real doubt as to the application of the competition rules. What we have in mind is a system of opinions whereby companies could put questions to the Commission in cases where the existing general rules, measures and case practice do not provide sufficient guidance. Companies would be required to submit a supporting and explanatory memorandum alongside questions they put to the Commission. On that basis the Commission would issue an opinion in appropriate cases. Such opinions would be reasoned and published, and they would therefore contribute to the overall clarity of the rules. Accordingly, they would not only provide guidance to the companies asking a question but to other undertakings as well. The value of this instrument should not be underestimated, particularly when compared with the current system of comfort letters that are neither reasoned nor published. In this respect, it is important to take account of one of the effects of our more economic approach under Article 81. When one takes a global view of the Commission's proposal and takes due account of the impact of our reform work in the substantivefield,I believe that there will actually be more coherence, predictability and legal certainty in the proposed new system than in the current system.

VII. Conclusion Ladies and gentlemen: let me now conclude. The Commission is aware that reforming the enforcement of European competition rules poses major challenges. I have already mentioned some of these. However, the risks of inaction and complacency are even greater. We must adapt the current system to the radical change in the legal, economic and political environment in which it has to be applied: market integration, the Euro, successive enlargements and globalisation. In so doing we must devise a new system that will live up to the expectations of Europe's citizens. Our primary objective must be to strengthen the enforcement of our rules. I wish you much success in the study of the complex issues that are facing each Panel. The discussions will no doubt improve our understanding and assist us in constructing a new antitrust enforcement system that will ensure the success of EC antitrust policy in the future.

PANEL ONE COMPATIBILITY, EFFICIENCY, LEGAL SECURITY

1 PANEL DISCUSSION

PARTICIPANTS

Carl Baudenbacher Simon Bishop Ulf Boge Jochen Burrichter David Edward Claus-Dieter Ehlermann Ian Forrester Eleanor Fox Barry Hawk Ulrich Immenga

Giuliano Marenco Santiago Martinez Lage Mitsuo Matsushita Mario Monti Damien J. Neven Alexander Schaub Anne Spiritus-Dassesse Giuseppe Tesauro Karel van Miert

1 - Compatibility, Efficiency, Legal Security

Panel One: Compatibility, Efficiency, Legal Security • KAREL VAN MIERT—Thank you very much indeed for offering us this very clear view on behalf of the Commission. This is certainly very valuable for our further discussions during the Workshop, and also for the Commission itself. The 'input' of the participants around this table will certainly be considered very carefully by the Commission, particularly in view of some of the controversies and contradictions that have surrounded the reform proposal. We have now about ten minutes left for discussion or questions on your contribution as the main speaker of today, and then we will start with Panel One. Obviously, Alex Schaub will come back later to some of these issues and present the views of the Commission. As far as the procedure and timing of the reform is concerned, we have learned that there will probably be a formal Commission proposal by the end of September. • MARIO MoNTi^Before the end of September a proposal will be submitted to the Council. • KAREL VAN MIERT—So this is a very timely meeting, as it takes place even before the formal proposal is on the table. • DAVID EDWARD—Could I ask whether consideration has been given to enforcement in arbitration proceedings, and other alternative dispute resolution procedures? • MARIO SIRAGUSA—One of the big questions I address in my paper is: why should the reform induce national competition authorities to apply EU law? They are currently applying national laws very successfully, and these are themselves inspired by EU law. They apply the case law of the Commission and the European Courts, but in turn this is done by applying national law. I do not see anything in the reform that will induce national competition authorities and national judges to apply EU law. The fact that they can in theory apply Article 85 (3) is not enough in itself to induce this change in the Community environment. • SANTIAGO MARTINEZ LAGE—Does the Commission expect the Member States that copied the EC system of notification and prior administrative approval to also modify their national system? Or could they work with a system of legal exemption at the Community level combined with one of individual exemptions at national level?

13

14

Panel Discussion

• MARIO MONTI—We have the confident expectation that the reform, far from being a re-nationalisation of antitrust enforcement, would contribute to a wider application of Community law. Perhaps where our views differ is on the weight of the present situation, whereby Article 81 (3) is not directly applicable by national bodies, as an impeding factor in relation to the application of Community law. In our view, removing this obstacle would also mean removing an important factor now playing against the application of Community law. I should also stress—and I am sure that Mr Schaub and the other representatives of the Commission here present will argue this in more depth—that I see an indirect effect of this reform playing in favour of a wider and deeper application of EC competition law: this is the 'network effect' that, in our view, this reform will bring about among the different national competition authorities. On the last question, about national enforcement systems that rely on notification and prior administrative approval: will the legal exemption aspect introduced by the reform have an impact on the evolution of the national systems that are presently based on notification? Again, I would like others more knowledgeable from the Commission to comment on this later. My view is that in principle there is no immediate implication in the sense that adoption of the reform should bring about changes in the architecture of national competition laws as far as notification is concerned; though in the medium term we would expect national competition laws to take stock of any improvement we expect to be brought about at the Community level, so that there could be some change induced by 'imitation' if the Community system worked well, as we expect it will. On the first question, related to arbitration proceedings, I wonder whether Mr Schaub or Mr Marenco would like to step in? • CLAUS-DIETER EHLERMANN—I would suggest that we leave this argument for discussions under Panel Three. It is an argument that comes up in several of the written contributions for this Workshop, and it is the first time that arbitration is raised as a serious concern in a meeting like this. • DAVID EDWARD—One of our concerns is that the Commission's White Paper speaks of the European Court of Justice as a mechanism for ensuring consistency in the enforcement of EC antitrust law. At present, however, there is no possibility for arbitrators or people conducting alternative dispute resolution to make references to the European Court of Justice. • KAREL VAN MIERT—This will be addressed tomorrow, during Panel Three, as it is really a very important issue. • CLAUS-DIETER EHLERMANN—Mr Monti mentioned opinions as one of the Commission's replies to criticism about lack of legal certainty for companies in the new enforcement system. This is a sort of compensation for not accepting

1 - Compatibility, Efficiency, Legal Security

the possibility of 'voluntary notifications', that Mr Siragusa, for instance, pleaded and fought for so vigorously last October. Another thing that strikes me is the Commission's reluctance to allow for the so-called 'positive decisions', to be shared under a decentralised enforcement system with national competition authorities. I would be curious to know the Commission's position on sharing this modality of giving guidance to industry (or their lawyers), on whether an agreement is or is not covered by Article 81 (3): should this be decentralised too, or kept close to the Commission? • MARIO MONTI—A first approximate answer is that it would be consistent with the aim of developing a well functioning network of national competition authorities if there was some sharing on this aspect, unless there are technical reasons against such sharing that escape me at this stage. Even before the adoption of this reform, in several areas of EC competition policy, including merger control, there are recent examples of much deeper co-operation between the Commission and national competition authorities than in the past. This is also a sign of developing a common competition culture, which I am sure will be highly beneficial once this reform is in place. May I just add a more general remark. The debate that takes place here is not directly concerned with the linkages between this reform and the position of the Commission as a European institution. Yet I would like to stress, and this is part of my deep inner motivation in pushing politically for this reform, there is great consistency between the reform of Regulation 17/62 and the structural reforms in other areas of EC competition policy, such as mergers and state aid control, that we are currently bringing forward. At the same time, there is great consistency between what we are trying to do in the area of competition and the new role that the European Commission would like to play. I do not want to enter into the discussion on whether the White Paper is, or is not, an example of a march towards subsidiarity. In my view, technically it is so. At any rate, with this reform I see the Commission focusing on its core activities and ensuring a better discharge of its obligations. This is something that, at this historical stage, the Commission would like to do and needs to do for most of its activities. It may be of interest for you to know that my colleagues at the Commission look at this reform in thefieldof EC antitrust enforcement as an interesting example that could perhaps have a broader validity in terms of helping the Commission to narrow down its activities to the core and perform them better. This will be one of the key issues in the White Paper on Governance that the Commission intends to present in the course of next year. • KAREL VAN MIERT—Thank you very much, Mario. Perhaps an additional remark from my side: I think that more coherence between national competition rules and the European ones was the result of a process of 'soft harmonisation'. The Commission always refused to take formal steps in this direction, as we were very confident that, step by step, the Member States would introduce

15

16

Panel Discussion

systems similar to the European one, even Great Britain. I can remember very vividly the time when DGIV and I came to the conclusion that we should propose reform of the EC antitrust enforcement system. At that time, the British authorities were finally changing their own competition system in light of the existent European one. So we went to see the British government and said, 'it's a little bit unfortunate, now you are finally adapting your system to the European one and we are already thinking of changing our own system. We can understand that you might have some second thoughts now about your reforms. We are going ahead with what we have considered, and please, you go ahead as well. If later on we need Europe to change further, you can reconsider it. You can go ahead. We go ahead.' I think that is the way things have to proceed in Europe. Therefore, Mario, I think what has happened previously will continue to happen in the future, but this is just a more general remark I wanted to make on the basis of my personal experience. Thank you very much indeed also for informing us about further steps in the direction of the adoption of the reform. We shall see how this debate continues once the proposal is before the Council of Ministers, but it is good to see that things proceed as scheduled and that the Commission is pursuing this reform as has previously been agreed. We will now continue with Panel One: compatibility, efficiency, legal security—three vital items that are related to this reform. I invite Professor Bishop to start the discussion. As an economist, he will make a valuable contribution to this debate. • SIMON BISHOP—Let me say straight away that, in general, I am in agreement with the overall approach in the White Paper. Indeed, as an economist, I welcome any move towards a more economic effect-based approach in the enforcement of EC antitrust rules. This is not only for commercial self-interest, but also out of the belief that the reform will actually help the implementation of public policy. I have been asked to look in my written contribution at two questions. The first is whether the existing case law provides adequate guidance for the application of EC antitrust rules by national bodies. The second is whether 'positive decisions' are indeed no longer required. It seems to me that these questions are really sort of one, wrapped up together. I think it is a bold statement to say 'we had forty years of case law, and therefore parties and their legal advisors can determine whether something is contrary to Community law or not'. The White Paper is a radical change not only in terms of the implementation of EC antitrust rules. The introduction of an economic based enforcement system means that after the implementation of the White Paper the new EC competition policy will be different from the old one. On this basis, how can the old case law provide an adequate guide for new case law? Indeed, an economic effect-based approach introduces in some sense increasing uncertainty. Barry Hawk says in his written contribution as well that, the more economic-oriented the approach in enforcement, the more uncertainty

1 - Compatibility, Efficiency, Legal Security

17

there will be. An economic based approach to enforcement means basically the implementation of the rules on a case-by-case basis. Therefore, a particular practice that is prohibited in one industry might have pro-competitive effects in a completely different industry, even if implemented by firms with the same market shares, or even higher market shares. In the White Paper there is a sort of recognition that the new policy will be different from the old one, otherwise why should we need Guidelines and Notices for clarifying the new situation? One particular enforcement area dear to my heart is that of firms with high market shares (since these are most, or many, of my clients). Where does the new enforcement system leave them? I would hesitate to use the word 'dominant' in relation to suchfirms.Firms with high market shares are not necessarily anti-competitive, and an economic effect-based approach to enforcement should recognise this. Many agreements that firms with large market share enter into are pro-competitive, so the notion that firms considered 'dominant' cannot obtain an exemption under Article 81 (3) strikes me to be at odds with the new approach. Moreover, since the economic effects of a particular practice can vary from industry to industry, where does this leave afirmwith a high market share? If the practice is prohibited in industry A, does it necessarily mean that it is going to be anti-competitive in industry B? Or, under other circumstances, is industry B completely different? Often it will be the case that the circumstances are not completely different, so we cannot know a priori whether a given practice is anti-competitive or pro-competitive. It therefore seems to me that large firms will need some guidance under the new enforcement system, because it is a bit naive to say that firms know when they are doing the wrong thing. In certain cases they will know, for instance if the firms are engaging in a cartel; but if a firm has introduced a selective distribution agreement, for example, how can it know whether the agreement is contrary to EC antitrust rules or not? Under an economic effect-based enforcement system, it could be either, even if the firm has a high market share. I therefore believe there is a need for 'positive decisions'. I do not know, however, whether such decisions should be taken following 'voluntary notification', or in another way. At any rate, under the new system the Commission should be required to issue a decision whenever opening an investigation, be it a negative or a positive decision. In some sense, this would mirror what happened under the Merger Regulation, where the issuing of decision after decision, both positive and negative, helped transparency, and helped people understand where the policy was going. In sum, an economic effect-based approach to enforcement does not imply that firms with high market shares always do bad things. This is an economic fact, and such firms will need some guidance in the future. • KAREL VAN MIERT—I would now like to invite Mr Boge, the new President of the German Cartel Office, to comment. I have worked for a long time with his predecessor, but unfortunately I do not have the privilege to work with him

18

Panel Discussion

that much anymore. Mr Boge is going to speak in particular about the abolition of the notification system and some issues related to the decentralisation of enforcement.

• ULF BOGE—First, I would like to thank you very much for being invited to this workshop, and to thank Commissioner Monti for the clear picture he gave us on the next steps in the preparation of the reform. As you all know very well, the Bundeskartellamt supports the objectives of the Commission's reform proposal. However, we consider it particularly important that the reform should not lead to any reduction in the substantive level of protection of competition in the European, and indirectly in the national enforcement systems. This aim is the most important for all of us. Mr Marenco dealt in detail with the question of the compatibility of the Commission's reform proposal with the EC Treaty in his written contribution for this workshop. In his opinion, the reform proposal can be realised by merely adopting a new Regulation to replace Regulation 17/62, without amending the EC Treaty. Other lawyers think differently, and I regret that Professor Mestmacker cannot be with us today, because this was one of the main topics he has dealt with in the past. However, I do not want to go into the details of this legal dispute here, and would prefer to concentrate instead on matters of content. In my opinion, there is no doubt that the changes proposed in the White Paper will have a direct and appreciable effect on national competition laws and their application. This is the reason why, at the start, the Bundeskartellamt had major objections to the abolition of the notification and prior authorisation system that had proved successful in German enforcement practice. During consultations between the Commission and the Member States, however, the Bundeskartellamt proposed a compromise solution in this respect. The aim of this proposal is to achieve the necessary transparency under the new system by means of an obligation on the part of the enterprises to simply inform the competition authorities of their co-operation agreements. This information could thereafter be made available on both the web of the national competition authorities and that of the Commission. I would be pleased to go back to this proposal in more detail later on. Another reason for our proposal is that we should find a way to work together with the other competition authorities, and not to work against each other. We need competition advocacy, and we can only do it if we work together. I do not wish to criticise the more economic approach to enforcement suggested by the White Paper and mentioned here, but we are concerned about the ideas expressed in the recent draft block exemption regulations and guidelines for horizontal agreements; those view horizontal agreements as positive in principle, and consider them to have potential anti-competitive effects only when associated with market power. Our concern is that this might considerably weaken the ban on cartels. More careful consideration should be given to

1 — Compatibility, Efficiency, Legal Security

whether we really want to send the signal that horizontal agreements other than pricefixing,production and market allocation agreements are generally acceptable, or are acceptable if they fall below an appreciable market share threshold. As far as the allocation of cases between national competition authorities is concerned, the Commission's reform proposal includes a list of allocation criteria, which certainly appears to be practicable. This would probably lead to results similar to those achieved by an assessment based on economic or competitive criteria. As to ensuring a uniform application of EC antitrust law, this requires a consistent system of types of decisions and appropriate mechanisms of co-operation and co-ordination. The competition authorities of the Member States will then be able to assume more responsibility in the enforcement of EC antitrust law, even if some do not have much enforcement experience. The consultation and co-ordination mechanisms envisaged for both the Commission and the national competition authorities with respect to prohibition and positive decisions or withdrawal of the benefit of block exemption regulations appear to be quite suitable means for achieving enforcement consistency. At the same time, the obligations to provide information should be kept as simple as possible, and decisions should be dealt with in more detail by the Advisory Committee only in really important cases. Questions arise in relation to the Commission's so-called 'positive decisions', which are not exactly in line with the direct applicability of Article 81 (3), and should therefore be abandoned in favour of consistency in the new system. In any case, small and medium-sized enterprises should not be subject to stricter decisions than the large companies, which by their very nature are more likely to be able to point out the existence of a significant Community interest. I also believe that attempting to solve the problem of legal certainty by means of positive decisions and opinions is not without problems. If the purpose of the reform is to encourage undertakings to assume greater responsibility, a certain level of legal uncertainty in the 'grey areas' of enforcement is a logical consequence, or just the other side of the same coin. Of course, this will not prevent competition authorities from continuing to hold, in critical cases, informal discussions with the undertakings concerned. But we should not try to fool ourselves and the business community: the degree of legal certainty attainable by giving binding exemption decisions cannot be achieved with positive decisions, opinions or informal discussions. No detailed investigation has been carried out yet as to how the national courts will apply EC antitrust law in practice, and what the effects of decentralisation will be in this respect. We would welcome it if the competence to apply Article 81 (3) was to be concentrated in a few specialised courts, as has been the case in Germany for many years. France decided to do likewise, but it is unlikely that all Member States wish to follow this course. The Commission made a range of proposals destined to alleviate the potential conflict between decentralisation and the need to ensure coherence in enforcement. Among these are the obligation for national courts to provide information to the Commission about cases involving the application of

19

20

Panel Discussion

Articles 81 (1) and (2), the possibility for the Commission to intervene as amicus curiae in national court proceedings, and perhaps even the right for the Commission to appeal against decisions by national courts. In addition, the Commission intends to lay down guidance on the interpretation of EC antitrust rules by means of issuing 'positive decisions', opinions and policy guidelines. However, such proposals are not without risk. On the one hand, the desired effect of reducing the Commission's workload may unintentionally be endangered by such proposals. On the other hand, the independence of the judiciary cannot be infringed, and the principle of the separation of powers must be observed. The highest national courts and the European Court of Justice will ultimately guarantee that EC antitrust law is applied uniformly. I would like to sum up by saying that the objectives of the reform cannot be all achieved to the fullest extent at the same time. It might therefore be only a question of optimising the results of the reform by introducing an information system such as that suggested by the Bundeskartellamt. Our proposal is meant to ensure that our common goal, which is to guarantee effective competition on the single market, is not jeopardised. • IAN FORRESTER—It was an enormous pleasure to have participated in the first meetings of this workshop, and it is intriguing to see that the 'parents', or 'the fathers of the church', who upheld the traditional order of things at previous meetings, are now suddenly so enthusiastic to reform that those who originally pointed to the necessity of reform have become nervous, alarmed and worried about what the 'children' are getting up to. I argued in the past that there were three possible reforms, or three possible 'families of reform', to address the well-known problems of the existing enforcement system. The first was that the Commission could change its approach to the interpretation of Article 85, the second was that it could share its enforcement responsibilities with others; and the third was that the Commission could take more decisions by changing its own internal procedures. The White Paper proposes thefirsttwo of these reforms, and implicitly— though here I might exaggerate—suggests decreasing the number of its own decisions. I warmly welcome the notion of reform and I want to offer some reassurance to the Commission about why the proposed reforms are broadly constitutional, but I also want to add one or two warnings. The biggest pleasure in writing my paper for this workshop was browsing through the Council's archives and looking—like Giuliano Marenco, although in a sense in competition with him—at what the fathers of the Treaty discussed back in 1956 and 1957. It is plain from reading those documents that they were confused, and I think it's fair to conclude that the text of the Treaty was not intended to limit our interpretation of it in any special way today. The best proof of this is that, in April 1958, the British Embassies in the six Member States wrote to ask: 'What does this mean? Articles 85 (1) and (3) are very confusing. Please ask the European Commission to explain to us.' At the meetings

I - Compatibility, Efficiency, Legal Security

that followed in Brussels, the Italians said: 'Well, very interesting. Let's give a long answer.' The Germans said: 'Let's be brief, and the French said 'Let's not even give a temporary reply, because that would oblige us to give a complete reply later, and we don't know what the answer should be.' Therefore, if in 1958 it was unclear what the Treaty was actually meant to mean, it seems legitimate for the Commission to look at things with fresh eyes in the year 2000.1 submit that the Treaty provisions would pursue three objectives: first, the punishment and prohibition of the unacceptable; second, intelligent examination of doubtful matters with a view to giving prompt guidance and; third, abstinence from interfering in benign matters. I therefore consider that overall the Commission's reform proposals make good constitutional sense. I do, however, have a number of practical concerns. The first is related to the apparently diminished role that the Commission assigns to itself in the new enforcement system. At the moment, the entity best equipped to take EC antitrust decisions is the European Commission. It has the power to gather evidence, it has enormous experience and great prestige; it is in a leadership position in EC antitrust law enforcement. That is the good news. The bad news is that some Member States are not in such good condition. The legal diversity of Europe is in some ways a source of pride, but in other ways it should be a source of embarrassment. It is highly probable that some national competition authorities will do their job well, some will do it badly, and some will not do it at all. This has to be recognised up front when proposing a major reform of the system. Are there reforms the Commission could make, which would help to alleviate the enforcement difficulties that we are familiar with? I believe that the European Commission takes too few decisions in the application of Articles 81 and 82. The Commission produces admirable decisions in thefieldof merger control, but in the area of Articles 81 and 82 it takes maybe a dozen, or fifteen, decisions per year. This is far too few. The Commission thinks about many things, but decides too few. If the Commission were to set itself deadlines, it would be a great improvement. Deadlines of 45 days, or 75 day—it doesn't matter; deadlines after which something is going to happen—a transaction will be approved or whatever. There are examples, notably in the Scottish criminal jurisdiction, where deadlines are a real nuisance, but if you comply by them, the system works well. Deadlines are a spur to efficiency, and not the toleration of the present situation, whereby there are no deadlines for not deciding. At present it is much easier to prevent a decision from being taken than to procure the making of a decision, and it is not uncommon for decisions to take five years before they are delivered, or sometimes even longer. That is far too long. The abandoning of the existing enforcement procedures that work well should be done only if the outcome will be better. I believe that lawyers exaggerate the legal certainty conferred by the present system, but legal recklessness is another matter. If the Commission were to say 'we are not going to get involved in uninteresting cases in the future. We will only get into the 'sexy cases', those which have high profile and high importance, and the others the

21

22

Panel Discussion

Member States should look after,' this would be a waste of the Commission's resources, which could be more effectively deployed by taking more decisions, subject of course to judicial review. I know that judges are 'repulsive, interfering, unreasonable individuals', and lawyers are 'nasty, unfair people who say unkind things about good civil servants . ..' Yet an agency will never work well if its goal is perfection. The Commission cannot achieve a 100% positive record before the European Courts, and if it aimed at that it would not be doing a good job. Therefore short, pragmatic, intelligible decisions as in the field of merger control, which imply taking a position and answer to a particular problem, those are useful decisions that the European Court can deal with in an appropriate way. My last observation is on the subject of notifications. I believe there is a market for authoritative guidance on EC antitrust law, and at the moment the Commission has a dominant position in that market. One of the particularities of this 'market-place' is that few others are well equipped to do the job, and if the Commission were to say 'notifications never, consultation never, look after yourselves', that would be wrong, because an interplay between the outside and the inside is highly commendable in any enforcement field. On the subject of opinions, be they binding opinions or 'guidance' opinions, the Commission implies that these are not too far away from re-introducing a 'reformed' notification system. I believe the Commission is too hostile to the possible perpetuation of the notification system, which is currently flawed because of the lack of deadlines. To sum up, the Commission is to be warmly commended for proposing the reform, and a reform that is 'constitutional', or compatible with the Treaty, but the difficulties of getting to a broader, wider, more consistent enforcement system are underestimated in the White Paper. The reform could be bettered if coupled with an internal revision of the Commission's procedures. • KAREL VAN MIERT—I can tell you one thing that we never did consider: guidelines on how to find out whether cases are 'sexy' or not—unless this was done in DG IV behind my back. But, as far as I know, there is no such document in preparation. Now I would like to invite Professor Eleanor Fox, who is going to tackle the question of legal certainty. Let me quote one sentence from her Paper, 'I believe that it [the reform] can be implemented efficiently with no meaningful loss of legal certainty' Eleanor, if you could explain to the audience why this is so, as I do agree with you. • ELEANOR FOX—If you will allow me, I will start in a different place than legal certainty, and come to that later on. This is an enormous opportunity to get things right or get things better. I was asked to look in my written

1 - Compatibility, Efficiency, Legal Security

contribution for this workshop at the main things that are wrong with the current enforcement system. I came up with four, and some, but not all, have been extensively talked about. I will give you my list of four problems, or 'lost opportunities'. Then I will suggest how I think these problems could be solved. Finally I will touch upon remaining problems, including legal certainty. My list of four of course includes, first, the notification and prior authorisation system. The system does create busy work, and it shifts the resources of the Commission to areas with the least pay-off. That is clearly a problem, and it must certainly be addressed. The second problem is that the Article 81 (1)—Article 81 (3) dichotomy is an artificial segmentation and an artificial barrier. If this is an artificial barrier, is not the European Union against it? The problem is not just that under 81 (1) you look at negative aspects and under 81 (3) you look at the positive ones. Under 81 (1) the threshold is so low that it catches restraints that should not need justification. I will come back to that. The third problem is constituted by the ambiguities and difficulties under 81 (3). The first point under this heading is the extent to which exemptions can be given for anti-competitive agreements on the grounds of social policy or economic policy. The second is the extensive discretion that the Commission now has, and that national agencies will possibly get under the new system, in granting conditions under which deals can go through. The fourth problem relates to national courts. Under the current system there is clearly an inefficiency of enforcement in national courts. National courts cannot adjudicate on 81(3) justifications, and that is clearly handicapping national courts, and is very inefficient. Afinalpoint in relation to national courts, still in the context of the 'fourth problem': there is an enormous loss of opportunity resulting, from not enlisting the national competition authorities and courts to work together to enforce law, to become engaged with the law, to root themselves in the competition culture. Now the solutions to these problems, some of which have already been proposed. Problem number one, the notification and prior authorisation system, could be solved at least 90% of the way, and maybe more, by simply abolishing the notification and prior authorisation requirement without a devolution of enforcement. Problems number two and three, related to Article 81(1) and (3), could be solved by integrating the competition analysis under 81 (1) and leaving the granting of exemptions for anti-competitive agreements to the Commission, not to national competition authorities, so that whether an agreement is anti-competitive or not is decided under a 'broader' 81 (1). Problem number four, the lost opportunities for enlisting the national courts and national competition authorities and the inefficiency of handicapping the national courts by stopping them in their tracks when they come to an 81 (3) problem, could be solved by broadening the scope of 81 (1) and bringing all competition analysis under it. Of course, this is not exactly what the Commission proposes, and I believe will create problems when enforcement devolves towards national courts. First of all, national agencies and possibly

23

24

Panel Discussion

courts will have .the power to grant exemptions. It seems to me that the power to exempt anti-competitive agreements for, say, environmental policy reasons, is dangerous to devolve. In addition, national agencies and possibly courts will have the power to grant conditional exemptions. Although the courts will probably have to consult the Commission in such cases, I question whether this is wise. The solutions I would propose include, of course, abolishing the notification and prior authorisation system, bringing the competition analysis under Article 81 (1) and authorising national bodies to perform full competitive analysis with the Commission reserving the power to grant exemptions. In relation to coherence, uniformity and legal certainty under the new enforcement system: I do not think these are significant problems where there is a rule of law. I also think that it is very useful to have a certain amount of diversity coming out of the national courts'jurisprudence. There seems to be a perception that national courts are not supposed to be diverse or independent, but harnessed agents in the service of the Commission. A different model ought to be considered whereby national courts and national competition agencies become more autonomous in applying Community law, and not necessarily the Commission's interpretation of Community law. There are three areas of enforcement in which I believe this should be so. The first is the scope of Article 81 (1), though the European Courts have already expanded the scope of analysis under 81 (1) in European Night Services.1 The second is the viability of Consten and Grundig2 in light of intensified interbrand competition. The third is the application of EC antitrust rules to 'foreclosure' types of restraints, such as in the ice-cream cases. I believe these are all controversial enforcement areas. It would be helpful and useful for the European law to be bubbling up with more than one thinker lighting the sparks, and with cases eventually decided by the Court of First Instance and the European Court of Justice. Worrying so much about uniformity is not necessarily good. Just two points to conclude. First, it is possible that there will be very little practical change as an outcome of the reform. National courts already have the authority to apply 81(1) and 82, and they sometimes already do a full competitive analysis in actions to enforce private contracts, which the courts will have now the power to adjudicate on the whole, that the change will occur. Finally, I hope the reform will not be just a system of minimal devolution, and that national authorities and courts will be engaged in applying EC competition rules and establishing networks of co-operation. Networking and cross-fertilisation will be very helpful and enriching to the whole system.

1 Joined Cases T-374/94, T-375/94, T-384/94 and T-388/94 European Night Services and Others v Commission [1998] ECR 11-3141. 2 Joined Cases 56 and 58/64 Consten and Grudig v Commission [1966] ECR 299.

1 - Compatibility, Efficiency, Legal Security

25

• BARRY HAWK—Claus Ehlermann asked me to give, in my written contribution for this workshop, a US perspective on legal certainty. I do not know if I am going to give a particularly 'US perspective', but I would like to suggest some fundamental notions or ideas relevant to legal certainty. Some of these notions or ideas are so fundamental that they might strike many of you as banal. My excuse for mentioning them is that I sometimes wonder, particularly when I hear the anguished cries for legal certainty of some of my European friends, whether some of these fundamental notions are either not recognised or have been rejected on good grounds. First, I think it's helpful to define what we mean by 'legal certainty'. There are at least two, and I will suggest even three, important different meanings of this concept. The first is legal certainty about the predictability of the outcome of a case, or 'uncertainty about liability'. The second is uncertainty about the legal consequences flowing from the finding of a violation, or 'uncertainty about penalties'. In other words, uncertainty as to whether you are going to be fined or not. An almost unique European phenomenon within the same category is uncertainty about the enforceability of contractual provisions falling under 81 (1). The third is uncertainty about jurisdiction, which raises a very different set of considerations from those related to the predictability of outcome or the predictability of penalty. I shall not spend time on it, as this is more about a choice of law or how to resolve conflicts between jurisdictions, and there is a huge body of knowledge on how to address issues of uncertainty about fora. As to predictability about the outcome, I think there is clearly a tension between economics and legal certainty in competition law, as Simon Bishop has already pointed out. This is competition law, not real property law, and there is no need to run in and notify. In property law one must have stability and know who owns the property, so one registers a title back to the Doomsday Book if necessary. In competition law the need for stability or legal certainty is not the same, especially if one wants some economics in it. The more economicsoriented the approach in enforcement, the less legal certainty there is going to be. There is tension and a trade-off between the two. The only way to ensure absolute legal certainty is to have a. per se rule of prohibition or a per se permission rule. With per se rules there is no inquiry into the effects of particular agreements—be it in terms of price, output, market share or market power—or into their justifications. Agreements are classified as either lawful or unlawful, and the outcome is totally predictable. The difficulty consists of deciding what categories of agreements fall under the per se prohibition. A fundamental judgment must be made in order to identify those categories of agreements that have so few redeeming (or competitive) benefits and cause so much harm so as to justify the sacrifice of cases on the margin. Despite the differences of statutory structure and the horrible bifurcation of Article 81, per se rules of prohibition in the EU are largely similar to those in the US. They concern hard-core cartel practices, minimum retail price maintenance and—in the EC—certain restraints in vertical agreements because of

26

Panel Discussion

market integration. However, in European Night Services71 the Court of First Instance does everything but use the term 'per se rule'. In fact, if Eleanor Fox, or I, or anyone else wrote a casebook on comparative antitrust, the European Night Services judgment should be in it. It is one of the best decisions ever written, contrasting very nicely with the US Supreme Court's mysterious decision in California Dental.* There is, however, a price to pay for legal certainty, because per se rules involve a certain degree of arbitrariness. The issue arises even before actual enforcement, because a decision must be taken as to which categories of agreements fall under the per se rules. These are policy judgments that are to be made. There may be categories of agreements that, generally speaking, are almost always beneficial and are so rarely harmful that they can be permitted on a per se basis. It is difficult to find such an example in US law; one may be exclusive distribution agreements, but then only if not associated with market power. In the EU, the best examples of per se permissive rules are block exemptions on vertical restraints. The introduction of a market power threshold in the new block exemption on vertical restraints is the cause of this renewed passion about legal certainty. The introduction of market share analysis sacrifices legal certainty, yet market power considerations make the rules more economically meaningful. If, however, there are a lot of agreements that cannot be classified as 'always lawful' or 'always unlawful', then what is to be done? Their assessment requires an inquiry into the effects (good and bad, beneficial and harmful) of each particular agreement. In the US we call that 'rule of reason', and I went into this in some length in my paper. There is far more uncertainty under the US law, or under the 'rule of reason', than has ever been dreamt of under Article 81 (1) and (3). As an example of this, read the US Supreme Court's latest attempt to talk about the rule of reason in California Dental, which was a complete disaster. If societies go through three stages, (e.g. barbarism, civilisation, and decadence) then maybe we have become decadent, since the Supreme Court tries to figure out what is the rule of reason under Section 1 of the Sherman Act. So, read European Night Services, do not read California DentaP. As to legal uncertainty about the consequences of a violation, or legal uncertainty about fines, I think that is a red herring under Article 81. In almost all cases you know whether you are going to be fined or not. I therefore do not think fines are an issue, and the legal consequence of a violation is the unenforceability of the contract. The US law on unenforceability of contracts violating antitrust rules is almost identical to the EC law. If a forty-year-old contract is found to violate antitrust rules, it has always been unlawful. It has never been enforceable. In the United States there is no debate about legal uncertainty on the enforceability of contracts. Firms, both European and US ones, prefer the highly uncertain US rule of reason. In fact, they love the uncer3 4

See above. California Dental Associations v FTC, 526 U.S. 756, 119 S. Ct. 1604 (1999).

1 - Compatibility, Efficiency, Legal Security

27

tain US rule of reason as opposed to a per se prohibition rule. European firms complaining about legal uncertainty under Article 81 never mention the uncertainty under Section 1 of the Sherman Act, because when we shifted in the US from per se prohibition of vertical and territorial customer restraints to rule of reason, they were very happy. In the second place, the rule of reason is good for defendants. It is good for the manufacturers, because it is far more costly to prove unreasonableness than to prove a. per se violation. I believe this explains why there is almost no debate about legal uncertainty under Section 1 of the Sherman Act, even though I would suggest there is far more legal uncertainty under the rule of reason of the Sherman Act than under Article 81. • KAREL VAN MIERT—Thank you very much for this stimulating and outspoken contribution. I now invite Mr Giuliano Marenco to present his contribution. Almost two years ago, when we started the internal discussions at DGIV about the reform, I asked the Commission's Legal Service to be part of the exercise and tell us whether our proposal was in line with the Treaty. Giuliano Marenco played a very important role in the whole debate, and I am personally very thankful for all he did. • GIULIANO MARENCO—The question of the reform's compatibility with the Treaty has been raised mainly by the German government and in German academic circles. My impression is that, with the passing of time, interest in this subject is diminishing, so that I am afraid that the last two persons to be interested in it are Professor Mestmacker and myself. Three kinds of objections to compatibility have been made. The first objection, and I believe that Professor Mestmacker was particularly interested in this, is that a legal exception system would run counter to the direct applicability of 81 (1), which, as he says quite rightly, enshrines the prohibition principle. According to Professor Mestmacker, a legal exception system embodies a control of abuse system rather than a prohibition system. I believe that this is a misunderstanding, and this is not what the Commission wants. The Commission does not want to institute a control of abuse system. A legal exception system means that a restrictive agreement which does not fulfil the four conditions of 81 (3) is void and prohibited from the beginning, and does not have to wait for a decision by the administrative authority to be so, as would be the case in a control of abuse system. The second objection is about the drafting of 81 (1) and (3), and in particular the introductory formula of 81 (3), which says that the provisions of 81 (1) can be declared inapplicable to agreements fulfilling its four conditions. It has been claimed that this formula must mean that a prior administrative decision is necessary. I have tried to find an answer to this objection by elaborating on the vagueness of this formula, which is not as clear as the supporters of the

28

Panel Discussion

present system maintain, and above all by insisting on the fact that the Treaty delegates to the Council the adoption of rules implementing the provisions of 81 (3). Such a legislative program would be practically empty if the Treaty itself had already organised a particular enforcement system. I have also looked at the Travaux Preparatories of the Treaty, like Ian Forrester, and I share with him the view that this has been the most amusing part of the exercise. However, I do not share Ian's conclusion that the authors of the Treaty had vague ideas. Every time I reflect on the structure and the wording of the Treaty of Rome I feel obliged to think that the authors of the Treaty were particularly lucid and intelligent, and could foresee the future in a way that we could not imagine. My conclusion on the historical aspect is that, although there is no express confirmation of this, the authors of the Treaty deliberately did not decide which system should govern 81 (3), but instead left this task to the Council. The third objection is about the non-justiciability of 81 (3) because of the complex economic assessments involved in its application. Here I have one main response. It is true, as the opponents of the reform say, that the European Courts have professed self-restraint in the review of the Commission's exemption decisions, but I think that this self-restraint is not a result of the Treaty. It is rather the result of the system set up by the Regulation 17/62, to the extent that it gave to an administrative authority the task of applying the four conditions in 81 (3). It is inevitable that the review of the Courts is limited to what an administrative judge does in respect of administrative decisions. If this system is changed, the more penetrating control of the ordinary judge should prevail. However I have not been able to answer to the main and most interesting question: that is, why has opposition as to the compatibility of the reform with the Treaty come mainly from one country? I have asked my old friend and mentor, Norbert Koch, who was in the Bundeskartellamt, in DG IV and in the Commission's Legal Service. He has very wide experience, but even talking to him I have not found a satisfactory answer. Maybe it will come out of this debate. • KAREL VAN MIERT—Thank you very much indeed, Mr Marenco. I would not dare to answer that last question either, although from time to time I might be tempted. I invite now Professor Matsushita to give us the benefit of a view from someone from far away, but that person being someone who is very knowledgeable about what is happening in Europe. • MITSUO MATSUSHITA—I am from far away, and perhaps outside of the context of this European reform. My view is that of a layman, yet I believe that the debate surrounding this reform is quite relevant in so far as our situation in Japan is concerned. I am referring in particular to the deregulation of the economy, which in Japan has become an important issue. In a regulated economy, different kinds of regulations authorise certain activities, and their effects

1 - Compatibility, Efficiency, Legal Security

29

are similar to the prior authorisation system in EC competition law. Such regulations are currently being done away with, one by one. For example, just about one month ago, the Japanese competition law was amended to remove the exemption of 'natural monopolies—such as gas, electricity etc.—from its scope of application. Therefore, now the prior exemption is gone and competition law applies; this causes anxiety within industry as to what will happen. At present, both the Ministry for International Trade and Industry and the Fair Trade Commission in Japan are working to come up with some kind of guidelines on the application of competition rules to various industries. As far as I understand, this sort of situation is somewhat similar to what happens in Europe. I would like to make a few comments on the White Paper, and about Article 81 (1) and (3). I must confess that it is really difficult for me to understand the relationship between Article 81 (1) and (3). If you look at Article 81 (1), it prohibits agreements restricting competition within the Community. Article 81 (3) says that all agreements coming under the scope of Article 81(1), including distribution agreements, can be exempted from the prohibition if they benefit consumers and do not eliminate competition in a substantial part of the relevant market. Yet if the agreements satisfying the exemption requirements under Article 81 (3) do not eliminate competition on a substantive part of the relevant market, why do they fall under the scope of Article 81 (1). In other words, why are they prohibited? And if they are indeed prohibited, why can an exemption be granted? It seems to me that there is some sort of a conceptual difference between those two provisions. Article 81 (3) might just confirm the fact that the agreements satisfying its requirements are not unlawful under Article 81 (1). I feel that the relationship between these two paragraphs of Article 81 should be clarified. The White Paper proposes that the existent notification and prior authorisation system should be replaced by a system whereby national courts apply Article 81 (3) directly. The premise of this proposal seems to be that an agreement can be judged as lawful under Article 81 (3) without needing prior administrative approval. This will be accomplished by amending Regulation 17/62, but without changing the basic provisions in the Treaty. This also seems to endorse the view that the nature of Article 81 (3) is not really supporting a new legal situation. In Japan we had the concept of the so-called 'depression cartels', which eliminated competition completely yet were accepted on industrial policy grounds. Such agreements were approved regardless of their effects on competition. This situation is, however, quite different from that under Article 81 (3), because the application of the latter includes competition assessments. My next point is about the decentralised enforcement system proposed by the White Paper. If there are a number of agencies granting authorisation instead of just one, there will of course be divergent interpretations of the law, and there may be some risk involved in this. This is not really a European problem only; we will have that same problem in Japan when courts decide on antitrust

30

Panel Discussion

cases (there are about 100 such cases already). There are, of course, differences between enforcement in civil courts and enforcement by competition agencies. Though it seems to me that, as long as the European Court of Justice makes the final judgment, there will be convergence in the long run. Meanwhile, what is probably needed is some sort of communication mechanism between the national competition authorities, national courts and the European Commission. A few concluding remarks: the first is that, under the new system, the interpretation of Article 81 (1) and (3) will probably be more 'integrated'. This will lead to a more 'rule of reason' approach in the application of Article 81(1). The second remark is that the new enforcement system will reduce the scope for Commission intervention and increase the role of the judiciary. This is a trend we are also aiming at in Japan. Of course, there is a risk of interference with the independence of the judiciary, but I believe that an increased role for the judiciary is an essential feature of a deregulated economy. • KAREL VAN MIERT—The following presentation is that of the paper coauthored by Professors Petros Mavroidis and Damien Neven. Professor Neven will take the floor on behalf of both. • DAMIEN NEVEN—The basic framework of our paper can be summarised in one phrase: diversity is an opportunity, and divergence is a risk, but the challenge of decentralisation is to capitalise on the benefits of diversity—such as the creativity and professionalism of national competition authorities—while reducing the risks of divergence without reintroducing centralisation through the back-door. This is to say that trying to ensure consistency in enforcement by introducing alternative mechanisms leading again to centralisation would defeat the whole purpose of the reform. In our view, the system of multiple enforcement envisaged in the White Paper could be improved on two accounts. The first is that more attention should be paid to the benefits of diversity and the need to capitalise on its benefits, in line with what Eleanor said earlier. We think that the US experience is extremely valuable from this perspective, yet little attention has been paid so far to the US 'experiment', though some useful lessons can be drawn from it. As to the second account, we argue that in the White Paper the Commission understated, or underplayed, two important sources of divergence in enforcement. The first, and rather obvious, source of divergence is associated with differences in the institutional frameworks of national competition authorities. National competition authorities pursue different objectives, operate within different institutional frameworks and have different constraints. Some might be more independent than others, some might be more transparent than others, and some might be more responsive to the problems of business than others. To the extent that such institutional constraints vary substantially across the

1 - Compatibility, Efficiency, Legal Security

Member States, this will be an important sort of divergence in the new enforcement system. We suggest in our paper that, in order to address this potentially important source of divergence, the Commission should perhaps impose certain institutional standards on the Member States, for example, transparency requirements such as the publication of decisions and possibly also criteria for the nomination of head officials or regarding the status of civil servants. Let me now turn to what we regard as a more serious source of divergence, to the extent that it can lead to some fragmentation of the internal market. The system envisaged by the White Paper is likely to lead to simultaneous multiple enforcement, in other words, it is likely to lead to numerous instances whereby several national competition authorities will consider the same complaint or the same case. We believe that such occurrences have been underplayed by the Commission in the White Paper. The reason for such occurrence is associated with the test for determining whether EU law applies. The test is whether trade between the Member States is affected. This test is very simple to apply under the current enforcement framework where there is a single enforcer, because exactly how trade is affected does not matter; what matters is that trade between the Member States is affected. However, the matter is no longer so clear when there are several enforcers. To give an example illustrating this point, consider an agreement between an Italian and a Greek ferryboat operator. Obviously, one thing is that both the Greek and the Italian national competition authorities could handle a complaint about the agreement, but the question arises whether the Portuguese competition authority, for example, or the Luxembourgeoise competition authority, could also handle such a complaint. What then would be the competence test? In such cases, should something be added to the usual test as to whether trade between the Member States is affected? We tend to believe that there should be no additional tests, simply because all the White Paper does is to enlarge the set of enforcers. The White Paper does not in any way alter the mechanism that triggers jurisdiction, so there is no reason why an additional requirement should be introduced. One can thus expect that Member States will very easily assert jurisdiction with respect to particular cases and can be expected to actually handle the complaints received. In a simultaneous multiple enforcement system, one must take into account the incentives under which national competition authorities act. It is realistic to assume that national competition authorities respond to incentives coming from their national constituencies. They are responsible in front of their national Parliaments, to their governments, to the citizens of their countries, and they will make decisions taking into account their national interests. They will not take into account, in a systematic fashion, the effects of their decisions outside their own jurisdictions. Let us now think what the consequences would be of a system of simultaneous multiple enforcement coupled with the incentive mechanisms that I just mentioned. To illustrate the case, think about a complaint against an agreement that only makes economic sense if implemented throughout the Community, and not

31

32

Panel Discussion

only in one Member State. In an enforcement system such as that we were discussing, it only takes one Member State to ban such an agreement. This means that simultaneous multiple enforcement will generally be more restrictive than enforcement by a single enforcer at a central level. Indeed, whereas a single enforcer might implicitly, at least, balance positive and negative effects across the Community, national enforcers will not do so—they will only focus on the effects of the agreement in their own jurisdictions. Hence, what will be required for such an agreement to be allowed is the absence of negative effects in sixteen Member States! This a serious concern, to the extent that it will lead to an enforcement system that is biased against agreements with a European-wide scope, and in particular against agreements having effects across different Member States, which in the end is one of the main objectives of the internal market. We consider that the solutions considered in the White Paper in order to counter-act such effects are not sufficient. I have no time to discuss this in more detail here, but in our paper we have effectively proposed additional measures. The first, simple and obvious additional measure is related to incentives, which is the underlying problem in this case. That would be to introduce a positive comity obligation for the national competition authorities, which would thus have to take into account effects in other Member States in their decisions. Yet, as we argue in our paper, this is unlikely to be sufficient, and would have to be supplemented with a stronger 'safety' mechanism. We can think of two such mechanisms. One relates to the allocation of cases, and the other consists of having the Commission play the role of umpire by retaining Article 9 of Regulation 17/62. However, as we argue in our paper, the Commission's role of umpire in the context of Article 9 should be further clarified, in particular with respect to the circumstances in which the Commission will take over particular cases. In this respect, the US experience is again extremely useful, as the US Supreme Court can also intervene in well-defined circumstances and take over particular cases. Our main conclusion is that, under the new enforcement system, the Commission, in its role of umpire, would have to manage regulatory intervention as well as resolve conflicts. • KAREL VAN MIERT—Thank you very much. I suggest that we return to this rather crucial issue later on during the discussions. I now invite Professor Ulrich Immenga to take the floor on behalf of Professor Mestmacker, and if you could be just as outspoken as he is, that would be welcome. • ULRICH IMMENGA—I will now be speaking on behalf of Professor Mestmacker, who apologises for not being able to be with us today, as he has been ill for the last few days. Professor Mestmacker asked me to present what he considers the most important points in his written contribution for this workshop, some of which have already been discussed here. The first relates to the compatibility of the reform with the Treaty, an issue that has been discussed here before, so I will not return to this point. Another

1- Compatibility, Efficiency, Legal Security

33

issue is that of the direct applicability of 81 (3). Eleanor Fox and others have already mentioned this as well, but Professor Mestmacker was going to examine this point more in depth and consider the Commission's reform proposal as a kind of constitutional challenge. I quote from Professor Mestmacker's paper: 'The direct applicability of Article 81 (1) does not follow from Regulation 17/62, but from the Treaty itself. The European Court of Justice stressed that Regulation 17/62 did not, and could not, modify the effects of Article 81 (1) and 81 (2). The Court has consistently ruled that Article 81 (1) and (2) are, by their very nature, susceptible of creating direct effects in relation to individuals, thus creating for these individuals rights that the courts of Member States must safeguard. In interpreting the concept of 'national authorities' in Article 9(3) of Regulation 17/62, the Court distinguished between national competition authorities and national courts. As far as national courts are concerned, Regulation 17/62 could not take away their competence to apply Article 81 (1), because that would deprive individuals of rights conferred upon them by the Treaty, by virtue of direct applicability of Article 81 (1).' Then I quote briefly from the ECJ decision in the Marty Lauder case: 'Since the prohibitions contained in Article 81 (1) and 82 tend by their very nature to produce direct effects in relation to individuals, those articles create direct rights for the individuals concerned that national courts must safeguard. To deny, by virtue of Article 9 Regulation 17/62, the national courts'jurisdiction to provide that safeguard would mean to deprive individuals of rights that they hold under the Treaty itself.'5 And Prof. Mestmacker continues: 'It follows that a Regulation such as that contemplated by the Commission, providing for the inapplicability of Article 81 (1) when the conditions for an exemption are met, is compatible with the Treaty only if it does not interfere with the direct applicability of Article 81 (1). This depends upon the direct applicability of Article 81 (3): if the exemption provision, contrary to the Commission's assumption, does not satisfy the requirements of direct applicability, then the tying of Article 81 (1) to Article 81 (3) would deprive the former of its direct applicability. The most important indications as to whether Article 81 (3) is directly applicable can be encountered in the European Courts' decisions reviewing Commission decisions that granted or refused to grant exemptions. The European Court of Justice and the Court of First Instance have consistently held that the Commission is entitled to a wide margin of discretion in such cases. The veryfirstleading decision reviewing the application of Article 81 (3) by the Commission outlined the limitation of judicial review: the Commission, the Court observed, must render difficult economic judgments. The judicial review of such judgments is restricted to finding whether the relevant facts were correctly ascertained, and the law was correctly applied to the facts. In later decisions, the principle of limited judicial review in Article 81 (3) cases is spelled out in greater detail. In other words, the ECJ will not replace the

Case 37/79 Marty Lauder [1980] ECR 2481.

34

Panel Discussion

Commission's expert judgment with its own, and will not enter into a balancing of the economic advantages and disadvantages of a cartel. It has been argued that, even though Article 81 (3) is predicated upon a margin of discretion, it is nevertheless directly applicable because it does not provide for political discretion. This is a distinction that has no basis in the case law of the ECJ. In dealing with the limitations of judicial review in Article 81 (3) cases, the ECJ starts by observing that the Commission is responsible for the shaping and implementation of Community competition policy, and this calls for complex economic judgments. This applies to decisions in individual cases as well as to Regulations providing group exemptions under Article 81 (3). In a recent study for the Commission,6 the authors reiterate the Court's jurisprudence as to the complexity of economic judgments involved in the application of Article 81 (3), and the consequently political character of these decisions. The Court's jurisprudence shows that the conditions for granting exemptions under Article 81 (3) are not suited to judicial review, therefore a Regulation providing for the application of Article 81 as a whole would deprive Article 81 (1) of its direct applicability. Such a Regulation would infringe the Treaty' In his paper, Professor Mestmacker also addresses the particularities of the application of Article 81 as a whole in national court proceedings. His argumentation follows upon the previous discussion on the kind of judicial review performed by administrative courts. I would consider his argumentation on this point as a counter-proof to the understanding that private enforcement in national courts cannot be compared with judicial review in administrative courts. However, this is my own interpretation, and I will continue quoting from Professor Mestmacker's paper: 'The Commission expects the application of Article 81 as a whole in proceedings of three kinds: contractual liability proceedings (between parties to an agreement), noncontractual liability proceedings (between a third party and one or more of the parties to an agreement), and applications for injunctions. In dealing with private enforcement, the narrow limits of Community law must be observed. The only Community law sanction is Article 81 (2), providing for the invalidity of restrictive agreements. All other sanctions and remedies are a matter of national civil law. There are, nevertheless, important indirect effects of Article 81 (2) in tort and unfair competition law. For private enforcement to be effective, there is obviously a trade-off between the legal certainty of the parties to an agreement and the effective protection of third parties' rights. The Commission appears to be more concerned with the legal certainty of parties to an agreement than with the protection of potential victims.' • KAREL VAN MIERT—This argument will be taken up again tomorrow. I would like now to invite Alexander Schaub, the Director General of DG COMP, and then the last speaker of the first Panel, Mr Tesauro.

6

European Commission: Die Anwendung der Art. 85 und 86 des EG- Vertrages durch

die Gerichte der Mitgliedstaaten, Catalogue no. CV-O6-97-812-DE-C, available at http://www.europa.eu.int/comm/competition/publications/SUDIES, p. 133, no. 87.

1 — Compatibility, Efficiency, Legal Security

• ALEXANDER SCHAUB—I have been asked to deal in my written contribution for this Workshop with two issues that are not directly inter-linked. The first is why the German critics of the proposed reform are wrong, and the second concerns 'non-infringement decisions'. As the second topic seems to be much more sensitive, controversial and relatively open at this stage, I will start with it. In the White Paper it was proposed that in the new enforcement system the Commission should have the power to adopt individual positive decisions, or 'non-infringement decisions', as we now prefer to call them. Such decisions would declare that a particular agreement is not infringing Article 81, and would be of declaratory nature. In our view, constitutive decisions such as the present exemption decisions, creating rights for the past, present and future, and that can be revoked by the Commission only, are not compatible with the legal exemptions system envisaged in the White Paper. They would conflict with the sharing of enforcement competencies. I believe that, so far, this is broadly accepted. However, the fact that the Commission non-infringement decisions will be of declaratory nature does not mean that they will be without legal effects. The effects depend on the operative part of the decision. I would agree that the current negative clearance decisions bind only the Commission, since according to Article 2 of Regulation 17/62 they merely state that the Commission finds that there are no grounds for it to intervene. A decision declaring that a particular agreement is not infringing Article 81 would, in my view, have stronger legal effects. It would establish the Commission's substantive assessment based on the facts available to it, and would have certain effects on national courts. If the facts on which such decision is based change materially, national courts would obviously be free to come to a different conclusion. Such a finding would in no way be in conflict with the Commission's decision. If, however, the facts are exactly the same, in my view national courts must take into account that a Commission decision is a Community act, subject to a separate system of judicial control. Although national courts apply Community law and play a very important role in its enforcement, they are not competent to review the validity of Community acts. This is the sole responsibility and competence of the Community Courts. When the Commission and the national courts adopt decisions in the field of EC competition law, they both apply the primary rules of the Treaty. If a Commission decision is annulled on substantive grounds, it is because the decision violates the Treaty, or in other words, is invalid. In the Community legal order only the Community Courts are competent to declare the Commission's decisions invalid. If a national court declares null and void an agreement that was the subject of a definitive Commission non-infringement decision, it is calling into question the validity of the Commission's decision. Therefore, in my view, in such cases the national court must make reference to the European Court of Justice based on Article 235 EC Treaty. Furthermore, if the Commission non-infringement decision has not yet been adopted, or it is not yet definitive, the national courts are obliged to avoid decisions in conflict with it, otherwise the uniform application of Community

35

36

Panel Discussion

competition law and legal certainty would be threatened. As it is clear from the Delimitis judgment,7 such conflicts can be avoided in a number of ways: the national court can ask the Commission for an opinion, can follow the Commission's decision, it can suspend proceedings or make a reference for preliminary ruling to the European Court of Justice. We fully acknowledge that these are very complex issues, on which different views are possible and indeed are held within our own institution, which as of today does not yet have an official point of view in this respect. Much of this complexity has to do with the fact that the issue involves the relationship between two legal orders applying the same body of rules, but which are nevertheless assigned distinct tasks. Hopefully, the European Court of Justice will cast new and valuable light on the relationship between the Commission, the national courts and the Community Courts when rendering the judgment expected in the famous Masterfoods case.8 I wanted to add that divergences of views on this issue are based on the idea that the principle of separation of powers makes it impossible that decisions by an administrative body bind the jurisdiction of courts. In my view, however, the issue is not one of separation of powers, but of separation of judicial control systems. The decisions of a Community body can only be reviewed by Community courts. Let me very briefly add an important point that was not clarified in the White Paper, which is why we consider it desirable that the Commission be empowered to adopt non-infringement decisions. In the Community legal order the Commission is responsible for the implementation and orientation of Community competition policy. In the new enforcement system, Community competition policy will primarily be developed by means of general instruments, such as Block Exemption Regulations and Guidelines, and by means of individual prohibition decisions. In our opinion, however, the Commission should also be able to influence Community competition law enforcement through the adoption of non-infringement decisions. This will allow the Commission to provide guidance on the Commission's approach to new competition issues. It will also allow the Commission to influence the orientation of Community competition law enforcement in a system of parallel competencies. The Commission is the only authority that covers the entire Community territory. It therefore has a special responsibility to ensure consistent application of Community law. Non-infringement decisions are not exemption decisions in disguise. They would only be adopted exceptionally, in the Community public interest, and never just at the request of undertakings. • GIUSEPPE TESAURO—I will only make some telegraphic considerations. On the issue of the compatibility of the reform with the Treaty, I think that most 7

Case 234/89 Delimitis v Henninger Brau [1991] ECR 1-935. Case C-344/98 Masterfoods Ltd. v HB Ice Cream Ltd, on reference for a preliminary ruling from the Supreme Court of Ireland, judgment of 14.12.2000 (not yet reported) 8

1 - Compatibility, Efficiency, Legal Security

37

of the pertinent observations were already made. I think that the wording of Article 81 (3) leaves open the question of how the exemptions to the general prohibition should be administered. This formulation is broad enough to admit both ways of enforcing the prohibition system. The only limit it establishes is compliance with the prohibition principle, and that is the objective of the White Paper, I believe. Another argument against the compatibility of the reform proposal with the Treaty is that it deprives Article 81 (1) of direct applicability by tying it in with Article 81 (3). I think that the problem with Article 81 (3) is not one of direct effect—Article 81 (3) can have direct effect, of course—but whether national authorities and courts are competent to apply it. According to Article 9(3) Regulation 17/62, national competition authorities and national judges do not have the competence to apply 81 (3). The direct effect of Article 81 (3) would not be a problem if national bodies had the authority to apply it. Point 50 of the Delimitis judgment9 is already clear on this subject: national competition authorities and national judges already have the competence to look at the 81 (3). I also think that all the problems of legal certainty and enforcement efficiency under the new system are best approached by looking at Articles 81 and 82 as part of the wider Community legal system. I have a question for myself and for you: are Articles 81 and 82 really different from Articles 28 and 30? I do not think so. The entire Community legal system is made of rules and exemptions. If the Community legal system as a whole has been applied by national authorities and national judges, why not Articles 81 and 82 as well? What is the difference? Enforcement is 'much more difficult' for whom? National competition authorities are specialised bodies, and I do not think that it is much more difficult for judges to apply 81 (3) than the old and dear Article 36. There are of course good judges and bad judges, but this is a rule of life. There are judges that are not very familiar with economic issues, but can an 'economic animal' be found among our best judges in the Court of First Instance and the European Court of Justice? I do not feel that we must be afraid on this point. The real problem, of course, is the difference in nature, in terms of the rules applied and in terms of the legal consequences of the decisions taken by national competition authorities. From the legal certainty point of view, a word about the Commission putting some strange objects on the table (such as decisions that are not really decisions, or papers that are 'non-papers'), which I think is dangerous. We are now in a system of comfort letters, 'spiritual comfort' letters. The judge has some difficulty in finding the real nature of this kind of object: communications, resolutions, comfort letters, non-infringement decisions. This could be seen as an element of legal insecurity. Of course it is a good system for the lawyers. But what is a non-infringement decision? What is the legal value and/or the consequence of positive or non-infringement decisions? 9

See above.

38

Panel Discussion

• KAREL VAN MIERT—We now have an hour left for discussions. After the stimulating contributions of this morning, I do not think there is any need to stimulate further, though perhaps for a second we might come back to the question of whether the proposed reform is in line with the Treaty or not. Mr Marenco could perhaps go into the arguments that have been made by Professor Immenga on behalf of Professor Mestmacker. Then Alex Schaub could perhaps explain very briefly why this type of arguments were mainly invoked in Germany. • GIULIANO MARENCO—The argumentation against the compatibility of the reform with the Treaty is, if I understand well, extremely complex. It consists of saying that the direct applicability of 81 (1) will not be able to deploy its effects to the extent that 81 (1) is applied together with 81 (3) and the latter is not justiciable. This line of argumentation links two kinds of problems: the question of the direct applicability of Article 81(1) and the question of the justiciability of Article 81 (3). On direct applicability, I think that everybody agrees with Professor Mestmacker that Article 81 (1) embodies the prohibition principle, and anything that would weaken it would run counter to the Treaty. On that point, as I said earlier on, it is wrong to say that the legal exception system weakens the prohibition principle. On the contrary, it implies a stronger and more stringent application of the prohibition principle than under the current system. The current enforcement system has some gaps as to direct applicability and the application of the prohibition principle. One is that notified agreements are immune from fines. Another is that, with the Commission's exemption monopoly, in practice it is not possible for national judges to apply the prohibition principle under Article 81 (1). We all know how difficult it is for judges to decide on one half of a rule when the other half cannot be applied. A legal exception system is therefore a more stringent application of the prohibition principle. The real question in Professor Mestmacker's argument is, however, whether 81 (3) is justiciable. There is a declaration of self-restraint in the judicial review of Commission decisions based on Article 81 (3) on the part of the Community Courts, which is accompanied in fact by a very penetrating control of the reasoning followed by the Commission in its exemption decisions. The European Courts have more than once annulled Commission exemption decisions (for example, in the Eurotunnel and the European Night Services cases10). Yet even if there is some kind of self-restraint, I believe this is the result of the system set up by Regulation 17/62, which provided for an administrative decision to be taken on the applicability of 81 (3). Once this power is centralised with the administration, it is inevitable that the Court's review of the administration's dealings be limited to an administrative type of judicial review. The 10

Case C-408/95 Eurotunnel [1997] ECR 1-6315; European Night Services as cited above.

1 - Compatibility, Efficiency, Legal Security

39

self-restraint of the Court is nothing else but the traditional self-restraint of administrative judges in the review of administrative decisions. Once the new Regulation has suppressed the current system and replaced it with a system in which 81 (3) is directly applicable, the judges will have no longer reason to limit the scope of their review.

• ALEXANDER SCHAUB—The primary objective of the reform is clearly to improve the efficiency of EC competition law enforcement, and not to reduce the workload of the Commission. From this perspective, the criticism—(coming particularly from German circles) that the reform will actually reduce efficiency is very serious, and we have taken it very seriously. The main arguments invoked by those criticising the Commission's reform proposal from this perspective were as follows. One argument is that the reform would imply the abandoning of the prohibition system and a switch to a control of abuse system. It has already been explained that this is a total misunderstanding. Ten days ago I had the opportunity to discuss this with Professor Mestmacker and I got the feeling that he would no longer really insist on this argument; he admitted that there might indeed have been a misunderstanding. The White Paper did not make it sufficiently clear that illegal agreements would be invalid ab initio under the new enforcement system. Having clarified this aspect, I do not think that this argument remains important. A second argument is that the legal exemption system as a concept can only be less efficient than the present system, and I think that Giuliano Marenco has already replied to this. A third argument, that has been often repeated, is that without the notification system the Commission will no longer know anything about the market. Professor Moschel and Dieter Wolf made this argument. This seems to me a rather amazing and unfounded argument, because we have a lot of knowledge about the market coming from much better sources than notifications. On the other hand, we do not get a lot of useful information from notifications. The really critical cases are not only 'notified', but they are normally accompanied by complaints that are sometimes more informative about the case than the notification itself. Apart from these specific points, we generally believe that the present notification and prior authorisation system is not particularly efficient. First of all, there is no obligation to notify, and agreements are therefore increasingly being 'notified' only when the parties expect a negative clearance or exemption. In other cases, they are often not notified. Second, the notification system creates a largely unproductive workload. This comes back to our experience that, as of now, notifications are neither a very useful source of information, nor a very useful starting point for an investigation. Third, we believe—and this point has been detailed in the White Paper—that the present system prevents a proactive and focused enforcement effort. It is instead the source of unnecessary

40

Panel Discussion

delays in the handling of complaints and ex officio cases. And finally, in my view, culturally speaking the worst disadvantage is that a notification system places the wrong focus for enforcement. • ULRICH IMMENGA—I would like first to return to the issue of whether Article 81 (3) is justiciable. One aspect that has not been discussed here yet is that there is a clear difference between the Commission, which is not an independent cartel office, and national bodies. In its enforcement activity, the Commission must consider different Treaty objectives in the context of Article 81 (3): industrial policy, employment creation, and others, as it happened in the Volkswagen-Ford case.'' No national body, be it the competition authority or national courts, can consider the Treaty objectives in such a way, and I therefore believe that it is very important to reduce the scope of application of Article 81 (3). My second remark concerns the 'German objections' you have mentioned. One aspect that was only briefly discussed until now is whether the change contemplated by the Commission can be made through the adoption of a Regulation, or whether the Treaty itself must be changed. For example, one important issue is how to apply block exemptions. I think these are two very important points that have to be considered in order to understand the German point of view. • ULF BOGE—Alex Schaub mentioned that he was asked the question 'why the German critics are wrong'. I do not see this as a question. If somebody puts this question to Alex Schaub, then he has to answer it, and I understand that he hesitated a little bit because the question implies that it is wrong that the Germans are critical. I wonder if anybody has the wisdom to say what is right and what is wrong. I think we started a discussion in April 1999, and the Commission made a proposal, and at that time there was one proposal to reach the objectives we had. I will say it again: we never had problems with the objectives of the reform, as you know. The problem was: is the Commission's proposal the right way to achieve them? I think one must also take a look at history. About forty years ago, we had a discussion about which approach to take in Germany. It was then very clear that, for a variety of reasons, we wanted to have a notification system. The main reason was, however—and here there is no difference with respect to the current debate—that we wanted to have the most effective competition enforcement system. With respect to the current debate, I would say that a 'German concern' exists only in a very broad sense. Points of view differ in various German circles in so far as the legal aspect of the reform is concerned. Can modernisation be achieved by simply modifying Regulation 17/62, or should the EC Treaty be 11

Commission Decision 23/49/EEC Ford-VW of 23.12.1992, OJ L 20/14 of 28.1.1993

1 — Compatibility, Efficiency, Legal Security

amended? I am not a lawyer, though I have learned a lot from lawyers, and I wonder indeed which side is right. We have very respectable people in Germany who say it is impossible to bring about a reform without modifying the Treaty. Professor Immenga mentioned this and Professor Mestmacker was very clear on this point. In Germany we also have a 'Monopoly Commission', the most important body concentrating on the enforcement of competition rules, and if such a Commission has a very clear opinion on a matter I think it must be taken seriously. I am not the person to say whether it is right or wrong, but I wonder what we can do to find out which position is the right one. If there were a way of getting a clear picture on this, it would help as far as competition enforcement and legal certainty are concerned. It is up to the EC to decide to do it or not, and to find a way to do it. If the German Monopoly Commission and the German government are right—they have the same opinion on this point— what will happen within the next ten years when the first case goes to the European Court of Justice? I know this takes time—five or ten years—therefore I do not know when such a decision will have to be made. But what would happen if they were right after all? Mr Marenco said nobody could explain to him why this situation has occurred. I am sure he knows the Report of the German Monopoly Commission, and I think it is explained there. I believe he knows the history of the German enforcement system, and how it was developed. In contrast to Brussels, Germany has had very positive experiences with the notification system. We have had no problems with the existing system; you had a problem in applying it. We then said in Germany "yes, of course, one should try to modernise the European system". The only question was how to do it, how to reach our common objectives. I believe the debate with the so-called 'German critics that were wrong' lasted over one and a half years, and my judgment is that it went well. I think that critics and questions are natural at the beginning. As far as I know, many of the Member States had the same queries about the reform as those raised in Germany, and in the process the Commission understood many of the problems with the reform proposal as well. I would say there was aflexibleapproach over these months of discussion, and the Commission's opinions were modified after April 1999. We have Working Groups for the most serious topics, and I think they work well. As I said in my opening remarks, we are trying to help the process. In the end, this showsflexibilityafter a year or so of debate. I mentioned earlier that a 'compromise proposal' made by the Bundeskartellamt. From our point of view, the issue of legal certainty for companies is not really the most important issue. I agree that one has to ask, as Mr Hawk did: 'what is legal certainty?' When we advised the countries of the exSoviet Union, we always told them: 'what you need is a clear legal system, otherwise there will be no investment.' To ensure legal certainty for companies in our countries means something totally different. Uncertainty is a reality in political matters, and you never know what could happen if you go to court. It is in the hands of God, I was told, if you go to court. You never know what the outcome will be in political matters, why should it be different in competition

41

42

Panel Discussion

policy? However, if national competition authorities do not have the power to say 'OK, this is the acceptance under Article 81 (3)', then the burden is left to companies. There is also another problem that has not been mentioned yet: the difference between small and medium-sized companies and large companies. I think that as far as the 'grey zone' is concerned, it is much easier for large companies to obtain legal advice from their own lawyers, to ask others or to make an economic study. In so far as Article 81 (3) is concerned, something must be said about what is going on in the market. Large companies can pay for such studies, be they economic or judicial, but what about small and medium-sized companies falling under the threshold for the application of EC rules? This is a problem we still have to solve. My last point is that we must clarify the fact that our main objective is to protect competition. Legal certainty for companies cannot be our first objective. This also means that competition authorities must be given the power to gather information and the opportunity to ask questions of the companies. • CLAUS-DIETER EHLERMANN—The debate about compatibility is not only a German debate; this is exaggerated. There is also Professor M. R. Mok, who was an EC official in the early sixties and lived the history of Regulation 17/62, who said from the very beginning that there was a big compatibility problem. He draws his objections from the Bosch judgment of the European Court of Justice.12 The House of Lords might have been impressed by Dieter Wolf's statement about the incompatibility of the new enforcement system with the Treaty. The House of Lords implicitly makes the suggestion that it might be wise to amend the Treaty to clarify the situation and to substitute the word 'declare' in Article 81 (3) with a word that more clearly follows the direction of the new system. My second point is that the debate about the direct effect of Article 81 (3) is a debate about substance. Nobody here today has been able to explain exactly what is in and what is not in Article 81 (3). Among other reasons, this is so because there are so very few Commission decisions based on Article 81 (3), and even fewer decisions have been looked at critically by the Court. Are 'extracompetition' objectives to be considered under Article 81 (3)? This leads us back to the debate we had in 1997 in our European University Institute/RoBert Schuman Centre Workshop about the objectives of competition policy. More than anything else, decentralisation sharpens the debate about the substance and the content of Article 81 (3). Should it be only economics, in part with Article 81 (1), or more than economics? As David Edward pointed out in his paper, the Maastricht Treaty has not facilitated things by introducing as generalising other policy objectives, like environment, consumer protection and health, etc. 12

Case 13/61 Bosch [1962] ECR 89.

1 - Compatibility, Efficiency, Legal Security

43

• ELEANOR FOX—I have two remarks regarding the claim of incompatibility. One follows directly from Claus Ehlermann's point. One claim of incompatibility, as reported by Professor Immenga from Professor Mestmacker's paper, is that no national body can consider Treaty provisions containing EU industrial policy objectives. Why not allow national courts and national competition authorities to grant exemptions only on competition or efficiency grounds, thus excluding any possibility that national courts may apply industrial policy considerations under Community law? I think that if every Member State has the power to apply industrial policy under Community law, you will have greater uncertainty and much more industrial policy than you have today. My second remark relates to the claim that Article 81 (3) is not justiciable. I thought that Professor Tesauro made a very interesting point about Articles 28 and 30. The Cassis de Dijon13 exemptions can and are meant to be applied by national courts. I would like to hear why, if this can be done under Article 28, it cannot be done under Article 81 (3). • MARIO SIRAGUSA—I would like to go back to some of the other questions that were discussed this morning. I was very interested in what Alex Schaub said that there is a re-consideration of the role of the Commission in the new enforcement system, at least in the sense of issuing 'positive decisions'. The White Paper left very little room for that kind of Commission decisions, but I see that the Commission is finally starting to see the light. As Alex Schaub announced today, there will continue to be positive decisions made by the Commission. I think it is very important to have positive decisions, because of what Claus Ehlermann was saying. If we are moving towards a system in which national courts and national competition authorities apply Article 81 (3), it is very important that they have instructions about how to interpret Article 85 (3), for all the reasons you were discussing. What is the content of Article 81 (3)? The best way to clarify this is to continue to have a flow of Commission decisions interpreting the conditions for the granting of exemptions under Article 81 (3). It is true that this can also be done through Guidelines and Notices, but the best way is through case law. The greatest achievement of the European competition enforcement system has been to develop a body of case law (decisions of the Commission, the Court of First Instance and the Court of Justice). I would see the Commission renouncing its role of decision-maker as a great problem that allows it to generate jurisprudence and influence the application of the rules. Second, if there will be 'positive decisions' made by the Commission, what will their role and effects be? There is a very important change implied here. Under the present system— and here I share some of the 'German concerns'— 13 Case 120/78 Rewe-Zentral v Benndesmonopolverwaltung fur Branntwein [1979] ECR 649.

44

Panel Discussion

the Commission has very powerful means to shape policy and influence enforcement. The monopoly over the application of Article 81 (3) gives the Commission a tool to shape its policy. If the Commission issues a decision under Article 81 (3), everybody has to respect it. The jurisprudence of the European Court of Justice clearly indicates that national competition authorities and national judges are bound by such decisions. Even the application of national law cannot deprive Article 81 (3) decisions of their effects. Under the new system, the effects of Commission decisions are jeopardised. The discussion of these aspects in the White Paper is very poor. The White Paper even went so far as to say that Commission 'positive decisions' will have no binding effect. Now the Commission is going back on this a little. I saw in Emil Paulis' paper that the Commission is trying—rightly so—to say that, on the basis of the principles in the Treaty, a Commission 'positive decision' must have some binding effect as it is a Community act. Still, we are moving towards a situation in which such decisions will have a much less clear binding effect. The Commission is giving up the only clear tool of competition policy-making that, together with the block exemption regulations and individual decisions, has been the central pillar of the entire construction of our system. My third observation is concerned with the role of the European Courts under the new enforcement system. The main problem with the reform proposal is the weakening of the role of the Commission, and not the inability of judges to apply EC competition rules. I have great admiration for the judiciary. My proposal has always been to reduce the extensive application of Article 81 (1). It is not that I fear judges, but there is a fundamental distinction between EC competition rules and other provisions of the Treaty. Articles 28 and 30 always concern conflicts of laws, or whether national laws and regulations are in compliance with the Treaty, which is a totally different matter to that we are talking about here. Here we are talking about companies making investments across Europe, shaping alliances throughout Europe, engaging in certain behaviour over a number of years in the future, and where economic considerations are a very important part of the discussion. This is not a question of interpreting national law and determining whether it is in compliance with the Treaty, it is a question of forecasting and reviewing market situations and assessing whether a particular agreement has some redeeming virtue qualifying it for an exemption under Article 81 (3), which I think is a different exercise to the application of Articles 28 and 30. • IAN FORRESTER—Since the Commission is engaged in a fundamental procedural and substantive housecleaning, I think it is good that we give attention also to Article 81 (1). In the very beginning it was decided by the Commission, understandably for those times, that it was by the tolerant use of Article 81 (3) and not by the tolerant use of Article 81(1) that all sensitive decisions were to be made. That was decided forty years ago, and we understand why. That was a procedural disaster for the Commission because it created an inescapable

1 - Compatibility, Efficiency, Legal Security

demand for a scarce resource—exemptions—and only the Commission could deliver those. Finally, that demand became crippling, and that is why we're talking about reform today. I believe that, functionally speaking, the proposed reform will not have such a big effect on how national courts approach individual matters, because I believe that today—illogical and inconsistent with the German view, though it might be—a national judge wants to do a complete job. Once he accepts that the competition law issue is present, he wants to find a solution to it so that he can go on and the parties can be told what the relevant law is. I therefore believe that, functionally, a national judge today makes a comprehensive appraisal of what the situation is, except where the Commission has shown a likelihood that it is going to pre-empt him. I believe that agreements in the past were or were not enforced before national Courts, not because they were or were not 'notified' but because of their contents. That is going to be the same in the future. In practice, judges armed with the power to apply Article 81 (3) will do so, and it will not be such a big difference. It is a good thing. But remember that the Commission has not yet abandoned the original error, which was to give too great a power, too great a scope, too great an importance, to Article 81 (1). As part of the reform process, it is not enough to say merely that Article 81 (1) and (3) come together. I submit that the scope of Article 81(1) must also be re-calibrated, so that fewer agreements are caught by it, and thereby reducing everyone's workload and the opportunity for frivolous distractions by purely theoretical disputes. • CARL BAUDENBACHER—I shall take up an issue that has been touched upon by Professors Ehlermann and Immenga. Some feel—particularly in Germany, but not only—that the whole modernisation exercise will lead to a situation in which the Commission will have more possibilities to take industrial policy goals into account, as well as other policy goals outside competition policy. I have two questions in this respect. First, is this so? Second, isn't this unavoidable to a certain extent? I am not only thinking of the Maastricht Treaty, but also of what happened in Seattle. It seems to me that employment issues will be back on the table again after Seattle, and to a much greater extent than was the case in the last ten years. Another observation: if you look at afieldof European law that is very close to competition law, namely public procurement law, you will see that we have all these ideas of taking criteria outside pure competition ones into account, and all this is dealt with in an open way. If we take seriously what Giuseppe Tesauro has said, namely that competition policy is just like any other policy under the EC Treaty, then we should probably hold our eyes open and we should say: 'isn't it unavoidable that this is going to happen in the future?' • JOCHEM BURRICHTER—I would like to cover two points. The first relates to the argument that civil courts are not able to apply Article 81 (3) because the European Courts have granted a certain degree of discretion to the

45

46

Panel Discussion

Commission regarding the application of this provision. I totally agree with Mr Marenco, that this is the mere consequence of making the exemption procedure an administrative procedure. I would like to add that this might also be a consequence of the reluctance of the European Courts to deal with facts. Practitioners have deplored this practice. Very often and unfortunately this is still the case even after the establishment of the Court of First Instance. In my view, this is one of the reasons why this discretion has been granted, because even review of administrative decisions it is not necessarily limited to a kind of 'control of discretion'. In other jurisdictions, for example in Germany, the courts totally and fully review the decisions of the national competition authorities and do not grant them any discretion that they are not able to control. In Germany we have also had the possibility in the past to grant exemptions for rationalisation cartels, and the courts have clearly taken the liberty to examine whether the conditions for a rationalisation were fulfilled or not, whether the authority has applied the criteria correctly or not. This will also continue into the future, because we have taken over the criteria of Article 81 (3) in the new paragraph 7 of the German Competition Statute. I am sure that the German courts will take the liberty to control the application of these criteria. It also has been said that policy considerations are part of the application of Article 81 (3). I really cannot find anything in the conditions for the granting of an exemption under 81 (3) that allows a general reference to policy considerations. I would be more than happy if, in the application of Article 81 (3) by the courts, this nonsense would be abolished, if there has ever been a practice of this kind. The time is more than mature to avoid that mistake. My second comment refers to what Mr Schaub disclosed, namely DG COMP's considerations concerning 'positive' or 'non-infringement' decisions. Mario Siragusa has already explained why it is absolutely necessary to have such decisions in the future. I would back the idea that such decisions should be binding on national courts and national competition authorities, because the principle of primacy of EC law over national law clearly justifies such a binding effect. I even would invite the Commission to go a step further, in the sense that the Commission should even be able to over-rule decisions of national courts (for example, when a national court established that this provision was not applicable in a case where the decisive question was that of the applicability of Article 81 (3)). In extraordinary cases, the Commission should have the possibility and the competence to over-rule such judgments, even if they are res judicata. What would be the alternative otherwise? Let us imagine that a national court wrongly applied Article 81, the parties do not appeal the decision and no reference is made to the European Court of Justice. In such a case, the only possibility for the Commission would be to introduce a violation procedure against the Member State. This is legally and politically very complicated because a court acts independently. Giving the Commission the possibility of using a decision annulling a national court's decision in such a case would be a milder way in which to correct a wrong decision (which will occur and cannot be avoided), and it would guarantee more coherence in enforcement.

1 - Compatibility, Efficiency, Legal Security

47

• JAMES VENIT—I was puzzled by the request earlier this morning for more decisions from the Commission, although having heard Ian Forrester's second set of remarks, I begin to understand that maybe these would be decisions limiting the scope of Article 81. But it seems to me that there are a couple of fundamental points to be made on this issue. I think that there is a convergence of views on them, despite the smoke that might indicate that there is not much of a convergence. It is essential for a regulatory authority to set its own agenda and not to allow that agenda to be determined in a binding way by the consumers of the 'product' the authority is producing. Conversely, that authority should have the discretion to allow third parties to address it on issues that the authority regards as important. I think what all this means is that you have to get rid of the system of mandatory notification—that requires the Commission to deal with the 'notified' agreement—but you have to allow the process of complaints pass issues that might be interesting up to the regulator, and then let the regulator decide if it wants to deal with it. If it does not, the courts are of course available to deal with those problems. This is a very fundamental point that has been recognised in the reform proposal. I think a lot of the dissatisfaction that we heard earlier from the Bar as to lack of legal certainty under the new enforcement system or the impossibility to request opinions from the Commission will be rectified by limiting the number of positive decisions that the Commission can adopt. I have been involved in cases where, for instance, I am advising the client of the risk of fines. If one followed the classic judgment, I would have said 'yes, you do have a risk of fines and therefore you can't do this'. Under the notification system, we have been able to notify clients of 'grey area' agreements and spark a rather interesting debate within the Commission about a very important policy issue. Those opportunities should be preserved, but they can be preserved without a system of formal notification. From my perspective, the request to maintain 'voluntary' notifications would be irresponsible coming from those of us who have said for so long that the current system is not functioning. In fact, exemption decisions do not provide legal certainty because there are so few of them. It would be wrong for us to come back now and say that we are terribly concerned about the issue of legal certainty. I think the Commission is in the process of getting the balance right. We, the members of the Bar, have a responsibility to acknowledge that and to make constructive suggestions. But we also have to recognise that moving to this new world is not going to change the reality of what has been going on for a long time, which is basically companies deciding on their own, following the advice of their legal counsel, as to what is a reasonable course of conduct in terms of risk assessment and what is not. I am a little bit puzzled to hear from my own colleagues a kind of a pulling back from the reform and towards the way in which the old system has actually been working. Given that fact, it's very important to have reality and the formal rules accord with each other.

48

Panel Discussion

• GIUSEPPE TESAURO—Just a few words about the 'non-infringement' or 'positive decisions'. They are no problem at all, yet it is necessary to know whether such decisions are binding or not. My second point: a judgment could be wrong of course, but not only in the Member States, also in Luxembourg. Also Commission decisions could be wrong. We cannot reason on the basis of a paradox, of the extreme, of a crazy judgment. A crazy judgment is now possible, and will be possible also in the future. A crazy judgment is normal. It is the rule of our life, and we must also consider the hypothesis of a crazy judgment. • ANNE SPIRITUS-DASSESSE—Under the current enforcement settings, when a national judge is confronted with a case, he has to deal with it. He cannot wait for a Commission decision. In some cases, the judge has to decide if the contract is valid or not, and from this perspective he already has to take Article 81 (3) into account. He might wait for the Commission to carry out an economic evaluation, because he has this option, but he has to make a decision in cases in which he cannot wait, and this implies looking at Article 81 (3) and considering whether the Commission is likely to give an exemption or not. When the Commission has not been notified of the agreement at all, the judge has to deal with the case and has to make a decision because the period preceding a notification will never be covered by an exemption decision of the Commission. If the parties claim they will make a notification, this is alright, but it is for the future. The paradox is that, in such cases, the same contract can be declared invalid for the period before the notification, before it is clear whether or not it is valid for the future. With the notification, the parties might bring to the ' Commission even new elements modifying the original contract. A judge confronted with such a situation under the new system can act and play a dynamic role, instead of being bound to say 'for this period the contract is invalid and for the future we will see'. He could tell the parties 'you have to bring some kind of modification to the contract, then we can go forward'. During this time, the parties can consult each other or go back to the Commission. There are a lot of possibilities, and I think that this is one of the positive aspects of the reform with respect to national judges. What I mean is economicflexibility.The judge has to do something more 'economic' than just making legal assessments. I think this is perhaps the opportunity to give the judges a moreflexiblerule. I would like to add two general considerations. One is in line with what Mr Tesauro said earlier. As judges, we have to apply not only Articles 81 and 82, we have to apply all the provisions of the Treaty if they have direct effects. So what can happen in a competition case? The case must be decided on the base of Articles 81(1) and (3), we have to deal with the case very fast even if the case is not clear. We cannot wait for a Commission decision. In cases brought on the basis of Article 81, although the judge is not presented with sufficient evidence to conclude whether he is confronted with an anti-competitive behaviour or agreement, he may nonetheless be able to reach a 'negative' decision if he can

1 - Compatibility, Efficiency, Legal Security

49

deduce from the evidence before him that illegal state aid is also involved. Indeed, illegal state aid is a matter of Community public policy. The national judge has the obligation to take into consideration EC rules on state aid, though of course such rules will be applied in conformity with national procedural rules. Somebody was speaking earlier about the objectives of the Treaty. Of course, the national judge has to apply all Treaty provisions, and in this perspective he has to look at the objectives of the Treaty. We have to apply the previous Article 5 of the Treaty, Articles 28 and 30, and the Treaty provisions on State aid; we have to apply a lot of rules. These rules grow every day. The problem is that many people consider that national law and European law are two separate things. I do not think this is so, I think there is only one law; I think there is one European law that includes national law. Even if a national judge applies national law, he has to take into consideration that a Community Directive might exist, for example, and that the national law has to be interpreted in the light of this Directive even if the case is purely national. We would otherwise be in total contradiction with the economic reality. The judge has a major role to play in the application of all Treaty provisions, and also the application of other Community legal instruments. My last comment is about legal certainty. From my point of view, there is no legal certainty at present, even when the Commission gives an exemption. If a case is brought before a national judge, the parties have executed the agreement the terms of this agreement are different from those that had been brought before the Commission and that were subject to an exemption. Then of course the judge cannot say: 'I am bound by the exemption'. He has to look at reality. The problem is what the judge should do, because he can ask the European Court of Justice for preliminary rulings only on legal issues, not factual ones. I am not sure whether, in such cases, the national judge can ask the European Court of Justice for a preliminary ruling. He would probably have to go back to the Commission and ask what its position is on the matter. All this confusion originates from the separation that has been made between national law and European law. This is one of the reasons of the misunderstanding. • ALEXANDER SCHAUB—I wanted to express my strong agreement with three points made during this discussion. The first is the point made by Mario Siragusa: we clearly need more Commission decisions on the substance of European competition rules, and not less. The purpose of the reform is to achieve this. How? There will be more decision-makers; not only the Commission, but also national competition authorities and national courts, and we will therefore have considerably more decisions. As you know, only five percent of the notifications are now closed with formal decisions. We have a lot of these strange animals called 'comfort letters', that were validly criticised by the companies up to the point when we started the reform. Then, all of a sudden, they discovered how marvellous these comfort letters are.

50

Panel Discussion

My second point concerns what Ian Forrester has said on the old issue of Article 81 (1). Mario Siragusa is the world's greatest expert on this, and we have discussed it here in Florence from the very beginning. I remember that the first year when we discussed it, the conclusion of the majority was 'what a pity that in the European legal culture we cannot tackle a change from one day to another of what has been practised over forty years'. At that time, the majority was saying 'as we can't do it in this way, we have to do it in a different way', but I fully agree that this does not mean we should forget about a certain reconsideration of Article 81 (1). In the early years a very broad interpretation of Article 81 (1) was not at all crazy. On the contrary, it was very intelligent, because when you start to develop a Europe-wide competition policy from scratch, it makes a lot of sense to have a broad area under control. Today it does not make such sense anymore, and I would very much favour what I think the Courts in Luxembourg are already doing, that is to be progressively more restrictive in interpreting the criteria of Article 81 (1), though without disruption in its application. This can only be a progressive exercise, and I think the Commission could pursue this exercise more actively. And maybe it could and should be pursued more actively by the Courts in Luxembourg also. The third point is about the role of the judge. I must say I find our continued discussion about this more and more surrealistic. We are in the 21st century, where every normally constructed person can easily discover that in our society not only legal aspects, but also directly intertwined economic and complicated technological aspects, are of major importance whether we like it or not. In such a new context, to remain with the idea that a judge can at best decide some legal stuff, but one cannot by definition expect that he has any economic or any technological understanding, seems to me to indicate a total resignation from society. We need judges at the level of what the society in this century needs, and I do not have the slightest doubt that the judges can do it. The funny thing is that most of the judges, as we have heard during this debate, have no fundamental hesitation. It is only the arrogant 'non-judges' who come to the conclusion that unfortunately these poor and regrettable human beings are not up to the job. • BARRY HAWK—I have a concrete proposal to make for the September draft Regulation. The proposal rests on many of the ideas and questions that were raised this morning, starting with Mr Marenco's very fundamental point that we do not know what the scope of 81 (3) is. Maybe you could just tell the Council of Ministers to adopt my proposal and be done with it The proposal is that, first, the full competition analysis should go into Article 81 (1), and second, only the 'non-competition analysis' should be kept under Article 81 (3) (environmental, industrial policy, employment, or whatever). However, those 'non-competition' objectives must be enumerated and go into Article 81 (3). In addition, the Commission should retain the monopoly to grant exemptions under Article 81 (3), and then publish its exemption decisions explaining why it

1 - Compatibility, Efficiency, Legal Security

51

decided to exempt an anti-competitive agreement for environmental, industrial policy, employment reasons etc. • SIMON BISHOP—My comments really echo what Barry Hawk just said from an economic perspective. Doing the whole economic analysis under Article 81 (1) makes a lot of sense. But if you want to maintain the distinction between Article 81 (1) and (3), it seems that the Commission's proposal to introduce market share thresholds determining the scope of Article 81 (1) is the right way to go, provided that we keep an open mind about agreements falling outside the market share threshold. Then the competition analysis can be done under Article 81 (3). Under this sort of approach, the objectives of Article 81 (3) should be totally clear because we are talking about introducing a more economic effect-based approach to enforcement, in which the whole focus is on consumer welfare and the economic impact of agreements. This has nothing to do with environmental standards, or employment considerations, or even Single Market issues. To the extent that these are still important in European policy, I think each of these objectives should be clearly distinguished, and it must be recognised that there will be major tension between competition objectives and something like Single Market objectives in some cases. • SANTIAGO MARTINEZ LAGE—I will not refer now to the role of national courts—as I think that is the subject of one of tomorrow's panels—but to the role of the Commission. Mr Monti started this morning by saying that the Commission thinks it's making an inefficient use of its limited resources, and so from now on it wants to concentrate on important problems and leave nonimportant problems to national judges and national competition authorities. Along the same line, the White Paper stated that the Commission thinks that it is receiving too many notifications, and most of them concern innocuous, inoffensive agreements and uninteresting matters. As a practitioner I do not share this affirmation. I think—and I could say that most practitioners, I hope, would share my opinion—that we 'notify' only of agreements that really raise problems. A lot of other agreements are not 'notified'. Another belief of the Commission, as expressed by Alexander Schaub, is that the Commission's mission is not to grant individual legal certainty but to implement EC competition policy. In my opinion, this way of thinking does not sufficiently evaluate the very important preventive function that the Commission is performing through the present notification system (in the sense that, when we 'notify' agreements that raise delicate issues, the Commission immediately reacts by saying 'be aware that we do not grant immunity from fines ...', or sometimes might warn: 'we will withdraw immunity fromfines,be aware that you have problems with your agreement'). The parties to the agreement should modify the agreement in such cases. Very often, with or without a comfort letter from the Commission, or without a formal notification, the Commission manages to avoid allowing an anti-competitive agreement to enter

52

Panel Discussion

into force and damage the market. If there will be only a posteriori control in the future, this may happen only years later. Those who have suffered damages can be compensated, but the market has already suffered the damage. The Commission should consider more seriously the possibility of voluntary notifications, a real tool for making competition policy. This is in line with Claus Ehlermann's thesis, and that of Mario Siragusa. I would be in favour of a kind of voluntary notification system and a right to obtain 'positive decisions' following such notifications. I think this would also help the Commission to prevent some dangerous agreements from entering into force and leaving the remedy of their consequences until two or three years later. • KAREL VAN MIERT—Thank you all very much indeed for your valuable contributions. Let me just wrap up this session by telling you that if I ever had the slightest impression that the White Paper and the updating of the EC competition rules would lead to a less effective European competition policy, I would never have taken the responsibility for that. But I continue to believe very strongly that this updating and modernising does lead to a more effective competition policy. Second, there has been this debate about what other policy considerations might be included under Article 81 (3), such as industrial policy. We are bound to create a real internal market. So we have to think ahead. I do not believe that with the new enforcement system we will have a real problem in the sense that the Commission might include old-fashioned industrial policy considerations in the application of EC competition rules. We have already got rid of that and it is hard to see how it could return, particularly at the level of the European Commission. Do not forget that we now have a European economy that is much more market-driven, and it will continue to be much more marketdriven in the future. EC competition policy has been used to integrate economies, that is true, and I never did apologise for that. We chose to try to stimulate and integrate the European economy through the application of competition rules. This is what we have been doing by applying Article 86, and this is something that is gradually becoming part of a fully-fledged European competition policy. On the day that a real integrated market is reached through the Euro and the Economic and Monetary Union, this aspect of our job will be reduced. One should not forget about it when we are talking about introducing a new competition policy approach for the next decades. Within the Commission we have moved from a formalistic legal approach to being completely open-minded about economic and technological realities and developments. This is a trend that will continue into the future. There will be less certainty in some respects; this is part of this development. As a last thing, I want to say: you have been talking here about crazy judges; there might even be crazy Commissioners, but not crazy Competition Commissioners!

PANEL ONE COMPATIBILITY, EFFICIENCY, LEGAL SECURITY

WORKING PAPERS

I Simon Bishop Modernisation of the Rules Implementing Articles 81 and 82

I. Introduction In April 1999 the EC Commission unveiled its proposals for reform of the rules implementing Article 81 in a White Paper.l The White Paper includes some radical proposals to decentralise the implementation of Community competition rules to Member States and to give a greater focus to economic analysis. One of the key issues raised by these proposals is the issue of legal certainty. In particular, the intention for the analysis under Article 81 to be based more on an assessment of economic effects suggests a case-by-case analysis in which the role of formalistic rules of thumb is necessarily limited. Under the proposals there will therefore be a trade-off between economic analysis and legal certainty. This paper assesses the arguments put forward in the White Paper in the sense that 'legal certainty' will be largely unaffected by the proposed changes. In particular it assesses two issues: a) whether the application of EU competition law is sufficiently transparent and known so that the business community will be able to limit any legal uncertainty to an acceptable level, and b) whether under the proposed system, individual positive decisions and comfort letters are no longer needed. The remainder of this paper is organised as follows. Section II provides an overview of the White Paper and the general issues raised by it. Section III discusses whether business will be able to assess the likely competition law consequences of contracts into which they will enter. In particular, this section assesses whether the case law provides an adequate guide. Section IV discusses whether under the proposed system individual positive decisions and comfort letters are no longer required.

II. Too much paper and not enough competition analysis In the White Paper the Commission states that a change in its procedural framework is required to ensure a balance between effectiveness of policy and simplification of control. Procedural change is required due to the ever-increasing 1 European Commission: White Paper on modernisation of the rules implementing Articles 85 and 86 of the EC Treaty, OJ C 132 of 12.5.1999.

56

Working Paper I

burden on the European Commission to deal with notifications under Article 81. This burden created by the current enforcement of Article 81 has two sources. First, the Commission, in applying Article 81 (1), has adopted a very broad interpretation of what constitutes a restriction of competition. Article 81 (1) states that all agreements that restrict competition are unlawful unless they meet certain conditions that allow for an exemption under Article 81 (3). This implies that nearly all contracts between firms are caught by Article 81 (1). In consequence, the Commission has come dangerously close to concluding that all contracts are therefore in breach of 81 (I).2 Second, agreements that breach 81 (1) are void and unenforceable before they are notified to the Commission, and only the Commission has the power to grant an exemption to such agreements under Article 81 (3). The need for notification of all restrictive agreements to the Commission has led to a mountain of paperwork and bureaucracy devoted to trivial agreements, and at the same time has prevented the Commission from devoting its time to the important issues. In addition, attempts to relieve the scarce resources within the Commission have had a stifling effect on business. Although the Commission has developed tools such as block exemptions in order to ease the burden, the formalistic nature of these tools has tended to divorce Commission policy from a more rigorous analysis.3 Given this situation, change is clearly required. With refreshing candour, the White Paper summarises the situation as follows: 'It is essential to adapt the system so as to relieve companies from unnecessary bureaucracy, to allow the Commission to become more active in pursuit of serious competition infringements, and to increase and stimulate enforcement at national level.'4

1. The options for reform—and the Commission's preference There are two broad categories of options outlined in the White Paper. One is to stay with the existing authorisation system, in which the Article 81 prohibition can be lifted only by an act of the EC enforcement agency. The other is a move to a directly applicable exemption system, whereby the Article 81 prohibition does not apply to agreements that do not meet certain criteria. Such a system avoids the need for prior notification. 2

The European Courts have on several occasions indicated that this is too restrictive an interpretation, but these indications do not seem to have changed the Commission's practice. 3 However, the Commission has successfully reduced notifications in a manner that has better economic rationale through its de minimis notice (European Commission: Notice on agreements of minor importance which do not fall under Article 85(1) of the Treaty establishing the European Community, OJ C 372 of 9.12.1997). 4 White Paper, paragraph 10.

I - Compatibility, Efficiency, Legal Security

57

The White Paper considers reforms of the existing authorisation system. But none of these really deals with the problems associated with that system. It is therefore unsurprising that the White Paper comes out clearly in favour of a radical shift to a directly applicable exemption system.5 The key feature of this proposal is that agreements will no longer need to be validated in advance by the Commission in order to be lawful. This places a responsibility on businesses to take a view on whether their agreements meet the criteria for exemption. At the same time, it places a responsibility on the Commission and the member state competition authorities to give guidance oh when agreements will be considered lawful.

III. Shifting the burden to business? The introduction of a directly applicable exemption system raises the issue of legal certainty. Under any system of law, it is desirable that firms should be able to enter into contracts reasonably secure in the knowledge that these contracts will be enforceable. But under Article 81, the danger that contracts may be rendered void by a legal challenge on the grounds that they contain restrictions that should have been notified to the Commission has been one of the main factors behind the excessive burden on the system.6 The proposed move to a directly applicable exemption system removes the element of legal uncertainty created by the problems of non-enforceability of contracts that, whilst they are benign, contain some element of restrictiveness. However, such a system still contains an element of commercial uncertainty where an existing agreement is found to be anti-competitive—i.e. the agreement (a) falls within the scope of Article 81 (1) and (b) does not warrant an exemption under Article 81 (3). What is more, this element of uncertainty is to be increased by the Commission's proposals to withdraw the old block exemptions which provide safe harbour to any agreement that meets a certain standardised form, and to make the new set of block exemptions conditional on some kind of market power test, embodied in a market share threshold. Yet in economic terms, such a change is entirely justified. Although Article 81 applies to all agreements that entail 'the prevention, restriction or distortion of competition', this immediately raises the question of when competition is prevented, restricted or distorted. The current interpretation holds that many agreements may be 5

Of the authorisation system options, the third, involving an ability to do ex post analysis, seems to have the greatest appeal. 6 In a few high profile cases, firms have been challenged in the courts on the enforceability of contracts that had not been notified. Thus, for example, the singer/songwriter George Michael tried unsuccessfully to escape from his recording contract with CBS/Sony on the grounds that it should have been notified but was not.

58

Working Paper I

covered by Article 81 (1), regardless of the market position of the parties involved or of the economic impact of the agreement. In other words, the legal interpretation of what constitutes a restriction is extremely wide. Such a wide interpretation captures many agreements that have adverse effects for competition. From a policy perspective, whether an agreement can be retained to prevent, distort or restrict competition depends on its impact on market outcomes and its precise form. Standard economic theory states that, unlessfirmspossess and exercise market power, they are unable to affect adversely competition. In other words, in the absence of market power, agreements between firms do not restrict or distort competition. To illustrate this point, consider a shoe manufacturer who wishes to restrict the supply of its products only to those retailers who agree not to stock the products of other manufacturers. Is such an exclusivity agreement anti-competitive? The answer is that it depends on the facts of the particular case. For example, if the shoe manufacturer has a relatively small market share of retails sales of shoes—say 10 per cent—then it is unlikely that the exclusivity agreement would adversely affect competition, since competing shoe manufacturers can easily gain access to other retail outlets not subject to such restrictions. On the other hand, if the shoe manufacturer were able to sign up 90 per cent of shoe retail outlets the impact on competition might be different. It is therefore clear that the same type of agreement can have widely different impact on competition. A restriction agreed between two firms with no market power has different economic effects from an identical set of terms concluded by parties with appreciable market power. Modern competition law enforcement must take business reality into account. The White Paper's proposal to conduct a more economic approach to the application of Article 81 (1) is to be welcomed. Hence, the move towards an economic effect-based approach in enforcement will involve the evaluation of complex economic issues where it is hard to draw generalisations between cases, despite superficial similarities. This raises the question of how undertakings are to know whether their agreements fall foul of the law. It might be argued that the ex post prohibition of agreements found to be anti-competitive raises no particular concerns for the enforcement of Community competition rules. But such arguments rest on the assumption that those businesses that fall outside the proposed block exemptions are able to determine with precision whether their agreements either fall outside Article 81 (1)—a relatively straightforward issue to assess—or qualify for an exemption under Article 81 (3)—a much more complex task. Most agreements—with the notable exception of explicit cartel agreements—cannot be said a priori to be anti-competitive. Indeed, in many cases, a given agreement may have both proand anti-competitive effects that require a careful assessment of the relative balance between them. Given that that the new enforcement system is bound to create some additional uncertainty, the crucial question is how does the Commission propose to

1 - Compatibility, Efficiency, Legal Security

59

limit the damage. The White Paper claims that there are four areas in which predictability (if not certainty) will be created:7 —The existence of 40 years experience of implementation and case law; —The use of wide ranging block exemptions based on market share thresholds; —The adoption of Commission guidelines and individual decisions to clarify the scope of its application of Articles 81 (1) and 81 (3); and —The implementation of preventive and corrective mechanisms to ensure consistent and uniform application of Community law by national authorities and courts. 1. The existence of 40 years experience of implementation and case law The White Paper suggests that business will be able to infer the treatment of specific agreements by reference to the lessons that can be learnt from the 40 years of case law and decision-making practice, which is said to have 'largely clarified' the law. This is a bold statement to make, for two reasons. First, the fact that the Commission is proposing a radical change in implementation, the removal of block exemptions from undertakings with market shares above certain thresholds (likely to be set at a level below that traditionally signifying dominance), and a more economic approach to the application of Article 81, indicates that the new regime will differ significantly from the old one. Any decision-based 'precedent' that can be gleaned from the old system is therefore likely to provide an unreliable guide to future policy and enforcement. Second, and more importantly, the adoption of a more economic approach reduces the usefulness of 'precedent' when assessing the likely competitive effects within a particular industry. As discussed above, a particular type of agreement can have widely different effects on competition depending on the particular characteristics of the industry—including inter alia the market position of the undertakings concerned, the nature of the products or services and the nature of competition. This assessment is necessarily complex and must properly be conducted on a case-by-case basis. Precedent provides only minimal guidance. 2. The use of wide-ranging block exemptions based on market share thresholds The White Paper promises new block exemptions that will cover a much wider range of agreements—'the vast majority'. The effect of these new block These are to be found at para. 78 of the White Paper.

60

Working Paper I

exemptions will be to give those undertakings that are able to benefit from them greater freedom in running their respective business. This is an admirable intention that should be welcomed and supported. However, while this commitment provides legal certainty for many undertakings, for those to whom the proposed block exemptions do not apply, this is not the case. This wider range of block exemptions is to be conditional on the firms involved falling below a market share threshold. The Commission's current thinking on vertical restraints, for example, involves a near per se approval for all vertical restraints, provided the relevant market share lies below a 30 per cent 'market power' threshold. Such an approach is justified on the basis of economic theory. However, there remain some important issues to be resolved. First, for those firms that do not fall below the market share threshold, legal uncertainty will remain. It is likely that many undertakings with market shares above the relevant market share threshold will not wish to enter into contracts before undertaking a detailed legal and economic assessment. Even those that do enter into agreements, legal uncertainty may persist for some time. Second, there is the issue of market definition. Although the framework for defining relevant markets is now widely accepted, it would be wrong to see market definition as simply a mechanistic task. Moreover, market definition under Article 81 and Article 82 raises a number of difficulties that are not encountered in merger inquiries.8 This means that market definition and any subsequent market share analysis can sensibly play only a screening role. The market share threshold approach is appropriate, therefore, only if the competitive effects of agreements entered into by undertakings with market shares above the threshold, including those with dominant market shares, are genuinely assessed without prejudgment.

3. The adoption of Commission guidelines and individual decisions to clarify the scope of application of Articles 81 (1) and 81 (3) The Commission also promises enforcement guidelines and more published decisions aimed at explaining its enforcement policy in those difficult areas where firms exceed the market share safe harbours. This is in principle good news, signifying the Commission's determination to enter into an open debate on the economic objectives of competition law with industry and its external legal advisers. Such a debate should help clarify where the often hazy line between pro- and anti-competitive effects is to be drawn. However, the stated need for such enforcement guidelines and published decisions is at odds with the claim that existing case law provides sufficient guidance to future implementation. The application of Community competi8 See Baker S. and S. Bishop: Market Definition in Non-merger Inquiries, Research Paper for the Office of Fair Trading (forthcoming).

1 - Compatibility, Efficiency, Legal Security

tion law is therefore likely to continue to evolve and may entail some seemingly radical departures from previous case law. Indeed, it is the proposed change in the analytic approach that warrants the need for such publications. As the White Paper notes, the forthcoming notices and guidelines that explain the Commission's policy and provide guidance for the application of Community competition rules by national authorities are particularly well suited to the interpretation of rules of an economic nature.9

4. Ensuring the consistent application of Community competition rules As noted above, one of the stated reasons for the proposed reforms is the need to deal with the ever-increasing burden on the European Commission to deal with notifications under Article 81. As the White Paper notes: 'In an enlarged Community with more than twenty Member States, centralised detection of, and action against, infringements of the competition rules will be increasingly inefficient and inappropriate. Application of the rules will have to be decentralised more to the Member States' competition authorities and to the national courts.'10 The decentralisation of European competition law enforcement to national competition authorities and to national courts appears to offer a ready solution to this problem. Yet we should not pretend that such a proposal raises no concerns. The devolution of greater powers to implement Community competition policy brings with it the possibility of inconsistency in the application of those rules. That fear is recognised by the Commission, and the White Paper promises to implement a series of measures to ensure consistent and uniform applications of the EC Competition rules throughout the Community. It is not clear how such safeguards would work in practice. Even at the level of specialised antitrust authorities, differences in approach and interpretation of competition law exist throughout the Community. This can be seen even in dealings with the Commission, where the composition of the case team can have an important bearing on the approach taken in the case. These differences in approach are likely to be even more evident once the enforcement of EC competition law is devolved to the Member States. These differences arise both from different rationales for the establishment of national competition laws, and, to even a greater extent, from different legal traditions. While there is convergence between Member States in the area of competition law, these differences in approach still persist, and that needs to be recognised. Moreover, national courts are not well equipped to deal with competition issues. The application of EU competition involves addressing complex 9 10

White Paper, paragraph 86. White Paper, paragraph 42.

61

62

Working Paper I

economic issues, and the Commission's own reforms will require national courts to keep up with higher enforcement standards. While the same fundamental economic issues have always been present in the application of Community competition rules, the Commission's own approach in enforcement is becoming increasingly more economics-based. This trend will raise difficulties in the consistent implementation of Community competition law by the national bodies. National courts are often reluctant or unable to consider some of the core economic issues that are routinely raised in competition matters. My own experiences before national courts illustrates some of these difficulties. In one case, the national court was required to consider some detailed econometric analysis on an issue contested by experts on both sides. It was clear that the judge had neither the necessary training nor the support to evaluate properly the competing claims. In another case, the relevant national court was asked to assess whether an agreement infringed Article 81. Although largely sympathetic to the economic analysis that was presented to him, the judge remarked that this type of case was better dealt with by the specialised competition authority, which, in his words, was more able to assess the economic arguments. This is not to denigrate the ability of the judges, but merely to point out that the evaluation of economic issues, particularly those of an empirical nature, is particularly complex. Indeed, similar views were expressed by Mr Justice Ferris, a UK High Court judge, to the House of Lords' EC Select Committee on the reform of EC competition law procedures. When asked about the ability of judges to apply the kind of complex economic criteria required by Article 81 (3) of the EC Treaty, and which are, of course, replicated in the UK Competition Act, Mr Justice Ferris stated that 'the sort of feeling that judges can decide economic issues to my mind is wholly misplaced . . . They cannot make value judgements except in a very limited field, certainly not in relation to general economic questions.' In principle, these are not insurmountable problems. Indeed, it should be recognised that the courts pay an important role in the implementation of US antitrust rules, as the recent Microsoft case illustrated. However, the role played by US courts is rather different to that envisaged by the White Paper for the national courts of Member States. One possibility would be for judges to be given training in the application of competition law on both legal and economic aspects.11 Yet even such training introduces its own delays and level of bureaucracy. A further area of concern is whether the implementation of Community competition rules by national competition authorities and national courts creates greater scope for protectionism. The White Paper raises this possibility with reference to those Member States that have chosen to remain outside the 1 ' Training for judges was recently suggested by the Confederation of British Industry in relation to the introduction of the UK Competition Act 1998. This new competition law is closer to the EC competition law model.

1 - Compatibility, Efficiency, Legal Security

63

Euro. Yet such protectionism is as likely, if not more likely to arise within the Euro-zone countries, particularly those where the culture of national champions has been traditionally prevalent.

5. Summary The proposals contained in the White Paper are to a large degree predicted on the belief that undertakings are more than able to understand and foresee how Community competition rules will be implemented. But the foregoing considerations indicate that those businesses falling outside any new block exemption rules will face considerable legal uncertainty as to the applicability of their agreements.12 Such uncertainty will increase the burden on business. While business will be able to conduct their own legal and economic analysis prior to entering into agreements, and thereby reduce legal uncertainty somewhat, only in rare instances will such analysis provide unambiguous answers as to the outcome of any potential Commission investigation.

IV. A need for individual positive decisions and comfort letters The proposed changes will require the Commission to provide on-going guidance not only to national competition authorities but also to undertakings. In this sense, there is a need for published Commission decisions, both of prohibition and positive ones. It can also be argued that one of deficiencies of the current system is that it leads to too few published decisions. This not only hinders transparency, but also the development of a consistent policy approach. Moreover, this can be contrasted with Commission's approach to the control of mergers, where a decision is published following each notification, whether or not a detailed investigation is required.13 The adoption of market share thresholds will allow the Commission to focus on fewer cases, allowing a more detailed analysis of the issues arising in those particular cases. This, combined with the requirement to publish a decision, will help to shape the Commission's policy in a consistent and coherent manner. The White Paper states that positive decisions need only be taken in exceptional cases, where it may be necessary to provide guidance on the Commission's approach to certain restrictions. Yet, as mentioned above, one type of agreement properly exempted in one case need not necessarily qualify 12

As noted above, a distinction can be drawn between cartel agreements where the illegality of the agreement is clear and other types of agreements where this is usually not the case. 13 Unless the notification is withdrawn.

64

Working Paper I

for being exempted in other circumstances. It is, therefore, important that decisions be adopted whenever the Commission opens an investigation. Under the EC Merger Regulation, the publication of the Commission's decisions has probably done more to generate transparency and consistency than any other development in EC competition law. The same should be true in the application of Article 81, once a series of reasoned Commission decisions dealing with proand anti-competitive agreements between firms possessing some market power builds up. In particular, positive decisions will be needed for those undertakings whose market share lies above the block exemption thresholds to be introduced by the Commission. In those cases where the agreement does not give rise to anticompetitive effects, and therefore meets the criteria of 81 (1), a positive decision, or at a minimum a comfort letter, should be issued. In the absence of such a decision, the undertaking will continue to face legal uncertainty as to whether its methods of commercial activity are acceptable or liable for challenge in the future. Under these circumstances, pro-competitive investments may be severely affected. Positive decisions, as well as negative ones, will therefore provide important signals as to the likely application of Article 81 (3), especially to undertakings with market shares above levels traditionally associated with dominance.

V. Conclusions The basic thrust of the White Paper is a welcome one. The application of Article 81 needs to be transformed from a drawn out paper shuffling exercise into one that reflects the reality of real-world markets. The proposals contained in the White Paper represent part of a general and welcome shift towards greater emphasis on economics in the application of competition law in Europe. However, a number of issues remain unresolved by the proposals. As an economist, I would point to the following: 1. The reliability of market share threshold tests. Market share thresholds play a crucial role in limiting uncertainty to those cases where a firm has market power, but no one who has experience with the application of competition law will imagine that market share tests provide unambiguous guidelines. Moreover, it is important that market shares are not seen as a tolerance threshold in which agreements entered into by undertakings falling outside the proposed block exemptions, particularly those with a dominant market share, are perceived as being per se anti-competitive. 2. Article 81 's 'schizophrenic' nature. Many of the problems encountered in Article 81 cases arise from the Commission's dual role as a competition agency and as guardian of the Single Market. Refusal to acknowledge that

1 - Compatibility, Efficiency, Legal Security

65

these objectives may conflict can reduce the transparency of the decisionmaking process.14 3. The sheer 'newness' of the challenge the Commission is facing. As discussed above, existing case law provides little useful guidance and the approach set out in the White Paper requires the Commission to invent a whole new approach to the economic analysis of restrictive agreements. Of course, there are models (e.g. the US model) from which to learn, but the learning curve the Commission is facing is a steep one. The Commission must have the confidence to break with tradition if it is to take the right path in policy development. The Commission's quest for modernisation in this area is to be both welcomed and supported. While the proposed new system will generate uncertainties of its own, the benefits of a less formalistic assessment and a streamlined procedure are clear. Indeed, in many cases, those who now shout loudest about the unpredictability and the 'chaos' created by the proposed reforms will be those who have in the past complained most bitterly about the 'straight-jacket effect' of the existing system.

14

For example, where pro-competitive effects of exclusive sale territories can only properly be achieved by market segmentation.

II Dr. UlfBoge The Discussion on the Modernisation of EC Antitrust Policy: An Update on the Bundeskartellamt's Point of View

I. Preliminary remarks Since the publication of the White Paper over a year ago, the Commission's reform proposals have been intensively discussed and commented upon. Various working groups and meetings of Directors-General have, in the meantime, contributed towards clarifying the position of the Member States. Progress has also been made on the content of the reform itself. The Fifth EU Competition Law and Policy Workshop is a good opportunity to take stock of the situation and to gain insight into the latest developments in the discussion. In response to numerous comments, some of them critical, the Commission has specified its proposals in a series of non-papers. On this basis it has been possible to discuss the main problem areas in more detail, particularly those concerning the compatibility of the reform with the EC Treaty, the different types of decisions and their effects under the new system, the achievement of coherence by setting up a network of competition authorities, the role of the courts and,finally,the issue of legal certainty for undertakings. This discussion, which is still in full swing, is important and necessary. The Commission, which in the initial stages focused mainly on just defending its approach against Member States' objections, has recently been more prepared to enter into a real debate. There is, of course, no doubt that the changes to European competition law enforcement procedures proposed in the White Paper will have a direct and appreciable effect on national competition legislation and on its application. The expected pressure towards harmoriisation under the new enforcement system would make it impossible for Germany to maintain its system of preventive control. In addition, the Bundeskartellamt would be placed under a new framework of conditions relating to the decentralised application of Article 81 of the EC Treaty. Yet the possible need to adapt is no reason for the Bundeskartellamt to refrain from participating in the discussion. The ultimate aim of the reform can only be to more effectively enforce the principle of competition. The Commission too has declared that it does not desire any reduction in the level of protection for competition. And that is precisely the reason why we will have to look into all the questions involving the competitive effects of the new procedural rules. It is therefore a good thing that the Commission is not putting anyone under pressure, and that it set the year 2003 as the target date for starting to implement the reform.

68

Working Paper II

Whether the proposals contained in the White Paper can be realised merely by introducing a new regulation, therefore without amending the EC Treaty, is a matter that has not been fully clarified. Yet we do not wish to go into that debate at this stage. The topics of interest here are, above all: —The fundamental decision concerning the abolition of the notification and prior administrative approval system in conjunction with the re-evaluation of horizontal co-operation; and —The issue of decentralisation and how it can be implemented as consistently as possible. Let us look first at the points that are more likely to cause problems.

II. The abolition of the notification and prior administrative approval system It is well known that, from the start, there were major objections in Germany to the proposal that the notification and prior administrative approval system be completely abolished. In Germany's experience, the obligation to notify enables important information to be obtained, opens up room for manoeuvre and has a preventive, deterrent effect on undertakings. Discussions with the other Member States and the Commission, however, have shown that, of those positive aspects of a notification system that Germany considers important, namely transparency and the preventive function, at least the latter is not convincing at the European level. One reason for this might be that notifications do not have any preventive effect in the Commission's enforcement practice. However, the transparency function of the notification system would also to a large extent disappear in the new system, despite its high practical relevance. By contrast to the area of vertical agreements, in the case of horizontal agreements it cannot be automatically assumed that there will be an influx of complaints in cases of restraints of competition, as there is no conflict of interests among those involved in the agreements.

1. The introduction of an information system on horizontal agreements Considering the above, we introduced into the discussion a proposal in the sense that the necessary degree of transparency be achieved by obliging enterprises to briefly inform the national competition authorities of their co-operation arrangements. This information would simultaneously be made available to the public, for example on the Internet. Vertical agreements would be

/ - Compatibility, Efficiency, Legal Security

exempted from this information obligation, as would horizontal agreements covered by block exemptions. The legal effectiveness of agreements would remain unaffected as long as the criteria for exemption under EC Article 81(3) were fulfilled, as no decision would have to be taken. Competition authorities and affected third parties would, however, be able to act and obtain more detailed information if they saw reason to do so. In this way, a kind of early warning system would be created in the so-called 'grey area'. With it would come a degree of de facto legal certainty, as cartel authorities and third parties would be informed at an early stage. However, no legitimate expectations would be involved in this procedure. The obligation to provide brief information of this kind would be an appropriate, suitable and necessary corrective measure in order to create transparency if the preventive exemption system was to be called into question. Unless such an amendment to the reform proposal is made, it is hardly conceivable that the reform would be widely accepted in Germany.

2. Leniency programs for hard-core cartels Hard-core cartels would, of course, be just as unlikely to notify in a directly applicable exemption system as they were under the current enforcement system. In Germany, we have been thinking hard for some time now about how cartels could be more robustly combated. The most important step we have undertaken was to recently announce a leniency program. This is influenced by the corresponding Commission notice, but it is also based on the American model of automatic leniency for the first undertaking to report a cartel. 3. The reassessment of horizontal co-operation agreements Along with the abolition of the requirement to obtain an exemption, the more economic effect-based approach in enforcement envisaged by the Commission also causes headaches in the Bundeskartellamt. The term itself does, of course, have positive connotations. But the term alone will not automatically achieve the common goal of more efficient competition law enforcement in the Member States. If 'more economic' ultimately meant that all horizontal agreements involving market shares of 15% or less would not fall under Article 81 (1), this might be 'more economic' for the competition authorities and the courts. However, suppose there is no longer any competition in the German furniture industry, for example, becausefivelarge manufacturers have joint buying arrangements that are all below this threshold. In this case, we would have not a 'more', but a 'less' economic approach. It is therefore clear at this juncture that reform of the

69

70

Working Paper II

procedural law cannot be viewed in isolation from current considerations about substantive law. As far as reforming the block exemption regulations for vertical agreements is concerned, there have been discussions about the fact that such agreements should be re-evaluated in the future, and why this is so. However, a similar discussion about horizontal agreements still has to take place. The recent draft block exemption regulations on specialisation and on research and development, as well as the directives for applying Article 81 of the EC Treaty to certain main types of horizontal co-operation agreements, take a new approach, namely that horizontal agreements have anti-competitive effects only when associated with market power. This means that the previously applicable requirements for an appreciable effect with respect to Article 81(1) will be considerably raised. We have the impression that a new competition policy model is being developed, along the lines of the rule of reason approach taken in US antitrust. In particular, the new distinction between agreements which merely have the effect of restraining competition and those in which such a restraint is the object of the agreement is reminiscent of the American distinction between agreements 'may well be reasonable' and those are 'per se unreasonable'. In justifying its new approach, the Commission refers to numerous judgements of the European Court of Justice. Yet if one examines this jurisprudence, one discovers that it mainly relates to vertical restraints. Therefore the Bundeskartellamt has considerable doubts as to whether the Court's jurisprudence does in fact permit such conclusions to be drawn. Another question is whether one can actually say that horizontal agreements are not necessarily damaging. From our perspective, there are major differences between horizontal and vertical agreements. Restraint of competition by vertical agreements is usually limited to the fact that there is a natural conflict of interests between the supplier and the purchaser, whereas competitors at the same stage of production or distribution generally have the same aims. The anti-competitive potential of horizontal agreements is thus much higher. Of course, this does not mean that horizontal agreements cannot also promote competition. That is, after all, the reason why there are exemptions from the ban in German competition law as well. Yet we should carefully consider whether we really want to send a definite signal that horizontal agreements other than price-fixing, production and market allocation agreements, are generally acceptable (or are acceptable if they fall below an appreciable market share threshold). Finally, like the draft block exemption regulations, the current version of the directives leaves much room for interpretation and manoeuvre, even in the area of hard-core restraints such as price-fixing and market sharing. This gives the Commission considerable discretionary powers and therefore room for manoeuvre in decision-making, which is a matter that really has to be discussed.

1 - Compatibility, Efficiency, Legal Security

71

These questions concerning the level of protection in substantive law terms are closely related to the modernisation of procedural law and require further clarification.

III. Decentralised enforcement First of all, it should be emphasised once more that the Bundeskartellamt fully supports the broad objectives of the White Paper. The Bundeskartellamt also believes that a refom of the existent enforcement system is urgently needed. This also applies to efforts towards greater decentralisation. For years the Bundeskartellamt has been advocating an extension of the competence to apply Article 81(3) towards the competition authorities of Member States, thereby doing justice to the increased significance of the principle of subsidiarity. The President of the Commission, Mr Prodi, has on several occasions emphasised his commitment to bringing Europe closer to its citizens by applying this principle more intensively. The Directorate-General for Competition is following this course by abandoning its exemption monopoly. However, decentralising enforcement brings with it problems such as how best to distribute the cases and how to ensure uniformity in the way the law is applied. These problems were already manifest during the notification and prior administrative approval system, yet they would become more acute during the transition towards a directly applicable exemption system. Regardless of how decentralisation is organised, a reasonable solution needs to be found to these questions. 1. The allocation of cases The Commission suggests that the allocation of competencies should be based on a number of criteria, including the nature of the case, the size of the relevant geographic market, the extent to which intervention would be effective in ending a contravention of the law, appropriate punishment and the location of the evidence. This list of criteria certainly appears to be practicable, and the results would probably be similar to those achieved by basing the allocation system on an assessment of the economic and competitive focus, as the Bundeskartellamt suggested in an earner Working Paper on decentralisation. However, all in all, we should not overestimate the potential for conflict of competencies. 2. Coherence and legal certainty By far the most difficult question is how to ensure the uniform application of EC competition law. This requires, above all, a consistent system of

72

Working Paper II

types of decisions and appropriate mechanisms of co-operation and coordination. Questions arise in relation, for example, to the so-called Commission positive decisions and opinions. Neither of these types of decisions is strictly in line with the directly applicable exemption system, and they should therefore be abandoned in favour of consistency in the new approach. If the aim is that under the new system undertakings assume greater responsibility, a certain level of legal uncertainty in the 'grey area' is simply the other side of the same coin. Of course, that will not prevent competition authorities from continuing in the future to informally discuss critical agreements with the undertakings. This is what it will boil down to in practice anyway. And this also makes sense if we do not want to have courts that are unable to operate because of their burdensome caseloads. The President of the European Court of First Instance has already expressed concerns of this kind. The degree of legal certainty that could previously be achieved through Commission binding exemption decisions cannot be achieved through declaratory decisions and informal discussions. Nevertheless, if one decides in favour of a directly applicable exemption system, this is partly because in the past Commission formal exemption decisions were issued only on rare occasions, and frequently such decisions have not been requested in European competition law enforcement. If under the new enforcement system the Commission will issue positive decisions and opinions, what form should these take? Certain criteria have been developed about when the Commission should issue opinions, yet there is still not enough clarity about their practical value and field of application. It is suggested that the Commission should be able to issue positive decisions in cases raising particularly important issues from the perspective of the Community interest, whereas national competition authorities should issue exemption decisions only when a formal decision is required in order to settle a case under national law. This would mean that the Bundeskartellamt would probably not be able to use this type of decision without a change in its procedural law. Yet it is precisely because the Bundeskartellamt regards them as being alien to the concept of a directly applicable exemption rule that it sympathises with the idea of comprehensively restricting the scope of application of such decisions. However, this should not lead to small and medium-sized enterprises being placed in worse positions than large companies. It is only natural that the latter will more easily be able to point out the existence of a significant Community interest. Conversely, small and medium-sized enterprises might have a greater need for clarification, because they are less likely to have the resources necessary to carry out a competition law analysis of their own. This is why it should be made possible for the national competition authorities to act to a greater extent on their own initiative, even if some of them do not yet have many years of experience in applying European law. The consultation and co-ordination mechanisms that are envisaged for both the Commission and the national

1 - Compatibility, Efficiency, Legal Security

73

competition authorities before making prohibition and approval decisions or withdrawing the benefit of block exemption regulations to certain agreements appear to be quite suitable means of achieving the required measure of uniformity. At the same time, the obligations to provide information should be kept as simple as possible, and decisions should be dealt with in more detail by the Advisory Committee only in really important cases.

3. National courts No detailed investigation has yet been carried out as to how the national courts will apply EC competition law in practical terms and what will be the effect of decentralisation from this perspective. Coherent application of EC competition rules would certainly be more easily ensured if the competence to deal with competition cases is concentrated in a few specialised courts. This is the case in Germany, and France is also planning to introduce a similar procedure. But not all Member States will follow this course. The Commission made a range of proposals on how to alleviate the inherent conflict between decentralisation and coherence in enforcement, such as comprehensive obligations to provide information, an amicus curiae function for the Commission, or even its right to appeal against decisions by national courts. It has also laid down instructions on how to interpret the EC competition rules, through positive decisions, opinions or guidelines. However, these 'safety mechanisms' are not totally unproblematic, among other things because of the fact that the desired reduction in the Commission's workload might again be unintentionally lost. At the same time, Germany believes that the courts' independence must not be infringed upon and the principle of the separation of powers must be observed. The topic of coherence should not be concluded without stating that, in our view, the highest national courts and the European Court of Justice are ultimately the guarantee that the law will be uniformly applied.

IV. Summary To sum up, the objectives of the reform, namely to apply the law more efficiently and less bureaucratically, to ensure legal certainty for undertakings, and to achieve decentralisation and coherence, objectives that we all aspire to, cannot all be achieved to their full extent at the same time. These objectives can, however, be optimised by placing the right emphasis on each. By proposing an information system, the Bundeskartellamt is attempting to find an effective compromise among these objectives. Above all, this will prevent the one outcome nobody involved in the debate wants: a watering down or undermining of the principle of competition.

Ill Ian S. Forrester, Q. C.! The Modernisation of EC Antitrust Policy: Compatibility, Efficiency, Legal Security

A Constitution cannot make itself; somebody made it, not at once but at several times. It is alterable; and by that draweth nearer Perfection; and without suiting itself to differing Times and Circumstances, it could not live. Its Life is prolonged by changing seasonably the several Parts of it at several times.2

I. Introduction The European Commission has proposed big changes to European competition law, the biggest changes since 1962, in terms both of substance and procedure. Change has been called for since at least twenty years, and resisted for about eighteen years. It is interesting that Director General Ehlermann presided over an administration which, broadly speaking, set its face against such fundamental reforms (though there were occasional flutters of boldness in theoretical individual decisions). Yet, translated to Florence, he has encouraged the discussion of outrageous heresies. Whether Dr. Ehlermann is to be regarded as a Luther (an insider who became a successful outsider), or as a Becket or Savonarola or Tyndall3 (three clerics who died violently but whose efforts are better judged by posterity), it is sure that these meetings of a small number of enforcers, professors, academics and practitioners have contributed to the process of reform. Indeed, at the preceding meetings of this group in Florence in 1996 and 1997 (and 1998, to some extent), the faces of Commission officials flushed and blushed to hear severe criticisms: the European system of interpreting and applying the Treaty's basic rule on anti-competitive conduct did not work, could not work, had to change. These remarks were not new, but the circumstances were different: the new Director General confronted by his persecutors, confronted by the statistics of a case-load which was impossibly 1 The author gratefully acknowledges the very generous assistance and advice offered by his colleague Ann Stanley, as well as a number of other helpful persons who have answered technical, policy and legal questions. 2 This quotation from the seventeenth political figure, Lord Halifax, appears in Senator Jordan's Preface to the annotated version of the Constitution of the United States of America, published by the Congress in 1972. 3 William Tyndall believed in lay access to the Bible. He emigrated from England to Vilvoorde, near Brussels, for peace to continue his translation efforts, but was captured there by secret agents of Henry VIII of England.

76

Working Paper HI

heavy if old doctrines were maintained, allowed some staff to voice heresy, to think seriously about radical change, and finally to propose it. The last century's approach to enforcement of competition by the European Commission was too much dominated by textual analysis of written clauses rather than by economic analysis of business reality. The system as established following the adoption of Regulation 17/624 did not function properly, and sometimes did not function at all. The problem had several elements: • Article 81 (I) 5 was given a broad scope and caught all sorts of agreements, including those which were basically pro-competitive but which contained restrictive clauses. There was little analysis of the merits or seriousness of the restriction, merely a noting of its presence, which entailed the consequent triggering of Article 81 (1). • This meant that thousands of basically desirable agreements were caught by the prohibition of Article 81 (1), void under Article 81 (2), andfinableunder Regulation 17/62. Only an exemption could cure these hazards; and only the Commission could issue an exemption. • Decisions by the Commission were rare because every decision was a major piece of rule-making, not a mere response to an individual problem. Perfection and comprehensiveness were the goals, not adequacy. • There was not enough application of the law in a transparent and procedurally predictable manner. Because too many agreements were prohibited, most agreements could not receive an exemption. • A theoretically necessary but practically unavailable exemption was therefore not a source of legal certainty, but of legal uncertainty. • To avoid this uncertainty, block exemption regulations came to be regarded as black letter, obligatory standards, and failure to respect them was dangerous and close to illegal. Various commentators suggested how the Commission could react to the unsatisfactory procedural situation. I argued that there were three ways.6 It could share more responsibility with the Member States, whose enforcement resources are notionally limitless; or relax its interpretation of Article 81 (1) to release basically benign agreements from the difficulties of needing an exemption which was practically unobtainable; or render the taking of decisions very much simpler and quicker, so that an individual decision became no more than an answer to a specific problem rather than a long, public, detailed ex cathedra pronouncement of general importance. I argued that richness of output rather 4 Council Regulation 17/62: First Regulation Implementing Articles 85 and 86 of the Treaty, 1956-1962 OJ Spec. Ed. 5 This paper will use the new numbering given to the articles of the EC Treaty based on Article 12 of the Treaty of Amsterdam (e.g. Article 81, 82, etc.), except in passages quoted verbatim and in contexts where use of the old numbering is more appropriate from a historical perspective. 6 Forrester I. (1994): 'Competition Structures for the 21st Century', Fordham Corporate Law Institute, Matthew Bender & Company, Inc.

1 - Compatibility, Efficiency, Legal Security

than perfection should be the goal. To general surprise, the reforms have been more radical than expected, with both procedural and substantive aspects. We were asked to consider whether the proposed reforms of the competition rules are workable, whether they are constitutional in the sense that they match the Treaty of Rome, and whether the world of law and business will be better off once they are implemented. I will consider these questions from the historical, textual and practical viewpoints, as to which I have three themes: legal certainty; the desirability of retaining a notification system; and the discrepancy between the new law being adopted in Brussels and the old law being received into the competition laws of the Member States and, a fortiori, the candidate States to the East. As to that discrepancy, I fear that the law which is likely to be applied nationally may well be old-fashioned: the process of decentralisation is very laudable in principle, in that it promises more enforcement, more law and a better permeation of competition law doctrines into the daily lives of regulators and business people; yet there is a risk that the law applied locally may be the old unreformed doctrines which the European Commission wishes to abandon, doctrines preoccupied by textual analysis and black-letter rules, rather than the reality of how a given marketplace works.

II. Can the reforms be squared with the words of the Treaty? 1. Historical The most enjoyable portion of this paper has been the conducting of some legal archaeology. I felt it would be pleasing to know whether and how the founding fathers had discussed what we are today discussing. If the drafters had crafted Article 85 (1), (2) and (3) of the Treaty of Rome with the firm confidence that effective enforcement could be achieved only by a rigid and wide-reaching prohibition which could be cured only by a Commission decision, then the Commission's Directorate General for Competition ought to be very cautious before scrapping the regime so presciently crafted by giants with legendary names like Spaak and Snoy et d'Oppuers. It is of course true that the European Court of Justice has not so far found it appropriate to go back to the Minutes of early intergovernmental meetings to interpret the Treaty. It has not referred in its judgments to the Minutes of meetings at the Messina Conference, or of drafting sessions at Val Duchesse, the castle on the eastern edge of Brussels where the negotiations took place. These documents might be said to be equivalent to the Federalist Papers, a famous dialogue between famous founding fathers about the future governance of the United States (centralised or decentralised power?), and sometimes cited as relevant to the framers' original intent. These and other early authorities are sometimes invoked in US constitutional

11

78

Working Paper III

debates, although sparingly. Chief Justice Marshall said that the framers of the Constitution expected that it should endure for ages to come and consequently be adapted to various crises in human affairs. It has never been suggested in Europe that EC constitutional doctrines must be interpreted according to the original intent of the founding fathers, nor in light of the Minutes of the drafting of the Treaty of Rome. Although there were many similarities between French agricultural rules and the EEC agricultural regime, the Court has never looked at the former for help in understanding the latter. However, the Court does make a comparative law review of national laws and jurisprudence in appropriate cases.7 So a 'bad' historical precedent would not be an incurable impediment to reform. However, a positive or favourable historical indicator could be helpful reassurance that the new regime is not contradicted by the earliest records relating to competition law. A request was therefore made to the keepers of the Council's Archives, and in March 2000 Ann Stanley and I spent some happy hours browsing.8 The documents available are the French language Minutes kept by the forerunners of the Council Secretariat, along with a few texts from the States involved.9 Those documents can be corroborated by an article written by Arved Deringer in 196310 and by early books on competition law.11 I have also spoken to some of the now-retired officials who worked in the Competition Directorate General in the early 1960s. 1.1. The big issues to be settled During the six months from September 1956 to February 1957, the negotiators had the task of reaching consensus on the competition law portion of the Treaty. There were broadly three models. The German approach contemplated preventive control: an anti-competitive arrangement should be void unless approved. The French approach left more authority to the enterprise, so that the arrangement could be implemented without prior approval, but could be rendered retroactively void by the authorities. The Dutch (there was also a Belgian-Dutch variant) provided for compulsory notification and provisional validity; if the competition authority showed that it was abusive, the invalidity 7 See the as-yet-unpublished papers presented at the colloquium on The Role of Comparative Law in the Emergence of European Law organised by the Swiss Institute of Comparative Law, Lausanne, April 2000. 8 The patience, good humour and creativity of the staff in charge of the Archives of the Council, notably Mr. Sanchez Martin, are warmly to be commended. 9 Giuliano Marenco of the Commission Legal Service has indulged in some parallel archaeology, and has set forth the references to all the documents to which he refers in a table. For posterity's benefit, a similar table is annexed to this article. 10 Deringer A. (1963): 'Les regies de la concurrence au sein de la CEE', 54 Revue du marche commun, 256-66. 11 Such as that by the late Graupner R. (1965J: The Rules of Competition in the European Economic Community, Martinus Nijhoff, one of the first serious studies in English of the subject.

1 - Compatibility, Efficiency, Legal Security

79

would take effect only prospectively. There was concern about price discrimination and refusals to supply based on grounds of nationality and even race or religion. There were thus a lot of what would now seem extraneous issues, and there were also real divergences both as to what were the enforcement targets and how enforcement should be structured. The secretariat of the Intergovernmental Conference for the Common Market and Euratom prepared a chart, dated September 24, 1956, which showed the main lines of divergence. It is attached as an annex to this article. In early September 1956,12 the German delegation considered that the provisions on this topic should use as their starting point the prohibition of agreements/cartels. However, exceptions should be allowed for certain kinds of agreements/cartels, it being clearly understood that these exceptions would be subject to checks against abuse.13 Conversely, the French delegation wanted a system similar to its own legal exception system,14 and was supported by Belgium, '[which] considered that it might be desirable, for psychological reasons, to replace the absolute prohibition of agreements/cartels by the notion of checks against abuse.'15 The French proposal was the closest to what finally emerged: 1. All situations or practices involving arrangements or monopolies with the object, or which may have the effect, of hindering the exercise of competition, are incompatible with the Common Market, in particular: —fixing or determining resale or sale prices; —restricting or controlling production, technical development and investments; —partitioning markets, products, customers and sources of supply; —permitting the absorption or domination of the market for a product by an undertaking or group of undertakings.

12

The following quotation, and all subsequent quotations of texts from this early period of Community history, are unofficial translations prepared for the purposes of this article. The original French texts are given in footnotes. 13 '. . . la delegation allemande a insiste pour que les dispositions en cette matiere partent du principe de I'interdiction des ententes. Des exceptions devraient toutefois etre admises pour certaines formes d'ententes, etant bien entendu que ces exceptions seraient soumises au controle d'abus.' Excerpt from the Minutes of Meetings of the Common Market Group (extrait du Proces-Verbal des reunions du Groupe du marche commun), Sept. 3-5,1956, Council Archives, CM3/NEGO/236, doc. MAE 252/56 at II(B); see also Sections 1 and 6, German Law Against Restraints of Competition of July 27,1957, published in R. Graupner (1965), supra note 11, Appendix B. 14 Deringer A. (1963), supra note 10; see also Articles 59bis and 59ter of the French Price Ordinance No. 45-1483 of June 30,1945, as supplemented and amended, published in R. Graupner (1965), Appendix C, supra note 11. 15 '[qui] s'est demande s'il ne serait pas opportun de substituer, pour des raisons psychologiques, le principe du controle des abus a celui de I 'interdiction absolue des ententes.' Extract from the Minutes, supra note 13; this was September 1956.

80

Working Paper III

2. Situations or practices whose authors are able to prove that they effectively contribute to the improvement of production or putting into circulation, or to the development of technical and economic progress, may be exempted from the above provisions. [The authors] must also show that the consumer receives a legitimate share of the profit resulting from these measures. 3. (Provisions concerning State and public service monopolies to be inserted).16

1.2. Discriminations based on nationality, race or religion There is no sign, however, that the delegations attributed overwhelming importance to what we are discussing forty-four years later in Florence. They were looking at older doctrines. Should price discrimination based on nationality be prohibited? Should discrimination be addressed in the competition articles or in a separate Treaty provision? At the end of October 1956, there was apparently unanimity among the Delegations 'to limit the ban on discrimination to those practices which are effected within the Common Market and where the operators who are treated differently are in competition with each other. .. Moreover, the making of price differentials in comparable transactions in order to obtain a dominant position should not be dealt with by the ban on discrimination but within the framework of rules on unfair competition.'11 The reasoning seemed to be that, although the Single Market's replacement of the six national markets would seem to necessitate a general prohibition of discrimination on the basis

16 I. Sont incompatibles avec le marche commun toutes les situations oupratiques d'ententes ou de monopole ayantpour objet oupouvant avoir pour effet d'entraver I'exercice de la concurrence, en particulier:

—enfixant ou en determinant lesprix de revient ou de vente; —en restreignant ou en controlant la production, le developpement technique et les investissements; •—en repartissant les marches, produits, clients et sources d'approvisionnement; —enpermettant I 'absorption ou la domination du marche d'un produit par une entreprise ou un groupe d'entreprises. 2. Peuvent etre relev&es des dispositions precedentes, les situations ou pratiques dont les auteurs sont en mesure dejustifier qu'elles contribuent effectivement a I 'amelioration de la production ou des debouches, ou au developpement duprogres technique et economique. Us devront egalement justifier que le consommateur obtient une part legitime dans le profit qui decoule de ces mesures. 3. (Dispositions a inserer concernant les monopoles d'etat et services publics). Id. Annex (Doc. Mar. Com. 17). 17 'pour limiter I'interdiction de discrimination aux discriminations qui sont commises a Vinterieur du marche commun et ou les partenaires a des transactions qui sont traitees de facon differentielle, sont en concurrence entre eux. .. En outre, il semble que les majorations ou minorations deprix operees, lors de transactions comparables, en vue d'obtenir une position dominante sur le marche ne devraient pas etre traitees dans le cadre de I 'interdiction de discrimination, mais devraient faire Vobjet de dispositions eventuelles relatives a la concurrence deloyale.' Note of Oct. 26, 1956 du President du Groupe du Marche Commun, Council Archives, CM3/NEGO/217, Doc. MAE 468 f/56.

1 - Compatibility, Efficiency, Legal Security

81

of nationality, 'the opening up of the common market is not a reason to forbid discriminations committed for non-commercial motives, especially since discriminations based on non-commercial motives such as religious or political beliefs are rarely encountered in commercial life.'18 The Chairman of the Groupe du marche commun therefore proposed as a basis of discussion the following text: Within the Common Market, it shall be prohibited to apply dissimilar terms in similar situations to buyers or sellers in competition with each other (according to their nationality), in particular to ask from them or offer to them different prices.19 Indeed, the working group discussed a number of versions of Treaty articles that would have formally condemned discrimination in economic matters based upon nationality. In mid-November 1956, the full text of Article 4Oa20 as proposed by the Chairman of the working group read: Discriminatory practices carried out by commercial partners in competition with one another for reasons of nationality are prohibited within the Common Market.* (However, undertakings are not prohibited from establishing their offers on the basis of the conditions offered by other undertakings with regard to delivery). * The German and Italian delegations have proposed that this article be deleted. However, in a conciliatory spirit, the Italian delegation would be prepared to agree to the article being worded as follows: 'If it is established that competition in the Common Market is distorted by discriminatory practices whose effect is to place buyers or sellers in the other Member States at a disadvantage owing to their nationality, the Council may, on a proposal from the Commission, unanimously adopt regulations with a view to prohibiting such discrimination.'21 18

'. . . I'ouverture du marche commun nest pas ... un motif de reprimer d'autres discriminations commises pour des motifs non commerciaux, d'autant plus que les discriminations fondees sur Vorigine ou sur les convictions politiques ou religieuses, par exemple, ne jouent guere de role dans la vie commerciale.' Id. 19 '// est interdit, a Vinterieur du marche commun, d'appliquer des conditions inegales a des transactions comparables, avec des acheteurs ou vendeurs en concurrence entre eux (suivant lew nationalite) et notamment de leur demander ou de leur offrir des prix differents.' Id. 20 Article 40 dealt with the principle of discrimination on grounds of nationality. Article 40a contained t h e prohibition of discrimination. Article 41 dealt with antidumping. Article 42 was the ancestor of Article 85 (1). T h e numbering of the Articles and paragraphs, however, changed week by week. 21 Sont interdites a Vinterieur du Marche commun les pratiques discriminatoires exercees, en raison de la nationalite, a Vegard de partenaires commerciaux en concurrence entre eux. * (Toutefois il nest pas interdit aux entreprises d'etablir leurs offres d'apres les conditions offertespar d'autres entreprises, a I'endroit de la livraison). * Les delegations allemande et italienne ont propose de supprimer cet article. Toutefois, dans un esprit de conciliation, la delegation italienne serait disposee a accepter que cet article soil redige comme suit: 'S'il est constate que la concurrence dans le Marche commun est faussee par la pratique de discriminations ayant pour effet de desavantager des acheteurs ou des vendeurs des

82

Working Paper III

Two weeks later, the delegations had agreed to eliminate the Article concerning discrimination on the basis of nationality.22 It is evident that the wisdom of dealing with anticipated nationalist commercial behaviour in the private sector was vigorously discussed and then discarded. With respect to discriminatory rules adopted for the public sector, it was decided that the promulgation of these should be prohibited, rather than prohibiting that they be obeyed. 1.3. A discursive early review of competition problems The French and the German delegates had a difference in approach as to whether unilateral price discrimination by a non-dominant company should be legal.23 The drafters also pondered whether abusive monopolies should be addressed together with cartels or in a separate provision. The French view was dirigiste, suspicious of price discrimination, perhaps rather tolerant of horizontal arrangements under supervision. The conflicting approaches to enforcement emerge clearly:24 • M. Muller-Armack (D) noted that: 'the question of discrimination or price differentials should not necessarily be considered from the point of view of practices which harm competition. Dumping, dual pricing etc. are often perfectly compatible with a free market. When it does not give rise to abuse and is not based on discrimination for reasons of nationality, discrimination in itself is in no way harmful to the competition regime but, on the contrary, is one of its normal features.'25 autres Etats membres en raison de leur nationalite, le Conseil, sur proposition de la Commission, peut prendre a I'unanimite tout reglement en vue de I 'interdiction de ces discriminations. ' Proposal submitted by t h e President with a view to drawing up provisions relating to o b s t r u c t i o n s to c o m p e t i t i o n , N o v e m b e r 14, 1956, Council Archives, C M 3 / N E G O / 2 3 6 , D o c . M A E 541 f/56. 22 T h e working group considered that: 'the field of application of the provisions concerning discrimination for reasons of nationality goes beyond the scope of the competition rules, and that these provisions, which contain one of the fundamental principles of the Common Market, should be placed at the beginning of the Treaty, for example as Article 2bis' ('le champ d'application des dispositions concernant la discrimination en raison de la nationalite depasse le domaine des regies de concurrence et que ces dispositions, contenant un des principes fondamentaux du marche commun, auraient leur place au debut du Traite, par exemple, comme article 2bis.'). Projet de proces-verbal des reunions d u G r o u p e tenues les 2 7 - 2 9 Nov. 1956, C o u n c i l Archives, C M 3 / N E G O / 1 4 6 , D o c . M A E 785 f/56 (Dec. 10, 1956). 23 M e m o interne of Sept. 7, 1956, Fascicule 5, Council Archives, C M 3 / N E G O / 2 3 6 , D o c . M A E / S e c . 29/56. 24 Id. 25 'La question de la discrimination ou de la differentiation de prix ne doit pas etre envisagee necessairement sous Vangle des pratiques nuisibles a la concurrence. Dumping, double-prix, etc sont souvent parfaitement compatibles avec un libre marche. Les discriminations en elles-memes, lorsqu'elles ne donnent pas lieu a des abus et ne remontent pas a des discriminations de nationalite, ne prejugent en rien le regime de la concurrence mais, au contraire, representent un aspect normal de celui-ci.

1 - Compatibility, Efficiency, Legal Security

83

M. Donnedieu de Vabres (F) 'on the contrary, insisted on the need to draw up precise legal rules with regard to price discrimination and differentials.' He further noted that: 'not all arrangements and concentrations should be condemned, since they may be compatible with economic progress.'26 M. Muller-Armack (D) stated that: 'a clear distinction should be drawn between monopolies and oligopolies on the one hand, and cartels on the other. Monopolies and oligipolies are not necessarily . . . incompatible with a competition regime. What must be abolished ... [are] the abuses to which certain monopolistic situations might lead.'27 He suggested: 'for cartels,. . . the principle of an absolute prohibition, while foreseeing the need to authorise the existence of certain cartels which could exceptionally be compatable with a competition regime.' M. Donnedieu de Vabres (F) responded that: 'the meaning of the French word "arrangement" does not correspond to the term "kartell" used by the Germans. It has a wider meaning, which is why it requires a more supple regulatory regime than that proposed by Mr. Muller-Armack. Moreover, authorised exceptions must have a special character, and be allowed on a case by case basis and not generally, as the French Delegation [sic] seems to accept.'28 M. Van Tichelen (B) 'declared that for psychological reasons it was dangerous to provide for condemnation of arrangements in principle, but that it would be advisable to let certain arrangements continue to exist for some time.'29 M. Linthorst-Homan (NL) responded that: 'psychological arguments must not win out against the objective needs of the Common Market, and that— for commercial and financial circles—it would be better to know the contents and scope of the Treaty from the outset instead of being obliged to dither in a situation of uncertainty.'30 26

'au contraire, insiste sur la necessite d'elaborer une reglementation juridique precise en matiere de discriminations et de differenciation deprix... .ilnefaudraitpas condamner toutes les ententes et toutes les concentrations, car ellespourraient etre compatibles avec le progres economique.' 27 'ilfaut bien distinguer entre monopoles et oligopoles d'un cote et cartels de Vautre. Les monopoles et les oligopoles ne sont pas necessairement . . . incompatibles avec un regime de concurrence. Ce qu'il faut supprimer. . . [sont] les abus auxquels certaines situations monopolistiquespourraient aboutir ... [q]uant aux cartels,.. . leprincipe de I'interdiction absolue, tout enprevoyant la necessite d'autoriser Vexistence de certains cartels, qui, exceptionnellement, pourraient etre compatibles avec un regime de concurrence.' 28 'le terme 'entente' au sens francais du mot, ne correspond pas a celui de 'kartell' employe par les allemands. Ilestplus large que le second et voila pourquoi ilexige une regiementation plus souple que celle prevue par M. Muller-Armack. En outre, les exceptions autorisees devraient avoir un caractereparticulier, etre admises Caspar cas et nonpas d'une facon generate comme semble I'admettre la delegation francaise' [sic]. 29 'declare qu'il est dangereux, pour des raisons psychologiques, de prevoir une condamnation duprincipe des ententes, mais qu'ilserait opportun de laisser subsister certaines ententes pendant quelque temps.' 30 'les raisons d'ordre psychologique ne doivent pas I'emporter sur les necessites objectives du marche commun et qu'il vaut mieux—pour les milieux commerciaux et

84

Working Paper III

• M. Muller-Armack (D) suggested as a solution to 'the problem of sanctions against the formation of arrangements',31 the adoption of 'the principle of the declaration of nullity "ipso jure".' • M. Van Tichelen (B) favoured notification: 'imposing an obligation to notify on undertakings.'32 Thus at the beginning of the drafting process, the founding fathers were by no means agreed either on what constituted problems or on what the remedies might be. The notions of refusal to supply and price discrimination which French law did not discard until 1996, forty years later, were much debated. The notion of nullity now provided in Article 81 (2) is mentioned first by the German delegate. It appears one month later in a Note by the Chairman of the Groupe du Marche Commun:33 'The abusive exploitation of a dominant position in the market must be governed by the same principles whether that position results from the scale of production of an undertaking or from an arrangement between several undertakings. However, there are differences between the situation of agreements and that of monopolies, which must be taken into account when drawing up the legal requirements concerning them. Thus, for example, one of the characteristics of agreements is that they are likely to hinder or prevent competition, and legal requirements should be introduced bearing in mind this effect and the extent to which it is produced or sought. With regard to monopolies, on the other hand, the more complete the monopoly, the less probable is it that any competition likely to be compromised or eliminated will exist. As a result, what should be prohibited in the case of monopolies is not hindrance of competition but only abuse of the dominant position in the market. . . . Finally, sanctioning the annulment of legal operations which lead to the acquisition of a dominant position in the market should probably apply only to agreements.'34 financiers—de connaitre des le debut le contenu et la portee du Traite plutot que d'etre obliges de se balancer dans une situation d!incertitude.' 31 ' le probleme des sanctions contre la formation des ententes . . . le principe de la declaration de nullite "ipso jure."' 32 ''I'imposition aux entreprises I 'obligation de I'enregistrement.' 33 D o c . M A E 468 f/56, supra note 17. 34 Lexploitation abusive a"une position dominante sur le marche doit etre regie par les mimes principes, que cette position resulte de I'ampleur de la production d'une entreprise ou d'une entente entre plusieurs entreprises. Toutefois, il existe entre les ententes et les monopoles des differences de situation dont ilfaut tenir compte en redigeant les prescriptions qui les concernent. C'est ainsi, par exemple, qu'une des caracteristiques des ententes est qu'elles sont susceptibles d'entraver ou d'empecher la concurrence, et c'est en consideration de cet effet et dans la mesure oil cet effet seproduit ou est recherche qu'ily a lieu de les introduire. En revanche, en ce qui concerne les monopoles, la concurrence susceptible d'etre compromise ou eliminee existe d'autant moins que le monopole en question est plus parfait. En consequence, dans le cas des monopoles, ce nest pas lefait d'entraver la concurrence, mais bien seulement I'abus de la position dominante sur le marche qui pourrafaire I'objet d'une interdiction. . . . Enfin, il parait opportun de reserver aux seules ententes la sanction de la nullite des operations juridiques qui conduisent a I'acquisition d'une position dominante sur le marche.

/ - Compatibility, Efficiency, Legal Security

85

At the beginning of November 1956 the Expert Group issued a proposal which also contained a provision embodying the notion of nullity:35 Agreements or decisions prohibited by virtue of paragraph 1 above [referring to the prohibition paragraph] are null and void in law and may not be relied on before any Member State jurisdiction.36 Agreements declared under the terms of paragraph 2 above [referring to the exemption paragraph] but prohibited by the Commission37 are equally null and void in law.

The delegations' positions on the proposed text can be culled from the following exchange:38 • Muller-Armack (D) 'expressed his full preference for a ban in principle on agreements, which was also accepted by the German legislation. B u t . . . this principle did not seem to be really envisaged in the proposal. . . , since even if point 1 [of the first article] provided for this ban, point 2 introduced too many exceptions. For this reason it would be preferable not to retain the 'principle of a ban' at international level, when this principle was not accepted in German national legislation; it would be more advisable simply to include in the Treaty a general declaration stating the incompatibility of arrangements with the Common Market, and to leave it up to the Member States themselves to make sure that this principle is put into practice during the first phase, through close collaboration and reciprocal consultation.'39 • M. Van Tichelen (B) 'declared he was very worried about the effect which the German proposal seemed likely to have on the objectives assigned to the Treaty and asked whether Mr. Muller-Armack would be prepared to adopt a

35 Projet de Redaction, Nov. 12, 1956, Council Archives, CM3/NEGO/217, Doc. MAE 527 f/56. A long document with some useful illustrations of alternative approaches to the drafting of the Articles. 36 'Les accords ou decisions interdits en vertu du paragraphe 1 ci-dessus sont mils de plein droit et nepeuvent etre invoques devant aucune juridiction des Etats membres.... La meme nullite frappe les accords declares aux termes du paragraphe 2 ci-dessus mais interdits par la Commission. 37 Reservation expressed by the Belgian expert; he would prefer the prohibition t o be laid down by a court. 38 Memo interne du groupe du marche commun of Nov. 13-15, 1956, Fascicule 10, Council Archives, CM3/NEGO/236, Doc. MAE/Sec. 70. 39 'exprime toute sa preference pour le critere de I'interdiction de principe des ententes, qui est egalement accepte par la legislation allemande. Mais,... ce principe ne semble pas etre reellement envisage dans le projet. . . , car si le (1 [du premier article] prevoit cette interdiction, le (2 introduit trop d'exceptions. Pour cette raison il serait preferable . . . de ne pas se tenir au 'principe de I'interdiction,' sur le plan international, alors que ce mime principe nest pas accepte par la legislation nationale allemande; il vaudrait mieux se limiter a inscrire dans le Traite une declaration generate qui enonce Vincompatibilite des ententes avec le Marche commun, en laissant aux Etats membres eux-memes le soin de veiller a la realisation de ce principe au cours de la premiere etape, moyennement une collaboration etroite et une consultation reciproque.'

86

•

•

•

•

Working Paper III

less extreme position if all the other delegations were ready to accept the principle of the ban adopted in the German legislation.'40 M. Muller-Armack (D) 'replied that his position on this problem was inspired by concern that should the proposed Treaty contain a ban in principle on arrangements, it would encounter strong opposition from certain circles with vested interests.'41 M. Hyzen (NL) 'drew the Group's attention to the hybrid nature of the proposal drawn up by the experts, which seemed to be a mixture of two different systems, one involving a ban and the other simply control of abuses. He ended by saying that every possible effort must be made to reach one single solution.'42 M. Donnedieu de Vabres (F) stated that 'a statement that arrangements and monopolies were incompatible and must be banned should not be abandoned. The fact that national laws on this subject were not harmonised was not an insurmountable obstacle. National laws should transpose the Treaty provision on agreements and monopolies, so that after a certain time international legislation could be drawn up.'43 M. Catalano (I) 'supported the German argument in favour of purely and simply deleting [the first article] which, unlike Mr. Muller-Armack, he thought was wholly superfluous, if not dangerous, since all transactions would be caught by the ban on discrimination which it laid down.'44

In parallel to these discussions in November 1956 about what we would call enforcement choices, the drafters' debate on prohibiting price discrimination on the basis of nationality was proceedings tied to their discussion on whether the rules enunciated in the Treaty should be directly applicable to the Member States. The President noted that most of the Member States at least agreed that

40 'se declare tres preoccupe par les consequences que la proposition allemande semble entrainer a I'egarddes objectifs assignes au Traite et demande a M. Muller-Armack s'il est pret a revenir sur des positions moins reculees au cas ou toutes les autres delegations seraient pretes a accepter leprincipe de I 'interdiction adopte par la legislation allemande.' 41 'reponden disant que sa position a I'egard de ceprobleme est suggereepar le soucique le projet du Traite, au cas oil il enonce le principe de I 'interdiction des ententes, nefasse I'objet d'une forte opposition venant de certains milieux interesses.' 42 'appelle Vattention du Groupe sur le caractere hybride du Projet etablipar les experts qui semble melanger deux systemes diffirents: celui de I'interdiction et celui du simple controle d'abus. II conclut en disant qu'il fautfaire tout effort possible pour aboutir a une solution unique.' 43 'il nefaudrait pas renoncer a I'enonciation de I 'incompatibility des ententes et des monopoles et a celui de leur interdiction. Lefait que les legislations nationales ne soientpas harmonisees a ce sujet nest pas un obstacle insurmontable. Les legislations nationales devraient incorporer les dispositions du Traite en matiere d'ententes et monopoles, defacon qu'apres un certain temps une legislation internationalpourra etre etablie.' 44 'se rallie a la these allemande de la suppression pure et simple [du premier article] qu'il considere, a I'oppose de M. Muller-Armack, absolument superflue si non dangereux, car Vinterdiction des discriminations qu'il enonce nepeut que frapper toute transaction.'

1 - Compatibility, Efficiency, Legal Security

87

the Treaty should only contain fundamental legal principles.45 He concluded that because the delegations could not reach an agreement on the text proposed by the Expert Group, they would refer the matter to a 'groupe restreint' which would consider: 1. a draft compromise setting out the principle of a prohibition of agreements and monopolies, to be drawn up jointly by the German, French and Italian delegations; 2. a draft compromise which would be presented by the President of the Common Market Group; and 3. a proposal setting out the principle of control of abuses regarding agreements and monopolies, to be presented by the Dutch delegation.46 In late November, the working group considered a set of draft competition articles, in which appear both nullity and the prohibition of discrimination on the basis of nationality:47 • first Article: • Discriminatory practices on grounds of nationality by commercial partners in competition with one another are prohibited within the Common Market. • The Council, acting on a proposal from the Commission, may adopt by qualified majority any legislation with a view to prohibiting such discriminatory practices. • second Article: • (reminder: question of anti-dumping legislation) • third Article: (1) All agreements between undertakings, all decisions by associations of undertakings and all concerted practices which are likely to affect trade between Member States and whose aim or effect is to prevent, restrict or distort competition within the Common Market are prohibited, in particular those which consist of: [list of 5 practices, much as they appear in what became Article 85 (1)] 45

M e m o interne of Sept. 7, 1956, Fascicule 5, supra note 23. 1. un projet de compromis formulant le principe de Vinterdiction des ententes et monopoles, a elaborer conjointement par les delegations allemande, frangaise et italienne; 2. un projet de compromis qui sera presente par le President du Groupe du Marche commun; et 3. une proposition formulant le principe du contrdle d'abus en matiere d'ententes et monopoles a presenter par la delegation neerlandaise. Projet de Proces-Verbal of Nov. 13-15, 1956, Council Archives, C M 3 / N E G O / 1 4 5 , D o c . M A E / 5 3 9 f/56; see also Projet de R e d a c t i o n p a r la delegation neerlandaise, of Nov. 15, 1956, Council Archives, C M 3 / N E G O / 2 3 6 , D o c . MAE/547 f/56, a n d Tableau synoptique des projets d'articles of Sept. 24, 1956, Council Archives, C M 3 / N E G O / 2 3 6 , D o c . MAE/325 f/56 (annexed to this paper). 47 Projet de redaction sur les regies de concurrence, Nov. 20, 1956, Council Archives, C M 3 / N E G O / 2 3 6 , D o c . MAE/602 f/56. 46

88

Working Paper III

(2) These prohibitions will not apply to agreements between undertakings, decisions by associations of undertakings and concerted practices with regard to which those concerned can supply proof that they contribute towards improving the production or distribution of products or promoting technical or economic progress, and that users receive a fair share of the resulting profits, and that they do not • impose on those concerned restrictions which are not indispensable for the achievement of these objectives, • give these undertakings the possibility, for a substantial part of the products in question, to fix prices, limit production or market openings or eliminate competition from other undertakings. (3) agreements or decisions which are prohibited by virtue of the previous paragraph are null and void in law and may not be relied on before any Member State jurisdiction.48 It is curious that the third paragraph, establishing nullity, refers back to the second paragraph, which was the precursor of the exemption. Did the drafters consider that nullity should be inflicted upon agreements prohibited by virtue of the second paragraph (as not eligible for an exemption)? If so, we could conclude that the drafters did not, at least at this point, interpret Article 85 as necessarily establishing a prior authorisation system. Only those agreements 48

Article 1 Sont interdites a I'interieurdu marche commun les pratiques discriminatoires exercees, en raison de la nationality, a I'egarddepartenaires commerciaux en concurrence entre eux. Le Conseil, sur proposition de la Commission, peut prendre a la majorite qualifiee tout reglement en vue de I'interdiction de ces discriminations. Article 2 (pour memoire: question de la legislation anti-dumping) Article 3 (1) Sont incompatibles avec le marche commun et interdits tous accords entre entreprises, toutes decisions d'associations d'entreprises, et toutes pratiques concertees qui sont susceptibles d'affecter le commerce entre les Etats membres et qui ont pour objet ou pour effet d'empecher, de restreindre ou defausser lejeu de concurrence a I'interieur du marche commun et notamment ceux qui consistent a: [suit une liste de pratiques interdites plus ou moins telles que celles de iArticle 85] (2) Seront releves des interdictions precedentes les accords entre entreprises, les decisions d'associations d'entreprises et les pratiques concertees au sujet desquels la preuve peut etre fournie par les interesses qu'ils contribuent a ameliorer la production ou la distribution des produits ou a promouvoir le progres technique ou economique, tout en reservant aux utilisateurs une part equitable du profit qui en resulte et sans • imposer aux entreprises interessees des restrictions qui ne sont pas indispensables a la realisation de ces objectifs, • donner a ces entreprises la possibility, pour une partie substantielle des produits en cause, de fixer lesprix, de limiter la production ou les debouches ou d'eliminer la concurrence d'autres entreprises. (3) les accords ou decisions interdits en vertu duparagrapheprecedent sont nuls deplein droit et nepeuvent etre invoques devant aucune juridiction des Etats membres.

1 - Compatibility, Efficiency, Legal Security

89

falling within the prohibition of the first paragraph and not benefiting from the exceptions laid out in the second paragraph would be automatically null and void. One week later, the nullity provision had been changed, and referred back to the paragraph concerning the prohibition, not the exemption!49 Unfortunately, the archival documents do not provide any explanation for the change. Another week brought a further change: the nullity provision, still located in the third paragraph, refers to agreements prohibited by 'the present article.'50 This version is very close to what ultimately became Article 85 of the Treaty, save that paragraphs (2) and (3) are reversed. The drafting group was charged with determining whether the removal of the words 'sont nuls de plein droif would weaken the scope of the paragraph. Thus, the drafters considered that it might be enough that an agreement could not be invoked before a national court. But in a draft dated February 14, 1957, the nullity paragraph was eliminated and the notion incorporated into the first paragraph containing the prohibition: Sont incompatibles avec le marche commun, interdits et nuls de plein droit, tous accords.. . 51 It was not until the late date of February 23, 1957 that the order of the paragraphs appears as it does in the final form of the Treaty.52 At this time, there was also discussion of whether the nullity should be prospective only, and whether the intervention of a national authority would be necessary to establish nullity. It is clear that the nullity provision bounced around frequently before definitively arriving in Article 85 (2), its final rest place. The provision was in paragraphs (3), (2), and (1) at different times. The uncertainty related not merely to its numbering, but, more importantly, to its functioning. Upon which triggering event was the agreement to be void, and from when did the nullity run? Having looked at the texts, it seems impossible to say that they clearly favour one answer more than another. I think it is fair to say that the two paragraphs could be regarded as complementary, each being a part of a single analysis but they cannot—in my view— be regarded as favouring either the control of abuse theory or the prohibition of restrictions theory. Nor do the words or travaux preparatoires indicate the relative weights to be accorded to Article 85 (1) and (3). Were the energy, authority, skills and resources of the enforcing agency to be deployed principally in applying paragraph (1) or paragraph (3)? The Minutes do not much help us as to the nature of the exemption process. Is the agreement which is 49 Projet de Redaction, Nov. 28, 1956, Council Archives, C M 3 / N E G O / 2 1 7 , Doc. M A E 657 f/56. 50 Redaction approuvee, Dec. 6, 1956, Council Archives, C M 3 / N E G O / 2 3 6 , D o c . MAE/788 f/56. 51 Projet de redaction of Feb. 14, 1957, Council Archives, C M 3 / N E G O / 2 3 6 . 52 We found n o explanation of why Articles 40(1), (2) and (3) became Articles 85 (1), (3) and (2).

90

Working Paper III

caught by Article 85 (1) unlawful because it restricts competition but curable by Article 85 (3) because it serves broad competition goals despite its restrictive features? In this case, the entirety of the analysis would be a competition law analysis. Or might the role of Article 85 (3) be to extend to the agreement the benefit of general economic and social goals? In either case, there is a two-step process. Now, the Commission got itself into early difficulties by finding too easily restrictive features that necessitated the use of the exemption mechanism. Instead of blessing the agreement under Article 85 (1) as benign in its entirety, the Commission used the costly remedy of Article 85 (3) (costly because the apparatus of formally deciding to grant an exemption was far too burdensome). Thus to the extent that Article 85 was based on a prohibition theory, then it was used too promiscuously.53 As a result, Article 85 (3) was used for the wrong purpose, to confirm the overall legitimacy of the deal, as opposed to . being reserved for the special occasions where the deal called out for creative thinking, tolerance or the grant of a short-term exceptional approval. A simpler way of saying this is to observe that if negative clearances had been issued more frequently, the Commission could have kept up with the demand for exemptions by sharing the task of blessing agreements with national courts. By contrast, under the regime adopted, national courts had no function, at least in theory, once they had concluded Article 85 (1) applied. The words as the drafters left them are clearly based on a prohibition. The reforms would mean that an agreement containing a restrictive feature would no longer be void until exempted; it would be void if its restrictive features did not satisfy the criteria prescribed in paragraph (3), rather than solely because it had restrictive features. That change is, at least in theory, a profound one. (I argue below that, in practice, the change is not an important one.) However, I do not find in the travaux preparatoires evidence that the drafters would have opposed such an outcome. The travaux preparatoires indicate, I submit, that the drafters could legitimately expect three features in European competition law enforcement measures. First, the prohibition and punishment of the unacceptable. Second, the intelligent and pragmatic examination of doubtful matters with a view to determining in a timely manner whether they are acceptable. Third, abstention from interference in benign matters. Granting direct effect to paragraph (3) is not inconsistent with that division; nor does it seem inconsistent with what the drafters intended. The debate in 1956 and 1957 was not as focused and as scientific as the one we are conducting forty-five years later. The delegates were exploring in a basically amicable and collegiate manner concepts that were unfamiliar to most of them.

53

Read a competition decision like AlcatellEspacelANTNachrichtentechnik, 1990 OJ L 32/19, which lumbers through a series of reasons why jurisdictionally the deal falls within the Commission's clutches (changes in the parties' level of autonomy in various respects), and then records the many virtues of the deal.

1 - Compatibility, Efficiency, Legal Security

1.4. British doubts about what it all meant While thefinishedversions of Articles 85, 86 and 87 of the Treaty of Rome may have pleased the drafters, they left many questions unanswered. On April 10, 1958, the British Embassy in Paris addressed to the French Government an aide-memoire recording that there was: uncertainty in commercial and legal circles in the United Kingdom about the effect of Articles 85,86 and 88 ... on current and future commercial contracts between parties in the United Kingdom on the one hand and parties in the countries of the European Economic Community on the other.54 The UK presented a number of pertinent questions, such as: In order that this uncertainty may be resolved, Her Majesty's Embassy would be grateful if the Ministry of Foreign Affairs would provide answers to the following questions: 1. Article 88 of the Treaty of Rome provides that until the entry into force of the provisions adopted in application of Article 87, the authorities of Member States shall, in accordance with their respective municipal law and with the provisions of Articles 85 and 86, rule upon the admissibility of any understanding and upon any improper advantage taken of a dominant position in the Common Market. To what extent, and in what ways, does the ratification of the Treaty of Rome affect or modify existing municipal law in thisfieldin France? Is a party to a pre-existing contract entitled to repudiate it if it falls within the prohibited class of contracts indicated in Article 85(1) and (2) and is not saved by Article 85 (3)? Would the ordinary commercial courts apply Article 85 if the dispute were brought before them by another party to the contract? (b) What would be the position in the event of any conflict between the substantive rules of existing municipal law in France and Articles 85 and 86 of the Treaty of Rome? 3. Is it possible to say what further legislative steps, if any, in connection with private restrictive practices, are required or contemplated by the Government of France as a consequence of the ratification of the Treaty of Rome? 4. In the event that the answers to these questions indicate that there is doubt whether or how far Articles 85 and 86 are of direct, immediate and independent application in France during the interim period, what steps if any are contemplated by the Government of France to resolve this doubt?55 On June 12, 1958, the French Permanent Representation proposed that the aide-memoire should be examined by a working group of the six Member States 54 Aide-Memoire of April 10, 1958 from U K Embassy in Paris to French Government, Council Archives, CM2/1958/748. Identical memoranda were submitted by embassies in four other capitals. The sixth memorandum perhaps got lost. 55 Aide-Memoire of April 10,1958, of U K Embassy in Paris to French Government, Council Archives, CM2/1958/748.

91

92

Working Paper III

in order to achieve a common line (or, if this were not possible, parallel lines).56 The Commission's reaction was unease. In a memorandum dated June 19,1958, the Competition Directorate General expressed its fears: The Commission fully understands the French desire to avoid any increase in existing levels of uncertainty about the interpretation of these articles, which could arise if six countries were to give different replies to the questions from the British government.57 The Commission (disingenuously or optimistically) expressed its desire to put an end as quickly as possible to this uncertainty, and asked for time to find 'a practical solution.' It asked the Member States to give it some more time; and at the fourteenth meeting of Coreper in Strasbourg on June 20 and 21, 1958, there was a discussion of the subject.58 The Commission offered to propose an answer even before the summer holidays. The Italian delegation favoured a precise and comprehensive answer, but the French feared the need to offer a very complete and thorough reply, and proposed to send not a provisional answer which would in its turn call for an exhaustive follow up, and therefore favoured a simple answer. Belgium was troubled by the fact that there was no competition law in Belgium (has much changed?), and felt the need to offer reassurance to the British. The Netherlands favoured six different replies. Predictably, the outcome was the constitution of a working group which—rapidly—produced a common reply stating that: The co-existence, foreseen by Article 88 of the Treaty, of national laws with the provisions of Articles 85 and 86 of the Treaty—a co-existence which national courts will eventually have to recognise—poses a problem of interpretation which can only be resolved definitively by the case-law of the Court of Justice in the context of the competence which is granted to it by the Treaty. Further, according to Article 89 of the Treaty, the Commission of the Community is charged—when it takes up its duties—with ensuring that the principles laid down in Articles 85 and 86 are applied, which could lead the Commission to adopt a position on this subject. The attention of the Commission has therefore been drawn to the questions raised. The Commission has informed the Government of . . . that it intends, in the shortest possible time, to decide on its position on the most urgent practical questions regarding the application of these Articles. These questions include the problem of what procedure should be followed in order to put an end as soon as possible to the present uncertainties emphasised in the memorandum from the United Kingdom.59 56

N o t e of June 12, 1958 from French Permanent Representation, Council Archives, CM2/1958/748. 57 Commission Memorandum of June 19, 1958, COM/58/132, Council Archives, CM2/1958/748 ('La Commission comprendpleinement le desir frangais d'eviter I'incertitude actuelle sur I'interpretation de ces articles, qui pourrait se produire au cas ou les six pays donneraient des reponses differentes aux questions du Gouvernement britannique.'). 58 Extrait d u Proces-Verbal de la 14eme Reunion du Coreper, June 20-21, 1958, Council Archives, CM2/1958/748, Doc. 636 f/58. 59 La coexistence, prevuepar Varticle 88 du Traite, des legislations nationales et des dispositions des articles 85 et 86 du Traite—coexistence dont lesjuridictions nationales auront

1 - Compatibility, Efficiency, Legal Security

93

This was agreed on July 11, 1958. So far as the Archives could reveal, the Commission never did produce its definitive reassurance about the effects of the Treaty of Rome on industrial property rights and exclusive territorial licensing. 1.5. Conclusion on the archival evidence The drafters were not contemplating the competition rules in the way that we have come to examine them. The text of Articles 85 and 86 was drafted in six months, the actual order of the paragraphs of Article 85 was still unsettled only days before the signing of the Treaty, and the preamble was forgotten until just ten days before signing.60 The year after signature, neither the Member States nor the Commission were able to respond to the UK's questions about how the competition rules would affect executed agreements. The drafters probably did not fully realise the implications that these provisions would have in the ensuing years. These conclusions lead to the following remarks. First, there was no single will or goal on the part of the drafters about the nature of the challenge to be confronted. Moreover, their concerns, to the extent they are discernible, are—not surprisingly—different to those of today's enforcers. They were Babes in the Wood by today's standards. They were evidently worried about price discrimination, especially based on nationality. They were worried about what we would now call intra-Community dumping and anti-dumping measures. A reading of the Minutes does not suggest that they were unanimous in seeing competition as an instrument for deregulation, for adapting business structures and for stimulating economic efficiency and consumer welfare. Second, they certainly looked at every possible variant, procedurally speaking. Should restrictive agreements be prohibited? Unenforceable? Void? eventuellement a connaitre—pose un probleme d'interpretation qui ne pourra etre definitivement tranche que par la jurisprudence de la Cour de Justice dans le cadre de la competence qui lui est attribute par le Traite. Par ailleurs, selon I'article 89 du Traite, c'est a la Commission de la Communaute qu'incombe la tache de veiller—des son entree en fonctions—a Vapplication des principesfixes par les articles 85 el 86, ce quipourrait amener la Commission a prendre position a ce sujet. C'est pour cette raison que Vattention de la Commission a ete appelee sur les questions posees. La Commission a fait part au Gouvernement de . . . de son intention de fixer, dans leplus court delai, sa position vis-a-vis des questions pratiques les plus urgents concernant Vapplication des articles mentionnes. Parmi ces questions figure le probleme de la procedure a suivre afin de mettre fin, le plus tot possible, aux incertitudes actuelles soulignees dans Vaide-memoire du Royaume- Uni. Draft reply by the 6 Governments to the Memorandum from the United Kingdom, Council Archives, CM2/1958/748, Doc. 678 f/58. 60

Commission europeene, DG X et Conseil Universitaire Europeen pour l'Action Jean Monnet (1999): 40 ans des Traites de Rome ou la capacite des Traites d'assurer les avancees de la construction europeenne, Actes du Colloque universitaire de Rome organise a la memoire d'Emile Noel, Rome, March 26-27,1997 (Bruxelles, Bruylant), at p. 63.

94

Working Paper III

Incapable of being relied on before national courts? Invalid after being declared unenforceable by national courts? Ex tune, ex nuncl There were no obvious winners or losers. What they produced was an amalgam of several countries' preferences. Third, the new regime was based upon a text drafted by a committee, a diplomatic compromise. The drafters' governments were themselves unsure of the implications of that text only one year after its drafting had been completed. My impression is of well-intentioned confusion about a range of problems, rather than a focussed, purposeful debate on accurately identified issues. Fourth, Community law has moved on. Now it has the benefit of primacy over inconsistent national law, and of direct effect. The direct effect of Articles 85 and 86 was only established in my professional lifetime in BRTv SABAM.61 These judge-made creations are so important that it is not surprising the earliest texts were obscure without them, and that in judgments such as Portelange62 and Bosch63 the Court found these issues difficult. As a result of all the foregoing, I submit that the answer to the question regarding the compatibility of the White Paper proposals lies primarily in the text of the Treaty itself, as explained by the European Courts and learned commentaries.

III. Textual considerations Can we fairly read Article 85 as allowing what the Commission favours? 1. The ECSC Treaty as a recent example The ECSC Treaty, signed in Paris on April 18, 1951, explicitly provides for prior authorisation. Article 65(1) contains a simple prohibition: All agreements between undertakings, decisions by associations of undertakings and concerted practices tending directly or indirectly to prevent, restrict or distort normal competition within the common market shall be prohibited, and in particular those tending: 61

Case 127/73 Belgische Radio en Televisie et societe beige des auteurs, compositeurs et editeurs v SV SABAM et NVFonior (BRT-I) [1974] ECR 51; commented on by the a u t h o r in ' C o m p l e m e n t o r O v e r l a p ? Jurisdiction of National a n d C o m m u n i t y Bodies in Competition M a t t e r s After S A B A M ' , 1974 C o m m o n Market Law Review, 171. 62 Case 10/69 S.A. Portelange v SA Smith Corona Marchant International and others [1969] E C R 309. 63 Case 13/61 Kledingverkoopbedrijf de Geus en Uitdenbogerd v Robert Bosch GmbH and Maatschappij tot voortzetting van de zaken der Firma Willem van Rijn [1962] E C R 89.

1 - Compatibility, Efficiency, Legal Security

95

(a) tofixor determine prices; (b) to restrict or control production, technical development or investment; (c) to share markets, products, customers or sources of supply. Article 65(2) gives the High Authority/Commission the power to authorise certain categories of agreement: However, the High Authority shall authorise specialisation agreements or joint-buying or joint-selling agreements in respect of particular products, if itfindsthat: (a) such specialisation or such joint-buying or -selling will make for a substantial improvement in the production or distribution of those products; (b) the agreement in question is essential in order to achieve these results and is not more restrictive than is necessary for that purpose; and (c) the agreement is not liable to give the undertakings concerned the power to determine the prices, or to control or restrict the production or marketing, of a substantial part of the products in question within the common market, or to shield them against effective competition from other undertakings within the common market. Authorisations may be granted subject to specified conditions and for limited periods. In such cases, the High Authority shall renew an authorisation once or several times if itfindsthat the requirements of subparagraphs (a) to (c) are still met at the time of renewal. The High Authority shall revoke or amend an authorisation if it finds that as a result of a change in circumstances the agreement no longer meets these requirements, or that the actual results of the agreement or of the application thereof are contrary to the requirements for its authorisation. Thus the drafters of the EEC Treaty had before them the example of Article 65 of the ECSC Treaty. They were confronted with a broad range of possible control mechanisms. Arguably, the most compelling Community precedent would be Article 65, a prohibition of illegal conduct curable under limited conditions by the Commission and by no one else. They elected not to follow it. This seems to confirm the constitutional propriety of ascribing to other bodies the grant of exemptions under the Treaty of Rome.

2. Article 85 (3) does not prescribe who shall decide upon exemptions Paragraph (3) states that the provisions of paragraph (1) may, however, be declared inapplicable.. . . There is no specification of which body may do the declaring: the Commission? The national courts? National regulators or competition authorities? It is clear from the Minutes already quoted that the delegates considered many hypotheses and came to a conclusion that did not decide for the future. When a Treaty, which in other respects is very precise about which institution shall do what, fails to specify who shall act, it means that the identity of the actor is left open. There are therefore good textual grounds for concluding that

96

Working Paper III

the Treaty of Rome contemplated exemptions being granted other than by the Commission. Nothing in Articles 87 to 89 of the EEC Treaty contradicts this conclusion. 3. Subsidiarity It is also relevant that the Treaty has taken formal note of the doctrine of subsidiarity. Article 3b, as amended by the Treaty on European Union, now reads: The Community shall act within the limits of the powers conferred upon it by this Treaty and of the objectives assigned to it therein. In areas which do not fall within its exclusive competence, the Community shall take action, in accordance with the principle of subsidiarity, only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States and can therefore, by reason of the scale or effects of the proposed action, be better achieved by the Community. Any action by the Community shall not go beyond what is necessary to achieve the objectives of this Treaty.

I do not find attractive the notion of subsidiarity as a constitutional principle, but it has been adopted, and it is difficult even for an anti-subsidiarity-ist like myself to deny that it supports the proposition that national decision-making is not a priori inferior to Community decision-making. The local court or competition agency is likely to know the facts more thoroughly than a European agency. So decentralisation is consistent with the Treaty.

4. Do Article 81(1) and Article 81 (3) constitute a single text? A more difficult textual question is whether one can fairly read Articles 81 (1) and 81 (3) together. This is a genuine and challenging legal problem with substantive and procedural consequences. Where only the Commission was in charge of the analysis, it did not make a great difference. If the agreement was basically desirable in its totality, but contained restrictive clauses, the exemption decision would recite a litany of restrictions, and then a counterpoint of beneficial effects. The Commission was regularly encouraged to be bolder in applying Article 85 (1) in such a way as to find that clauses which notionally limited the economic freedom of the parties were nonetheless not anti-competitive in that their dispositions were a natural concomitant of a basically pro-competitive agreement. The ease with which clauses could be found restrictive meant that enterprises were reluctant to pursue formal exemption since obtaining one always involved the risk of being asked to make heavy concessions as the price of approval. This phenomenon muddled the debate about the respective roles of Article 85 (1) and (3). Was the function of Article 85 (3) to cure, by reference to indus-

/ - Compatibility, Efficiency, Legal Security

97

trial policy, the competitive restrictions identified in Article 85 (1)? Or was it to be used only in the gravest cases, on pure competition grounds, where the criteria established by Article 81 (3) for the grant of an exemption had been stringently verified? If we looked at the decisions, however, the debate became irrelevant. The Commission's practice was not completely consistent, but generally if it granted an exemption it noted 'restrictive clauses' plus 'economic benefits', and then granted an exemption. Sometimes, there would be a bolder decision in which the restrictive features were minimised.64 Commission decisions, however, usually proceeded to a classical textual analysis of restrictive clauses, followed by a curative recital of their wholesome economic effects. Functionally, therefore, paragraphs (1) and (3) have been read together in the Commission's practice. This had big implications for the enforcement of the law. The Commission was unable to deliver the exemption decisions for which its theoretical jurisdictional reach created a need.65 Indeed, I submit that the core difficulty is not a textual problem at all but a problem of practicalities, of how the European Commission chose to structure its enforcement efforts. In the early 1960s, when Regulation 17/62 was being drafted, the staff of DG IV had no special reason to be optimistic about the receptivity of businesses, their lawyers or even national judges to novel doctrines, promulgated by a young bureaucracy, still less the use of competition law as a tool for cross-frontier market integration. Naturally, the officials were reluctant to see their creature stifled at birth. So they decided that the basic prohibition should be interpreted as having a broad reach. It would be by the tolerant use of Article 81 (3) rather than by the mild use of Article 81 (1) that sensitive choices were to be made. Producing a volume of decisions commensurate with the jurisdictional reach implicit in a broad interpretation of Article 81 (1) proved to be beyond the capacities of the Commission. The procedures adopted were too cumbersome, too thorough, too perfectionist, too slow. Some early decisions were taken very briskly66 but others took years, even decades.67 The enforcement system effectively broke down; certainly specialist practitioners in Brussels had an 64 {Consortium ECR 900, OJ L 228/31 1990, for example, or ElopaklMetal Box Odin, OJ L 209/15 1990, in which the Commission recited a lot of good reasons why the deal made sense as pro-competitive; dozens of decisions used this kind o f reasoning in the past, but always as justification for granting a n exemption under Article 81 (3), as opposed to negative clearance under Article 81 (1). 65 This approach to paragraph (1) may yet cripple the decentralisation effort, since if the authorities to whom power of enforcement is newly-given follow the example of the European Commission from 1970 to 1999, they will create for themselves the same kind of difficulty as the Commission is trying to escape. Too many deals caught by the prohibition, too many candidates for exemptive treatment, too much procedural manipulation by using unanswered notifications as proof of legality. 66 Decision 72/480/CEE WEA-Filipacchi OJ L 303/52 1972, raced from opening t o adoption in barely seven months. 67 Decision 93/46/EEC Ford Agricultural, O J L 2 0 / 1 1993, is one, maybe unfair, example.

98

Working Paper III

advantage over non-specialists elsewhere. It was easier to prevent a decision being taken than to achieve the taking of a decision. Doubts and theoretical disputes could delay controversial decisions for years or forever. Victory in defending a client accused of an infringement would lie in the absence of defeat; that is to say in the non-taking of a negative decision rather than the taking of a favourable decision. Thus the proposed reforms are largely a reaction to the Commission's inability to enforce the system rationally. When the key documents were adopted over thirty years ago, expectations about how they would be enforced were different.

5. Are the Bosch and Portelange judgments relevant? This is an appropriate moment to note the early judgments of the Court of Justice, where the Court had to consider the effect on 'old agreements', antedating the Treaty of Rome or Regulation 17/62, of the new competition rules. Did a restrictive agreement become void from the moment of its signature, from the moment of the Treaty's entry into force, from 1962, or from the moment of the notification? Or did it remain valid unless condemned? And if condemned, from what moment should the nullity run? The Bosch68 and Portelange69 cases offered an opportunity for the Court to consider how the competition rules ought rationally to be enforced. In Bosch, speaking of 'old agreements', those signed before the entry into force of Regulation 17/62, the Court said: . . . the authors of the Regulation seem to have envisaged also that at the date of its entry into force there would be subsisting agreements to which Article 85 (1) applied but in respect of which decisions under Article 85 (3) had not yet been taken, without such agreements thereby being automatically void. The opposite interpretation would lead to the inadmissible result that some agreements would already have been automatically void for several years without having been so declared by any authority, and even though they might ultimately be validated subsequently with retroactive effect. In general it would be contrary to the general principle of legal certainty—a rule of law to be upheld in the application of the Treaty—to render agreements automatically void before it is even possible to tell which are the agreements to which Article 85 as a whole applies.70

This seems consistent with the view that Article 85 is not a prohibition to which a potential cure is attached, but a conditional prohibition that applies if the cure is unavailable. Thus these paragraphs contradict the 'German view' that the reforms are improperly abandoning a regime of prohibition of infringements and adopting a regime of controlling abuses. They also contradict 68 69 70

Case 13/61 Bosch [1962] E C R 89. Case 10/69 Portelange v Smith Corona [1969] E C R 309. See Bosch at § 52.

1 - Compatibility, Efficiency, Legal Security

99

Commission practice, which was indeed to regard the denial of an exemption as having retroactive effect. The Court thought it would be wrong to prohibit and render void an agreement that might subsequently be validated retroactively. However, the Court then discussed Article 85 (2) which . . . seems to regard Articles 85 (1) and (3) as forming an indivisible whole. The Court was writing in an age when it was expected that the notification system would work rationally, with filing, examination and favourable or unfavourable response following each other in a rational manner and with reasonable speed. In Portelange,11 the Court maybe noted the possibility of a problem of delay. It stated: 15. In view of the absence of any effective legal means enabling the persons concerned to accelerate the adoption of a decision under Article 85 (3)—the consequences of which are all the more serious the longer such a decision is delayed—it would be contrary to the general principle of legal certainty to conclude that, because agreements notified are notfinallyvalid so long as the Commission has made no decision on them under Article 85 (3) of the Treaty, they are not completely efficacious. 16. Although the fact that such agreements are fully valid may possibly give rise to practical disadvantages, the difficulties which might arise from uncertainty in legal relationships based on the agreements notified would be still more harmful. 20. It must therefore be concluded that the agreements mentioned in Article 85 (1) of the Treaty, duly notified in accordance with Regulation No 17, are of full effect so long as the Commission has made no decision under Article 85 (3) and the provisions of the said regulation.

I do not see in this judgment a definitive statement about whether Article 85 (1) and (3) shall be applied in a sequential or in a unitary fashion. Both the Portelange and the Bosch judgments were considering practical procedural problems. They do not prove or disprove the Commission's case: they were rational responses to the procedural situation as the Court could imagine it. They were adopted when all parties could reasonably expect that a notification would be examined promptly and decided with only moderate delay. It would therefore be unreasonable for the denial of an exemption to have retroactive consequences that went back a very long way. As we have learned, however, very few notifications lead to a formal decision and none are decided quickly. One may imagine that, confronted with such a procedural situation, the Court would wish to follow whatever approach was not inconsistent with the words of the Treaty and was most likely to deliver a procedurally rational framework within which business people could organise their affairs with a reasonable level of stability. The Bosch and Portelange judgments are best read as the Court's reasonable expression of opinion as to how the notification system should work. I do not see them as making fundamental pronouncements relevant for all time in different procedural contexts. In any event, we should not feel

71

Portelange at §§ 15, 16 and 20.

100

Working Paper III

constrained by very early judgments when.the Court was feeling its way (remember that the Court made early mistakes. Sirena v Eda12 is probably the clearest).

6. Self-help and the rule of reason Although the notion that ancillary restraints could be pro-competitive was helpful, and although the European Court was notably encouraging when it had a chance to bless a more robust approach to Article 85 (1), the Commission was reluctant to contemplate a reading of Article 85 (1) which allowed businesses on their own to conclude that a deal with restrictive clauses and procompetitive effects was textually tolerable. My usual analogy for students is to say that DGIV was a catholic jurisdiction in which the affirmative approval of the priest was necessary: the layman was forbidden to discern the right path by consulting his conscience alone. He was eligible to receive blessing only after having made a full confession. To quote the White Paper,73 The Commission has already adopted this approach to a limited extent and has carried out an assessment of the pro- and anti-competitive aspects of some restrictive practices under Article 85 (1). This approach has been endorsed by the Court of Justice.47 However, the structure of Article 85 is such as to prevent greater use being made of this approach: if more systematic use were made under Article 85 (1) of an analysis of the pro-and anti-competitive aspects of a restrictive agreement, Article 85 (3) would be cast aside, whereas any such change could be made only through revision of the Treaty. It would at the very least be paradoxical to cast aside Article 85 (3) when that provision in fact contains all the elements of a 'rule of reason'. (Footnote47 referring to the Maize Seed1* and Pronuptiali cases omitted.) The Commission warned against . . . diverting Article 85 (3) from its purpose, which is to provide a legal framework for the economic assessment of restrictive practices and not to allow application of the competition rules to be set aside because of political considerations.76

72

C a s e 40/70 Sirena Sri v Eda Sri and Others [1971] E C R 69. E u r o p e a n C o m m i s s i o n : White Paper on the modernisation of the rules implementing Articles 85 and 86 of the EC Treaty, C O M (1999) 101 Final (Apr. 1999), § 57; 1999 O J C 132/1 [hereinafter White Paper]. 74 Case 258/78 L. C Nungesser KG and Kurt Eisele v Commission of the European Communities [1982] E C R 2015. 75 Case 161/84 Pronuplia de Paris Gmbh v Pronuptia de Paris Irmgard Schillgalis [1986] E C R 353 76 Mat§57. 73

1 - Compatibility, Efficiency, Legal Security

101

The Commission had already started changing its mind (but was unwilling to voice doubts in public) when the intriguing judgment in Night Services11 emerged. There, the Court of First Instance offered its version of the point at which Article 85 (3) should become relevant: . . . The Commission submits that, whilst the analysis of an agreement must take account of its economic context, it does not follow that the rule of reason—a concept which the Court of Justice has hitherto declined to embrace—should be resorted to . . . Consequently, it is necessary to balance the competitive benefits and harms of an agreement in relation to the granting of exemptions under Article 85 (3) of the Treaty but not in respect of the appraisal of restrictions on competition—which were, contrary to the United Kingdom's contention, fully explained in the decision—in accordance with Article 85 (I).78 Before any examination of the parties' arguments as to whether the Commission's analysis as regards restrictions of competition was correct, it must be borne in mind that in assessing an agreement under Article 85 (1) of the Treaty, account should be taken of the actual conditions in which it functions, in particular the economic context in which the undertakings operate, the products or services covered by the agreement and the actual structure of the market concerned (judgments in Delimitis, cited above, Gottrup-Klim, cited above, paragraph 31, Case C-399/93 Oude Luttikhuis and Others v Verenigde Cooperatieve Melkindustries [1995] ECR 1-4515, paragraph 10, and Case T-77/94 VGB and Others v Commission [1997] ECR 11-759, paragraph 140), unless it is an agreement containing obvious restrictions of competition such as price-fixing, market-sharing or the control of outlets (Case T-l 48/89 Trefilunion v Commission [1995] ECR 11-1063, paragraph 109). In the latter case, such restrictions may be weighed against their claimed pro-competitive effects only in the context of Article 85 (3) of the Treaty, with a view to granting an exemption from the prohibition in Article 85 (I).19 (emphasis added) The route chosen by the Commission in the White Paper is faithful to its traditional doctrines: The whole of Article 85 would then become a directly applicable provision which individuals could invoke in court or before any authority empowered to deal with such matters. This interpretation would have the effect of making restrictive practices which are prohibited by Article 85 (1), but which meet the tests of Article 85 (3) lawful as from the time they were concluded, without the need for any prior decision. Similarly, restrictive practices that restricted competition would be unlawful once the conditions of Article 85 (3) are no longer fulfilled. This new framework would mean that restrictive practices would no longer have to be notified in order to be validated. The arrangements for implementing Article 85 as a.whole would then be identical to those for Article 85 (1) and Article 86.80

77 Joined Cases T-374/94, T-375/94, T - 3 8 4 / 9 4 a n d T-388/94 European Services and others v Commission [1998] E C R 11-3141. 78 Id. at §134. 79 Id. at §136 80 White Paper, supra note 73 at § 69.

Night

102

Working Paper III

It is difficult to predict whether the analysis of a directly-effective Article 85(1) and (3) would follow the liberal approach espoused in Gottrup Klim81 and European Night Services,62 or the old approach in the joint venture cases like Alcatel, quoted above,83 or X/Open.84 As noted elsewhere, I suggest that in jurisdictions where competition doctrines are new, or where the European Commission's policy serves as the role model for the local enforcement agency, there is a risk that it will be by the application of the exempted part of the equation that most sensitive decisions are made. In conclusion, I do not think, however, that there is a strong textual case against reading Article 81(1) and Article 81 (3) together as a unity.

IV. Practical considerations 1. Legal certainty Legal certainty or legal security (the terms seem synonymous) is very frequently invoked as a prime concern for those responsible for enforcing the competition rules. Advocates General, article writers, and the Commission itself have each stated on various occasions how important it was to ensure legal certainty. The absence of such certainty would typically be due to the unavailability of a formal Commission exemption decision, or due to the inherent uncertainty of a competitive situation in which a formal clearance or condemnation had not yet emerged. I have never been wholly convinced that an absence of legal certainty was such a terrible thing; or indeed that the achievement of legal certainty about the application of the competition rules was in truth feasible. Competition law exists to encourage competition, not to prescribe or regulate rigidly or definitively. It is a plastic legal discipline. Its goal is to ensure the opportunity for economic operators to pursue prosperity in the marketplace and to engage in economic strife with each other. It is not realistic to imagine that a business manager can expect a high degree of legal certainty as to his 81 Case C-250/92 Gettrup Klim Grovvareforeninger v Dansk Landbrugs Grovvareselskab AmbA [1994] E C R 1-5641. 82 Joined cases T-374/94, T-375/94, T-384/94 and T-388/94, [1998] E C R 11-3141. 83 See supra note 53. 84 For a very conservative approach to Article 85 (1) see XlOpen Group, O J L 35/36 1987, where the Commission found that the information exchange and related non-disclosure agreements did not infringe Article 85 (1). On the other hand, the combination of the restrictions on membership and the fact that, even though the resulting standards were to be open and public, members would have access to them sooner than non-members, were caught by Article 85 (1) but exempted. One may wonder, as ever, whether this methodological contortion was really useful in order t o reach the conclusion that the whole arrangement was valid under the competition rules (the agreement was operated for almost a year before being notified).

1 - Compatibility, Efficiency, Legal Security

103

competitive situation. There are certain matters in running a business where legal certainty is expected and indeed demanded. The formal requirements for the execution of a binding contract of employment must be known, and if they are not known or are constantly changing, business people can properly feel let down by legislators and judges. The terms of the law governing the lease of office premises or the effectiveness of a bank loan must be clear, and their application ought to be predictable. In private life, the rules governing the formal validity of a will and the constraints upon a testator's freedom to dispose of his assets as he chooses, are precisely defined by the law, and prudent people organise their affairs accordingly. By contrast, any business is today exposed to swings in the respective values of foreign currencies, changes in how business is conducted due to the Internet and e-commerce, continually growing social costs, the emergence of new products, and the gradual de-regulation of many markets which, in former years, were heavily regulated. The uncertainties posed for a business by these external political, fiscal, monetary, economic and social factors, to say nothing of changes in customer taste and the perpetually unforeseeable course of events, are far more significant and arguably far more threatening than the doubts attributable to the unpredictability of the outcome of any competition law dispute. As long ago as the 1970s, prudently run companies were looking at the possibility of filing a notification with the European Commission, and concluding that this was one invitation they would not take up. Such companies would say to themselves that while it would be agreeable to have the formal confirmation of the European Commission that the arrangement to which they were party was valid and enforceable, there were countervailing drawbacks. The filing of a notification could draw attention to a transaction that might otherwise go unnoticed. More particularly, obtaining the Commission's approval could involve months and years of negotiation and concession. In most cases, the Commission's approval was not manifested by any formal decision. The heart of the doctrine concerning whether a notification was desirable was therefore based on a fiction, namely that most notifications led to a formal disposition by the Commission's services. It can thus be observed that many companies have decided over the years that they will not exercise their option to file a notification requesting an exemption. They accept some diminution in the legal certainty governing their affairs, but do not regard this as troublesome.

2. Fictitious advantages of the reforms At least one rationale for the change seems disingenuous. For example, it is contended that ... undertakings' legal certainty will remain at a globally satisfactory level, and in certain respects will even be strengthened. Thus, instead of depending on the Commission adopting an exemption decision, undertakings will be able to obtain

104

Working Paper HI

immediate execution of their contracts before national courts, with effect from the date of their conclusion, provided that the conditions of Article 85 (3) are satisfied. There is no presumption that restrictive practices are void under Article 85: the prohibition contained in this provision is applicable only when the conditions of prohibition are met.85

I call this fiction because companies did not find themselves having to wait for an exemption before being able to enforce a contract. In terms of legal certainty as that is conceived by the Commission, companies will not be better off. They will execute agreements containing clauses with restrictive features. If commercially convenient, they will endeavour to avoid complying with the contract on the grounds of competition law. There will be complaints or requests for the Commission's intervention to cure an uncertainty. The Commission's comments, whether framed as a formal decision (rare in the past and rare in the future) or in a simple administrative letter, will be given very great weight. While comfort letters did not formally solve the theoretical problem that only an exemption could cure a situation where there was a restrictive clause in a benign relationship, in actual practice I have never once encountered a judge who took the formal position that the agreement was necessarily damned since the Commission had not issued its formal blessing.

3. The supposed dangers of change Supporters of the German view contend that by eliminating the automatic prohibition of paragraph (1), there is a danger of a loosening of the protections against anti-competitive agreements. On this basis, since the prohibition only bites once the grant of an exemption is refused, doubtful agreements and even plainly unacceptable agreements, will effectively remain provisionally valid until they are denied an exemption; by contrast, at the moment doubtful agreements fall under the prohibition of Article 85 (2) regardless of whether they might be said to be eligible for an exemption. Thus the victim of an illegal agreement might be forced to respect it, being unable to invoke its nullity in the absence of an affirmative decision in its favour, whereas today he could invoke nullity without needing to discuss whether an exemption might be available. This is a substantial shift in the legal architecture of the competition rules. However, in my submission it will not make fundamental changes in practical outcome. At the moment, if Article 81 (1) is invoked before a national court, and if the court finds the presence of clauses deemed restrictive by the European Commission, in theory the court should either refrain from acting until the Commission has decided whether an exemption will be available, or should apply Article 81 (2). In actual practice, however, most judges are inclined to decide a case by enforcing or not enforcing the terms of an agree85

White Paper, supra note 73 at § 78.

1 - Compatibility, Efficiency, Legal Security

105

ment rather than by doing an incomplete job. Thus in actual practice the courts are today not greatly hindered by the Commission's monopoly over the granting of exemptions; and I presume that judges who feel comfortable with competition law matters will welcome the new power to grant exemptions. Conversely, unlawful agreements upon which one party seeks to rely are unlikely to be enforced, regardless of the fact that in strict theory they would be enforceable for so long as an exemption has not been refused. In the same way as agreements in the past were prosecuted and threatened with fines, not because of their not having been notified but because of their contents and effects, so agreements in the future will be threatened with nullity and fines not because they have not been granted an exemption, but because of their contents and effects. Functionally, Article 81 (1) and (3) are today read together. Functionally, they will be read together under the new regime. Neither outcome appears more or less likely to lead to genuine injustice.

4. Enforcement before thirty entities Europe is not a judicially uniform continent. In certain countries, judges are well paid and honoured, the court system receives high amounts of public funding, the orderly handling of litigation is a high priority of the central government, and litigants and lawyers involved in the process are correspondingly fortunate. In other countries, judges are poorly paid, back-logs of judicial business extend to years, procedures are formalistic, slow and unpredictable. Litigation in any country is to be approached only reluctantly, but in the latter countries it is quite unlikely to yield timely satisfaction (I refrain from using the word 'justice'). Three contiguous extremes may be noted. Litigation before the High Court in England is luxury justice, costly and risky, but efficient and relatively rapid. In the Netherlands, kortgedinge procedures are in effect interim interlocutory hearings that can be heard within a few weeks of the initiation of litigation. Very often the first hearing, although notionally merely preliminary, constitutes the entirety of the proceedings. Those who want a quick answer can obtain one in a brisk and common sense atmosphere. In Belgium, the situation is very different. The judicial backlog is immense and it is not uncommon for ten years to elapse between the initiation of the litigation and its disposition upon appeal. I lack personal experience, but it is commonly stated that litigation in Scandinavian countries is relatively rapid and efficient, whereas litigation before the Italian civil courts can be painfully slow. There are explanations for why the courts are so organised in each of the countries, and there is no early prospect of approximation of the different systems. It must also be acknowledged that there are huge differences in the experience, skill and resources of national competition agencies. Austria and Belgium are two countries where competition law enforcement is in serious difficulties. Although the need for an independent Kartellamt in Austria has been

106

Working Paper HI

recognised for years, and although a number of studies have been commissioned, and although there is cross-party agreement that the application of competition law needs to be improved, there appears little likelihood of change in the near future. The Ministry of Economics has the job of acting as competition authority, but it seems widely accepted that this is not successful. No competition agency has the right to apply Articles 81 and 82 of the EC Treaty. The Belgian Conseil de la Concurrence issued its second report with respect to the period 1994 to 1995. It has never issued a further report. In its second report, it stressed, as it had done in itsfirstreport, that due to staff and equipment shortages, it was able only partly to fulfil its duties under the Law of August 5, 1991, on the protection of economic competition. The reports acknowledge that the Conseil de la Concurrence has, on a number of occasions, been unable to deploy officials to examine a merger within the statutory deadline. The Conseil de la Concurrence lacks books on competition law, legal periodicals, economic periodicals, filing cabinets, computers, and even enough officials who have fluency in Dutch and French. As protest about these inadequacies, the seven members of the Conseil de la Concurrence resigned. I may also observe that the decisions taken do not always carry complete conviction. In one case of which I have heard, the Conseil approved a joint venture between fruit and vegetable traders on the grounds that the parent companies and the joint venture would be active in different markets, the one being active in the long-term market for the sale of fruit and vegetables, the parent companies being active in daily operations. It was alleged that there were considerable spill-over effects between the two markets, but the joint venture was cleared nonetheless. Other countries have very effective competition agencies. The Italian Autorita Garante della Concorrenza e del Mercato has had a tremendous impact in a short time led by eminent figures, and is well-funded and enjoying strong popular support. Likewise in France, the Conseil de la Concurrence and the DGCCRF at the Ministry of Economic Affairs are widely regarded as knowledgeable, vigorous and effective. The German and Dutch agencies are similarly well-regarded, and in due course the UK, which has lacked an effective system of dealing with small cartels, will probably be endowed with a good regime. A report on each Member State's competition agency is beyond the scope of this paper. The point remains that there are substantial divergences. It is sometimes argued that national judges cannot be trusted to do a good job of working with the unfamihar economic doctrines involved in applying the competition rules. If judges are trusted to handle subjects like medical negligence, taxation, white-collar crime, and the calculation of damages in a vast range of circumstances, it seems unreasonable and elitist to deem them incapable of applying the competition rules. That said, the judges of certain countries are more familiar with competition law arguments than others, and this will hinder the invocation of EC competition law claims or defences before certain courts. Thus, Belgian, English and Dutch judges, to name only three, are unlikely to be nervous of confronting European competition law questions, whereas Portuguese and Greek judges might be.

1 - Compatibility, Efficiency, Legal Security

107

Decentralising the government of a country can mean transferring authority from the central government to regional or local administrations. The roles and powers of the administrations to receive the new authority will be well known. If, tomorrow, the job of selecting the schoolbooks for primary pupils were to be effected by East Lothian District Council as opposed to a Scottish or UK Ministry, there is no doubt that the Council will know its task and be equipped to discharge it. By contrast, decentralising from the Directorate General for Competition to fifteen national authorities and more than fifteen judicial systems (regional judicial systems such as Scotland) represents quite a leap in the dark. Some agencies and courts are efficient, other are not. If the decentralisation were to work badly, substantial prejudice could occur for businesses located in Member States where enforcement is inefficient. The natural response will be forum shopping, which could take the form of pre-emptive action by initiating litigation in a slow jurisdiction, or by seeking to engage the attention of an efficient agency in a matter that is not primarily within its jurisdiction. There will be many matters where more than one court jurisdiction or enforcement agency could be involved. Indeed, such is the trend to globalisation that it would not be surprising for three or four countries to have an interest in a particular controversy. The existing mechanisms for cross-border enforcement cooperation appear to be non-existent or rudimentary in many cases. While the goal of sharing enforcement responsibility with the Member States is plainly desirable, the European Commission will have an immense responsibility to coordinate, educate and mediate. It should also be ready to decide, in cases where this is appropriate, and I submit that the White Paper seems to contemplate something close to abstention in a wide range of situations. To the contrary, I would submit that since the Commission is the premier competition law agency in Europe, with excellent prestige, resources and experience, those resources ought to be available for deciding competition cases. To exaggerate the proposition, it would be disappointing if the Commission were to become merely a spectator of imperfect national enforcement of the competition rules that have been largely shaped by Commission enforcement action. Pursuing this theme, the Commission must have an enforcement role more extensive than major cartels, legislation, major mergers, and coordination. Today, businesses with a competition problem can obtain guidance in Brussels: one-stop-shopping. The Commission's proposals go too far, I submit, in assuming that the move from centralised to decentralised enforcement will necessarily be a good thing. In many cases, especially at the beginning, substantial confusion seems likely. More precisely, there is a risk of wrong decisions, unjust decisions, or no decisions, and a step backwards from the level of enforcement, albeit imperfect, that we have today. Welcoming the proposal to share responsibility with national entities having resources beyond those of the Commission is not inconsistent with expressing alarm about the short- and medium-term problems which will probably arise if the Commission does not heavily and actively involve itself in the process of absorption into the Member States. The

108

Working Paper III

Commission is traditionally very cautious about taking a position in cases pending before a national court. Because so much weight is attached to its opinion, what it says may effectively decide the competition point in dispute rather than merely giving guidance. The Commission therefore expresses itself very cautiously and the formulation of its replies to a request for assistance from a national court will normally take months. Possibly because of the Commission's reluctance to take a false step in influencing the outcome of private litigation, its formal assistance has been requested on only a few occasions. Helping the process of decentralisation will involve giving help to national courts more quickly and less cautiously than has been the case in the past.

5. Notifications The Commission dislikes notifications because they clog its in-box. This seems surprising, at least by reference to the volume of notifications recorded in the Annual Reports on Competition Policy. In 1998, there were 216 notifications; in 1999, there were 162 notifications. These numbers do not seem to be an excessive burden for a skilled administration with long experience and immense inhouse knowledge. Moreover, a fair number of those must relate to transactions that are now eligible for a block exemption pursuant to the new doctrines on vertical restraints. In addition, a number are presumably principally national rather than European in scope. I therefore question whether the disadvantages to the Commission of notification are so serious as to outweigh the advantages of giving it thoroughly-documented, carefully-drafted descriptions of why particular transactions are to be regarded, in their entirety, as pro-competitive. When the private sector volunteers sensitive and potentially embarrassing information to a public authority, the public authority is better informed about market developments and techniques, learns who the big players are, understands where the controversies may lie, is better equipped to do its job, and is invited to guide major enterprises on how to respect the competition rules. These are important benefits. The notification system is capable of being abused. First, there is the notion of the so-called dilatory notification. The Commission's Notice on Cooperation with National Authorities86 defines dilatory notification as o n e where a firm, threatened with a decision banning a restrictive practice which a national authority is posed to take under Article 85 (1) or under national law, notifies the disputed agreement to the Commission a n d asks for it to be exempted under Article 85 (3). 8 7

86

E u r o p e a n Commission: Notice on Cooperation between National Courts and the Commission in applying Articles 85 and 86 of the EEC Treaty, O J C 39/6 1993. 87 W.at§55.

1 - Compatibility, Efficiency, Legal Security

109

Another category of potential abuse is what I call notification to achieve the higher moral ground, as follows. A firm enters an agreement. The agreement could be said to be caught by Article 81(1) and there are doubts about whether the other contracting party might wish, in the future, for commercial reasons, to try to wriggle out of the contract, invoking the competition rules. Filing a notification offers the notifying party the opportunity to claim the higher moral ground; the Commission has been informed about the deal and has not challenged it. As a result, the transaction is presumptively valid under the competition rules. The vice in this situation is not the requesting of the Commission's approval for a transaction, but the possibility that silence on the part of the Commission may be misrepresented as proving the agreement's acceptability. The problem is curable by deadlines. If the Commission were obliged to respond to every notification within, let us say, forty-five days, the problem would be eliminated. There is a market for authoritative guidance on competition law matters. In the context of the Merger Regulation,88 the Commission is arguably the world leader in delivering such guidance. In the context of Article 81 (1), it could do better if it were less perfectionist and worked to brisk deadlines. While I accept that the Commission is not slamming the door in the face of those who wish to seek its guidance, I am not convinced that abandoning the notification entirely would be an appropriate response to the problem of supposed administrative overload through having had to examine less than two hundred filings.89

6. Application in countries candidate for accession and other third countries Obligations to adopt and apply competition rules are imposed by the Europe Association Agreements of 1993 and 1994 upon Poland, Hungary, the Czech Republic, the Slovak Republic, Romania, Bulgaria, Slovenia; by Europe Association Agreements of 1998 upon Lithuania, Estonia and Latvia; by Partnership and Cooperation Agreements of 1995 upon Russia, Belarus, Ukraine, Moldova; and by Mediterranean Agreements of 1995, 1996, 1997 and 1998 upon Tunisia, Morocco, Jordan, the Palestinian Authority and Israel. Like obligations are imposed upon Turkey, and upon Norway, Liechtenstein and Switzerland. The commonest method of implementation of these Treaty obligations has been for the third country to reproduce more or less faithfully the language of Articles 81 and 82 of the EC Treaty. My anecdotal experience suggests that enforcement agencies in Eastern Europe are most comfortable 88

Council Regulation 4064/89 on the control of concentrations between undertakings, OJ L 257/14 1990. 89 If resources are n o t available, why n o t levy a charge of, say, €10,000 p e r notification?

110

Working Paper III

when applying a black-letter rule as set forth in a block exemption regulation, which they are likely to examine as if it was prescriptive and obligatory. Yet, the presence or absence of key clauses should not deliver certainty. Competition rules should move away from clause-driven unenforceability, for example whether know-how, which is the subject of an exclusive licence, should be recorded in writing in order for the licence to be exempted. These difficulties are well known to competition law specialists; and the Commission's theorists are correcting them. However, I question how the new enforcement jurisdictions will deal with their new responsibilities. Advising lawyers in these countries, practising before these administrations, certainly suggests that the competition law that is being exported from Europe to the East is the old law and not the new law. Again, the problem is capable of being solved, but it calls for the energetic involvement of the European Commission. We are talking about competition, not aesthetics. Our criticisms of the new regime should be rooted in practicalities. Will the new system work better than the old one? A system which may contain paradoxes and inconsistencies but which functions well is to be preferred over a perfectly conceived system that does not function. I therefore commend the boldness of the Commission. It observed that the system is not functioning. It has adopted a path that will yield many more decisions applying the competition rules, some of which will be good and some of which will be bad. In thisricherjurisprudence, there will be imperfections and injustices, but by the end of this decade, we should see the wholesome effects of the entering into daily business life of competition rules at national and international level. The difficulties of getting to that happy condition have been underestimated in the White Paper.

/ - Compatibility, Efficiency, Legal Security

111

ANNEX 1: COUNCIL ARCHIVES—SUMMARY OF TOPICS DISCRIMINATION* Projet allemand

Projet fran; ais

Art. 42 Est incompatible avec le marche commun et interdit le fait pour une entreprise de desavantager, dans des transactions comparables 1 , des acheteurs ou vendeurs en concurrence entre eux, en raison de leur nationality.

Art. X(Doc. Mar. Com. 37) Sont interdites a l'interieur du marche commun les majorations ou minorations de prix et les modifications des conditions de vente operees pour des transactions comparables 1 , a l'egard d'acheteurs ou de vendeurs en concurrence entre eux.

* These charts are based on a document found at the Council Archives entitled: Tableau synoptique des projets d'articles soumis par les experts des delegations concernant les regies de concurrence applicables aux entreprises, MAE 325 f/56, CM3/NEGO/236 (only the formatting has been changed). 1 Cette expression doit etre interpretee comme visant une inegalite dans les conditions faites a l'acheteur, qui depasse l'inegalite que justifie la difference de nature des transactions comparees.

Projet belgo/neerlandais

Sont incompatibles avec le Marche Commun et a interdire, dans la mesure oil le commerce entre Etats membres s'en trouve affecte, a) l'exploitation abusive, par les entreprises publiques ou privees d'une position dominante sur le marche; b) les accords entre entreprises dont Papplication constitue une entrave injustifiee a l'exercice de la concurrence. Ces pratiques abusives peuvent resulter notamment: —de lafixationde prix ou conditions de transaction; —de discriminations a l'egard des

Projet neerlandais

Sont interdites a l'interieur du marche commun les majorations ou minorations de prix et les modifications des conditions de vente operees pour des transactions comparables,' a l'egard d'acheteurs ou de vendeurs en concurrence entre eux, lorsque ces pratiques sont operees: a) en application de reglementations

Projet francais

Sont incompatibles avec le marche commun toutes les situations ou pratiques d'ententes ou de monopole ayant pour objet ou pouvant avoir pour effet d'entraver l'exercice de la concurrence, en particulier: —en fixant les prix; —en restreignant ou en controlant la production, la vente, le developpement technique et les investissements; —en repartissant les marches, produits, clients et sources d'approvisionnement; —en permettant l'absorption ou la domination du marche d'un produit par une entreprise ou un groupe

Projet allemand

Sont incompatibles avec le marche commun et interdits, dans la mesure oil le commerce entre Etats membres s'en trouve affecte, tous accords entre entreprises ayant pour objet ou pour consequence d'entraver la concurrence, notamment —en fixant les prix; —en restreignant ou en controlant la production, la vente, le developpement technique et les investissements; —en repartissant les marches, produits, clients et sources d'approvisionnement.

ENTENTES*

I

publiques;2 b) pour obtenir une position dominante sur le marche; c) pour des motifs autres que commerciaux. acheteurs ou des vendeurs en concurrence entre eux dans la conclusion des contrats; —
1 Cette expression doit etre interpretee comme visant une inegalite dans les conditions faites a l'acheteur, qui depasse l'inegalite que justifie la difference de nature des transactions comparees. 2 La question se pose de savoir si cette regie ne devrait pas etre completee ou remplacee par une disposition semblable visant les Etats membres.

Toutefois, lorsque la demande en d'entreprises; est faite, les accords ou decisions —en subordonnant la conclusion de de ce genre peuvent etre contrats a l'acceptation, par le autorises si le requerant est en partenaire, de marchandises ou de mesure de prouver qu'ils services qui, en raison de leur nature contribuent efTectivement a ou selon les usages commerciaux ne ameliorer la production ou la peuvent pas etre lies. vente, ou a promouvoir le Peuvent etre relevees des dispositions progres technique et precedentes, les situations ou pratiques economique. dont les auteurs sont en mesure de justifier qu'elles contribuent Le requerant devra egalement prouver que les consommateurs effectivement a l'amelioration de la production ou des debouches, ou au beneficieront raisonnablement developpement du progres technique des avantages qui resultent et economique. Us devront egalement de semblables mesures. justifier que le consommateur obtient une part legitime dans le profit qui decoule de ces mesures.

I

2

!

114

Working Paper III

DIVERGENCES—ENTENTES Le texte allemand est le plus restrictif: il prohibe seulement tout 'desavantage' fonde sur un motif de nationality. Le texte frangais est le plus large, en effet: a) il prohibe toutes majorations ou minorations de prix et toute modification des conditions de vente; b) il s'applique non seulement dans les relations entre Etats, mais aussi aux transactions qui ne depassent pas le cadre interne des Etats. Le texte neerlandais est intermediate: il limite l'interdiction generate prevue dans le texte francais aux trois cas a), b) et c). Les projets francais et belgo-neerlandais traitent sur le meme pied les ententes et les monopoles. Toutefois, le projet francais les interdit tous deux en principe, alors que le texte belgo-neerlandais les soumet au controle des abus. Le projet allemand distingue le cas des ententes, de celui des monopoles; il edicte l'interdiction des premieres et soumet les seconds au controle de l'abus. Les textes allemand et belgo-neerlandais limitent l'application des normes aux cas dans lesquels le 'commerce entre Etats membres s'en trouve affecte.' Dans le projet allemand, les ententes peuvent etre 'autorisees' dans certaines conditions; les ententes autorisees sont cependant soumises au controle de Tabus en application des dispositions inscrites sous b). Dans le projet belgo-neerlandais, ententes et monopoles sont soumis au controle des abus en application des dispositions inscrites sous a).

1 - Compatibility, Efficiency, Legal Security

115

MONOPOLES Projet allemand

Dans la mesure ou le commerce entre Etats membres s'en trouve affecte, est incompatible avec le marche commun et doit etre interdit, le fait pour des entreprises publiques ou privees qui ne doivent affronter aucune concurrence, ou du moins aucune concurrence serieuse, sur le marche commun (entreprises jouissant d'une position dominante sur le marche) d'exploiter abusivement leur position, notamment: —en demandant ou en proposant des prix ou enfixantdes conditions de transaction; —en exercant une discrimination injustifiee a l'egard des acheteurs ou des vendeurs dans la conclusion des contrats; —en limitant la production ou la vente ou en freinant le progres technique ou economique; —en subordonnant la conclusion de contrats a l'acceptation, par le partenaire, de marchandises ou de services qui, en raison de leur nature ou selon les usages commerciaux, ne peuvent pas etre lies.

A l'expiration d'une periode de deux ans a partir de la mise en application du present Traite, chacun des Etats membres devra avoir pris les dispositions legislatives ou reglementaires necessaires pour assurer la mise en oeuvre effective des precedentes dispositions, en ce qui concerne le commerce interieur de chaque Etat et le commerce entre Etats. La Commission sera consultee en vue d'assurer l'harmonisation de ces dispositions avant leur mise en vigueur par les Etats membres. Elle assure entre les Etats membres un echange de renseignements permanent.

Les Etats membres peuvent, avant l'expiration d'un delai de deux ans a partir de l'entree en vigueur du present Traite, conclure un accord additionnel contenant des dispositions en vue de la mise en oeuvre des principes enonces aux articles 42 a 42b et du reglement de la procedure. Si, a l'expiration des trois annees qui suivront l'entree en vigueur du present Traite aucun accord additionnel de ce genre n'a pris eflTet, la Commission edictera meme un reglement general cution comportant des dispositions de cette sorte, ce devant etre approuve Ministres statuant a la majorite qualifiee.1-2 La Commission prend des reglements generaux d'execution soumis au vote de PAssemblee pour la mise en oeuvre des principes enonces ci-dessus et fixe les mecanismes necessaires a ceteffet.1

Projet neerlandais

Commission prend des d'exereglements generaux qui sont soumis au vote dereglement l'Assemblee.par le Conseil de

Trois systemes sont en presence: a) le systeme frangais: chaque Etat prend les dispositions necessaires; la Commission coordonne; b) le systeme allemand: conclusion d'un accord additionel entre Etats membres; a defaut d'un tel accord dans une periode de 3 ans, la Commission edicte un reglement general, avec approbation du Conseil statuant a majorite qualifiee; c) le systeme neerlandais: laelle-

Divergences

La delegation neerlandaise a indique qu'elle prefere le systeme prevu par le Rapport, mais que, a defaut, elle pourrait se rallier au projet allemand sous reserve de Pintervention de PAssemblee. 2 II importe de preciser quand les reglementations d'application entreront en vigueur, car la procedure qui y sera etablie determinera l'application des dispositions de fond.

1

Projet francais

Projet allemand

ETABLISSEMENT DE REGLEMENTATIONS D'APPLICATION DES NORMES DU TRAITE

55

I"

i

Projet francais

La Commission cree une a) En cas d'application abusive ou de instance de recours qui non application des dispositions examine, soit d'office, soit a prevues au A. ci-dessus apres le la demande d'un Etat delai de deux ans, l'Etat dont un membre ou d'une entreprise, ressortissant est lese peut prendre les cas de pretendue provisoirement les mesures violation des principes de reglementaires necessaires et saisir concurrence loyale, tels de l'affaire la Commission. qu'ils sont enonces aux Lorsque les situations ou pratiques articles 42 a 42b. visees au B. ci-dessus affectent le L'instance de recours commerce entre deux ou plusieurs propose aux parties une Etats membres, chacun d'eux peut solution a l'amiable. Si les saisir la Commission pour parties n'ont pas pu violation de ses dispositions. La s'entendre dans un delai de Commission peut egalement se trois mois a compter de la saisir d'office. notification de la solution b) La Commission assure l'instruction proposee, la Commission ou des affaires en liaison avec les Etats un Etat membre pourra interesses et propose a ceux-ci les saisir la Cour de Justice. mesures entrant dans le cadre de leur pouvoir propre et qui sont de nature a porter remede aux situations ou pratiques incriminees. La Commission peut egalement prendre les mesures relevant de sa competence propre.

Projet allemand

La Commission, soit d'office, soit a la requete d'un Etat ou d'une entreprise, instruit les cas d'infractions aux reglements generaux. Elle propose aux parties les mesures propres a ecarter l'infraction constatee. Si, a l'expiration d'un delai de trois mois, elle constate que l'entreprise a laquelle l'infraction est reprochee, ne se conforme pas aux reglements generaux, elle applique les mesures suivantes: —dans le cas d'une entente, elle procede a l'annulation de l'entente; —dans le cas d'un monopole, ellefixeles prix et conditions de vente a appliquer par l'entreprise en cause, ou etablit des programmes de

Projet neerlandais

DECISIONS ET RECOURS

Systeme allemand: 1. creation par la Commission d'un instance, qui examine soit d'office, soit sur demande, les cas de pretendue violation; 2. l'instance propose aux parties une solution amiable; 3. a defaut d'acceptation de cette solution, recours a la Cour de Justice. Systeme francais: 1. il appartient aux seuls Etats de prendre les mesures requises et/ou de saisir la Commission; 2. la Commission propose aux Etats les mesures relevant de leur competence et prend les mesures qui relevent de sa competence propre; 3. en cas de non-application, la Commission saisit le Conseil, qui decide a la majorite qualifiee; 4. recours devant la Cour Systeme neerlandais: 1. la Commission instruit les cas d'infraction;

Divergences

5

I

Projet allemand

c) En cas de non application par les Etats interesses des mesures appropriees dans un delai de trois mois oil en cas d'application abusive des mesures prises, la Commission saisit le Conseil des Ministres, qui prend a la majorite qualifiee une decision immediatement executee. d) Un recours devant la Cour de Justice est ouvert aux Etats interesses en cas de violation des dispositions du Traite.

Projet francais fabrication ou des programmes de livraison a executer par elle.

Projet neerlandais

2. elle propose aux parties les mesures appropriees; 3. si l'infraction n'est pas ecartee dans un delai donne, elle applique ellememe des mesures correctives.

Divergences

I

I'

1 - Compatibility, Efficiency, Legal Security

119

ANNEX 2: DOCUMENTS FOUND AT THE COUNCIL ARCHIVES Date

Dossier N°

Document N°

CM3/NEGO/001

Introduction en 12 pages tragant l'historique des negotiations Propositions de redaction d'articles MAE 139 f/56 concernant les regies de concurrence et les distorsions MAE 148 f/56 Projet d'articles concernant les regies de concurrence MAE 252 f/56 Extrait du proces-verbal des partiel + Annex reunions des 03-05.09.1956 du groupe du marche commun MAE 233 f/56 Proposition de la delegation franchise concernant les discriminations en matiere de prix et les pratiques restrictives de la concurrence MAE/sec 29/56 Memento interne du secretariat Projet d'articles remis par M. Tiesing, 10.09.1956 MAE 325/56 Tableau synoptique des projets d'articles concernant les regies de concurrence applicables aux entreprises Proposition de base pour la deuxMAE 468 f/56 ieme lecture par le groupe du marche commun en ce qui concerne les regies relatives a la discrimination entre ententes et monopoles Texte relatif aux regies de concurrence adoptees par le groupe en seance restreinte Proposition neerlandaise amendee au cours de la reunion restreinte du 6novembre 1956 Proposition neerlandaise MAE 527 f/56 Projet de redaction sur les regies de concurrence MAE 541 f/56 Proposition soumise par le President du groupe du marche commun

27/5/1956

CM3/NEGO/236

6/7/1956

CM3/NEGO/236

3-5/9/1956 CM3/NEGO/236

4/9/1956

CM3/NEGO/236 (annexed to Doc. MAE 252/56)

7/9/1956 10/9/1956

CM3/NEGO/236 CM3/NEGO/236

24/9/1956

CM3/NEGO/236

26/10/1956 CM3/NEGO/217

6/11/1956

CM3/NEGO/236

6/11/1956

CM3/NEGO/236

7/11/1956 CM3/NEGO/236 12/11/1956 CM3/NEGO/217 14/11/1956 CM3/NEGO/236

Document Title

Working Paper III

120

Date

Dossier N°

15/11/1956 CM3/NEGO/236

11/1956 (?)

CM3/NEGO/236

11/1956 (?)

CM3/NEGO/236

11/1956 (?)

CM3/NEGO/236

Document N°

Document Title

MAE 547 f/56

Projet de redaction de la delegation neerlandaise sur les regies de concurrence Proposition italienne du document mar. com. 88 Proposition italienne Proposition de la delegation allemande Projet de redaction des articles relatifs aux regies de concurrence Memento interne du secretariat concernant les regies de concurrence Proposition de redaction concernant l'Article X presente par la delegation franchise Projet de proces-verbal Projet de redaction concernant les regies de concurrence Historique, article 85 Extrait du proces-verbal de la reunion du 06.12.1956 des chefs de delegation Redaction approuvee par les chefs de delegation des articles concernant les normes applicables aux entreprises Projet de proces-verbal des reunions des 27-29 novembre 1956. Redaction approuvee par le comite des chefs de delegation concernant le chapitre relatif aux regies de concurrence Projet de redaction Projet de redaction des articles concernant les entreprises publics et monopoles d'etat, par le secretariat

20/11/1956 CM3/NEGO/236 MAE 602 f/56 20/11/1956 CM3/NEGO/236 20/11/1956

MAE/sec70

CM3/NEGO/236

20/11/1956 CM3/NEGO/145 MAE 539 f/56 28/11/1956 CM3/NEGO/217 MAE 657 f/56 ch. del. 76 CM3/NEGO/236 6/12/1956 CM3/NEGO/236 MAE 781 f/56 10/12/1956 CM3/NEGO/236 MAE 788 f/56

10/12/1956 CM3/NEGO/146

MAE 785 f/56

02/1957 (?) CM3/NEGO/266

MAE262f/57

1/2/1957 8/2/1957

MAE472f/57

CM3/NEGO/236 CM3/NEGO/266

/ - Compatibility, Efficiency, Legal Security

Document N°

121

Date

Dossier N°

Document Title

14/2/1957

CM3/NEGO/236 MAE517f/57

Premiere lecture de la redaction proposee par les rapporteurs du groupe du marche commun

14/2/1957 25/2/1957

CM3/NEGO/236 CM3/NEGO/208 MAE 648 f/57

11/3/1957

CM3/NEGO/271 MAE 814 f/57

Projet de redaction du 14.02.1957 Projet de traite : groupe de redaction Projet de traite instituant la CEE

IV Eleanor M. Fox Modernisation: Efficiency, Dynamic Efficiency, and the Diffusion of Competition Law

I. Introduction Article 81 of the Treaty of Rome governs anti-competitive agreements in the European Union. Article 81 has a dual structure; agreements that, on a broad, first-blush cut may distort competition come within Article 81(1) and are void unless exempted. They are entitled to exemption if they meet four criteria: they must contribute to economic or technical progress, must provide a fair share of benefits to consumers, must not contain unnecessarily restrictive clauses, and must not lessen competition in a significant part of the common market. To obtain exemptions, agreements must be notified to the Commission. Under Regulation 17/62, only the Commission may grant exemptions. The White Paper on Modernisation proposes for Article 81 cases the elimination of the notification/authorisation procedure and the devolution of the authority to grant Article 81 (3) exemptions to the Member State courts and agencies. Provisions would be adopted to avoid Member State conflicts with Commission rulings and legislation, to allow Article 234 references of Community law questions to the Court of Justice, to allow Commission preemption, and to allocate cases, perhaps in accordance with the 'centre of gravity' model. This essay addresses the efficiency of modernisation. I conclude that the modernisation proposal is, in concept, efficient. It satisfies the traditional base points of efficiency: compliance and enforcement. Moreover, by promising fuller engagement of Member States in the Community enterprise, the proposed changes open the way for dynamic system efficiencies. Ultimately, I envision a system enriched by cross-fertilisation and deep co-operation, with development of rules of co-operation among the Member States and between Member States and the Commission. The process is likely to deepen the roots of competition culture within the Member States and across the borders between and among them; possibly, eventually, producing a model that could inform cross-border enforcement and co-operation among jurisdictions in the world trading system.1 1

To be sure, the European Union's common competition law and its goal of deep community integration differentiate the Member States of the European Union from the trading states of the world. But many lessons will probably be adaptable to a world competition system. See note 3 infra.

124

•

Working Paper IV

II. Efficiency In the context of modernisation, efficiency has numerous dimensions. It involves questions such as the following: 1) Will modernisation increase or decrease the degree of compliance with the law, especially with the most important proscriptions of law? 2) Will modernisation increase or decrease the effectiveness of enforcement in cases of non-compliance? 3) To the extent markets, needs, or priorities may change (might there be a shift to prefer more consumer welfare to more parallel imports?), is the modernised system more likely to facilitate efficient adjustment? 4) What is the probable effect of the system—pre and post modernisation—on the efficiency of business transactions? In a separate section, I discuss modernisation and the unleashing of dynamic efficiency by incentivising the Member States to carry out the law of the Community.

1. Compliance Hardcore cartels are reportedly pervasive in Europe, and the potential of the Community system to catch and deter them is far under-realised. The proposed system invites the Competition Directorate to take measures to enhance compliance with this most important proscription of law. (Modernisation is merely enabling in this respect; it is not self-executing.) If the promise of cartel enforcement backed by sufficiently serious fines and consequent deterrence is realised, modernisation will significantly advance market efficiency. Some commentators worry that abolition of the notification system will cause other more arcane violations to 'slip through the cracks', and that a lower prospect for enforcement will mean a greater prospect for non-compliance. But agreements that are problematic in their essence are not likely to be notified currently. Perhaps some clauses in agreements that would otherwise have been notified will escape the authorities' eyes. But the reformation of contract clauses in agreements that would otherwise have been filed is normally of a much lower level of concern than cartels; and, if they do cause harm, victims' complaints will probably prevent the slip through the cracks. In terms of compliance, then, one could expect the new system to be comparatively efficient.

1 - Compatibility, Efficiency, Legal Security

125

2. Enforcement The Commission's enforcement should become more efficient. A notification and authorisation system is costly. The Commission is forced to devote a large percentage of its resources to areas of little pay-off, and has too few resources left for areas of greatest pay-off. Block exemptions have alleviated the problem but not solved it; and indeed they have generated problems of their own by creating over-broad and under-broad frameworks as proxies for what is anticompetitive. The Commission's exemption monopoly is likewise costly to the Commission. Especially if the net of Article 81(1) is spread broadly, Article 81(1) catches a huge percentage of cases that pose no significant threat to competition but which require the attention of the exemption "monopolist". In early days, the combination of the notification/authorisation system and the exemption monopoly helped to create a coherent body of law; but, as the White Paper recognises, the law has matured and the common hand of the Commission is no longer necessary. Devolving of exemption powers to national authorities and courts is on the whole efficient. A national court seized of a contract dispute is handicapped today by not being able to rule on validity/invalidity of a contract under Article 81(3). In the case of agreements caught by Article 81 (1), the whole judicial system is stalled, pending an Article 234 ruling. Arbitrators seized of contract disputes are further handicapped; they cannot even make references to the European Court of Justice. Modernisation will shift some costs to national authorities and courts, and it will impose start-up costs on national judicial systems. But these costs are likely to be minor as compared with benefits for the system as a whole. After the modernisation program is implemented, national authorities and courts will be more fully applying Community law, and thus more frequently applying both Community and national law. This set of circumstances should motivate convergence of national law with Community law; it should tend to push these bodies of law into identity; not just harmony. Pure practicality (the ease of applying only one set of rules), combined with national needs and responsibilities for networking with the centre and with other Member States, can be expected to deepen and widen the common competition culture. Some commentators question the capability of national judges to handle the often complicated competition issues. This is simply a challenge to be overcome. Indeed, the national judges are already charged with applying Articles 81 (1), 82, and 86 when called upon. Modernisation would simply add Article 81 (3) to their repertoires; and if they have construed Article 81 (1) to catch less rather than more, or if they applied Articles 82 and 86, they already performed the complex tasks of competition analysis. The addition of Article 81(3) rights and responsibilities should be synergistic and thus helpful to the national judiciaries in performing the tasks with which they are already charged.

126

Working Paper IV

Potential overlaps of state jurisdictional authority could in theory constitute a systemic inefficiency. But I view this prospect, too, as an opportunity, not a problem. The phenomenon is similar to that generated in the United States by the existence of state and federal laws that may apply simultaneously to the same agreement, and by the authorisation of state attorneys general to bring proceedings under both sets of laws. The U.S. system functions well, without major problems. Collaboration among enforcers—state to state and federal to state—is increasingly both expected and effective. Likewise in the European Union, a networking/comity culture can be expected to develop. Finally, if modernisation frees up Commission resources to enforce Community law in the most critical areas, and if the Commission provides the structure and leadership for this enforcement (as noted similarly with regard to compliance), efficiency in enforcement is bound to rise.

3. Adjustment Under which system can law better adjust to new realities? Pluralism and dispersion of authority conduces to adjustment, while monopoly tends to retard progress. Dissipating the monopoly rights of a sole exempting authority means that more voices will be heard. The pluralism of the U.S. system has been a great benefit in connection with the adjustment of the law to the needs of the new economy. Problems are seen from different angles and windows, through eyes of Antitrust Division enforcers, Federal Trade Commissioners, state attorneys general, defendants, and private plaintiffs. Different federal courts have taken different tacks on hard questions, joining issue and paving the way for changes in law through the ultimate judicial vehicle, the Supreme Court. Community law, under the current structure, may and sometimes does change course. But adjustment is likely to be eased where states are active players and no one authority holds a gate-keeper monopoly.

4. Will modernisation affect the efficiency of business? Competition law should normally not harm the efficiency of business. But the scheme and by-products of Article 81(3) does do harm. This is so in three ways, apart from the transaction costs of notification and approval. First, businesses are attracted by the safe-harbour quality of the block exemptions, as they are expected to be. Yet the block exemptions are straight jackets of red and green lights (though this problem is being eased). The block exemptions steer transactions into second best forms. Therefore businesses do not structure transactions in the best possible way, even if no first-best provision is in fact anti-competitive.

1 - Compatibility, Efficiency, Legal Security

127

Second, almost every agreement between significant market players requires an exemption. The parties desire to get their deal through. They bargain. They often accede to conditions not required in the interests of avoiding harm to competition. It is a cost of the approval process. Third, if some principles of Community law are both inefficient and unnecessary or unhelpful in serving contemporary Community goals, then the law needs adjustment. Under the current system, established principles are unlikely to get tested. Modernisation may facilitate useful change. In sum, I believe that modernisation will lead to a higher level of compliance with the most important competition principles, more effective enforcement, better law, and more efficient business transactions.

III. Unleashing a new dynamic There is yet another facet of efficiency, and it is a dynamic efficiency that relates to the entire scheme and structure of Community law under the Treaty of Rome. The point I develop here is that modernisation felicitously complements basic Community methodology to make the European system work. Articles 81 and 82 must be seen within the broad scheme of Community law under the Treaty of Rome. The Treaty has been successful in creating one common market and one community by actively enlisting the Member States to help carry out Community law. It does this by the powerful tool of direct effect: where the command of the Treaty is clear and is capable of execution without implementing measures, the Treaty law is the law of the Member States. States have the duty to carry out the commands and spirit of the Treaty. When they fail to do so, they can be called to account by the Commission, by other Member States, and by injured persons, in whom rights vest that the national courts are required to protect. Thus, by the doctrine of direct effect and its emanations,2 the Community enlists all sources—Community, national and private—to make a more complete union. Paradigmatically, the Treaty law of free movement of goods and services is directly effective. As a result, it is deeply ingrained in the Member States. States must not unjustifiably restrain trade and competition in the internal market. Indeed, if they do so, they are directly answerable to those whose trade they restrain. The deep market integration that exists could not have occurred without direct effect of the law of free movement. 2

The remedies available to victims in national courts must be full and complete. '[Tjhose measures must be such as to guarantee real and effective judicial protection and have a real deterrent effect on the [violator].' Case C 271/91 Marshall v Southampton and South West Hampshire Area Health Authority (Marshall II) [1993] ECR 1^4367, para. 24 (Member State damage rules must include interest). Robust private remedies may be necessary to fulfill Member State duties. Jeremy Lever, Q.C., puts forth this proposition in 'Private Antitrust Enforcement in the European Community,' conference paper, Washington, D.C., June 23-24, 2000.

128

Working Paper IV

It is often observed that competition law is (in substantial part) the other side of the coin from trade law. If a community is based on the principle that trade must be free, then it must assure that trade is free from undue restraints by private as well as public actors. A respected and well-enforced competition law is therefore also a critical element of community. The drafters of the Treaty of Rome understood this synergy. The Treaty provides for common competition policy. Articles 81(1) and 82 as well as Articles 23 (ex 9) and 28 (ex 30) were given direct effect. But while states' consciousness regarding state restraints is high, and the forces to call straying states to account are strong, the same has not been true of private restraints. Why? The Commission has essentially operated as the monopolist in the development and execution of Community competition policy (though understandably so for the first several decades). Moreover, Article 81 (3) is not directly effective. Thus, States are, by law, forced to stay out of the heart of the process of opining on the validity of agreements. The Commission's notice on co-operation with national authorities of 1997, designed to encourage Member State enforcement of Articles 81(1) and 82, was only a tweak. It was not a strong signal. It has had little effect. And it is necessarily incomplete as long as the Commission retains the monopoly of exemption. By contrast, modernisation carries a strong message, both in rhetoric and in the fact of relinquishment of the Commission's monopoly to apply 81 (3). It nicely fits with globalisation and modern times—the growing consensus for more subsidiarity, for more horizontal networking, and for more grass rootswith-co-operation to anchor and spread a common competition culture.3 It is an idea whose time has come. Therefore, the promise of modernisation is much larger than the sharing of 81 (3) competence. I see it as the unleashing of a real 'direct effect' dynamic that will give each Member State an oar to pull, that will make all Member States valued players in the system, that will bring the Member States into a shared enterprise of policing, catching and deterring obstructions to free competition, 4 thus rounding the circle of direct effectiveness of the Treaty requirement that States must not restrain trade. 3

The analogies to the world system are fertile. Consider, for example, transposing a Fuji/Kodak problem to an internal market problem of the EU. A Member State has the duty to police border restraints. If the restraint is clear and significant, the Commission will wish to take action. The Member State would have the duty to aid the Commission if, for example, its processes and procedures can be of help in procuring evidence. If the restraint is not of general Community concern, the allegedly excluding state would have the duty to cooperate with the excluded one in ascertaining whether there is an anti-competitive restraint and, if so, in prohibiting it; and all States have the duty to build up their procedure, process and remedy arsenals for credible deterrence. 4 It may become important to give national authorities and courts the duty to carry out the Treaty law rather than the Commission's view of Treaty law, when they perceive a conflict, and to provide a right of appeal by parties or the Commission to the Court of First Instance or the European Court of Justice. The more the Member States are mere technicians in the service of the Commission, the less they will rise to the occasion of partner/players in the enterprise.

V Barry E. Hawk and Nathalie Denaeijer The Development of Articles 81 and 82 EC Treaty: Legal Certainty

I. Introduction Legal certainty is a complex and controversial concept with a number of meanings in antitrust law. First, and to us most importantly, legal certainty concerns predictability of outcomes in particular cases, i.e. with what degree of certainty can firms predict that a particular conduct or agreement does or does not violate antitrust laws. (See Section II.) Second, legal certainty concerns the consequences of a violation. How is one punished for a violation of antitrust laws? Under EC law, this would relate to fines and other penalties that can be ordered by the Commission, but also to the enforceability of contractual provisions under Article 81 EC Treaty. (See Section III.) Third, some commentators of the White Paper speak about legal (un)certainty as to the relevant forum or jurisdiction to hear a case. (See Section IV.) The effect of the White Paper on legal certainty should be viewed under these three meanings of legal (un)certainty. Legal certainty in the EC (as to predictability of outcome and consequences of a violation) supposedly comes from the ability to obtain a decision from the Commission declaring an agreement outside the scope of Article 81(1) or exempted under Article 81 (3). In our view, however, the principal cause for uncertainty under the current Article 81 system is the over-broad interpretation by the Commission of the scope of Article 81 coupled with the Commission's monopoly to grant Article 81 (3) individual exemptions. Because many agreements of economic significance contain one or more provisions 'restricting economic freedom', a very large number of agreements risk infringing Article 81 (1) and therefore, risk being void under Article 81 (2), unless exempted. What was meant to be a tool for certainty—notification for clearance or exemption—ended up generating more uncertainty primarily because of the long waiting period for the adoption of decisions and the limited precedent value of comfort letters in national courts in case of litigation. Other tools developed by the Commission to improve the process—the block exemptions—ended up being too formalistic and generated their own ambiguities and legal uncertainties. Thus, the White Paper reforms have the potential to increase legal certainty by eliminating the fundamental cause of uncertainty under the present regime, i.e. it eliminates the Commission's Article 81 (3) monopoly. In other words, the proposed move to a directly applicable exception under Article 81 (3) helps to moderate the threat to the enforceability of contracts (because the parties will be able to argue Article 81 (3) in national courts).

130

Working Paper V

II. Legal certainty about predictability of outcome The first meaning of legal certainty concerns predictability of outcome, i.e. to what extent can firms and/or their counsel predict whether a particular practice or agreement violates competition laws. The question of how to formulate competition rules so as to maximise legal certainty without unduly sacrificing a reasoned legal (and economic) assessment of the specific agreement or practice is not unique to EC law. It has also been heavily debated under U.S. law. At the centre of the debate lies the inherent tension between flexibility and certainty in competition laws. A certain degree of legal certainty is necessary to maintain the etat de droit and to enable businesses to operate. But competition laws, more than many if not most other sets of legal rules, must be flexible to reflect changes in economic thinking and in market conditions. Contrast competition rules with those governing transfer of title to real property where the paramount policy consideration is stability and thus there is far less flexibility in the legal rules. What distinguishes antitrust rules from other sets of rules is the inclusion in the legal analysis of an economics-based approach. The more economics-oriented an approach to Article 81 and 82 (e.g., involving market share and market power analyses), the more uncertainty one must accept in order to have an economically coherent application of the rules. The only way to ensure absolute legal certainty in antitrust laws is to have per se rules—either of prohibition or permission. A.per se rule requires that a particular practice or agreement always be treated in the same way, regardless of who engaged in the conduct and regardless of the effects of the conduct. There is no inquiry into the particular facts of the case, i.e. no inquiry into actual effects (benefits and harms), market power, market conditions, efficiencies, etc. The question to those passionate demanders of legal certainty is: 'Which practices and agreements do you want condemned and which permitted without any inquiry into effects, business justifications, efficiencies, market power, etc.? And what is the price to pay for such legal certainty?'

1. Per se rules of prohibition The underlying policy judgment for per se rules of prohibition under U.S. law is that certain practices have so little redeeming competitive benefit and can result in so much competitive harm that on balance they are condemned as a class. The U.S. Supreme Court in Jefferson Parish declared: (T]he rationale for per se rules in part is to avoid a burdensome inquiry into actual market conditions in situations where the likelihood of anti-competitive conduct is so great as to render unjustified the costs of determining whether the particular case at bar involves anti-competitive conduct.' Jefferson Parish Hosp. Dist. No.2 v Hyde, 104 S. Ct. 1551 (1984).

1 - Compatibility, Efficiency, Legal Security

131

Thus there is no case-by-case inquiry into the competitive effects (or market power) or into the efficiencies or justifications that the parties may want to put forward. Since by definition the nature of the per se prohibition rule forecloses an indepth analysis of the effect of the restraints on the market, the U.S. Supreme Court has limited its application to practices which are manifestly anticompetitive.2 In horizontal cases, the per se rule in the United States is largely limited to basic hard-core cartel practices. This is similar to EC practice. Indeed, the bifurcation of Article 81 and the (theoretical) possibility of an exemption obscure the fact that per se (-like) rules exist under EC law with respect to hardcore cartels (and restrictions on parallel imports). Indeed, the extraordinary and insightful judgment of the Court of First Instance in European Night Services comes very close to dividing Article 81 into per se offences and non-per se offences.3 In European Night Services the Court of First Instance stated: Before any examination of the parties' arguments as to whether the Commission's analysis as regards restrictions of competition was correct, it must be borne in mind that in assessing an agreement under Article 85 (1) of the Treaty account should be taken of the actual conditions in which it functions, in particular the economic context in which the undertakings operate, the products or services covered by the agreement and the actual structure of the market concerned (...), unless it is an agreement containing obvious restrictions of competition such as price-fixing, market-sharing or the control of outlets. In the latter case, such restrictions may be weighed against their claimed pro-competitive effects only in the context of Article 85 (3) of the Treaty, with a view to granting an exemption from the prohibition in Article 85 (1).

In vertical cases, the U.S. practice went from a complete per se prohibition of all vertical price, territorial and customer restraints to the concurrent application of a per se prohibition and a rule of reason.4 The only vertical restraint now subject to per se prohibition is minimum resale price maintenance. The list of hard core restrictions that the EC Commission includes in the new Vertical Restraints Block Exemption is substantially longer than the list of per se prohibitions in vertical cases under U.S. law, notably with regard to 2

The Supreme Court has made clear that the per se rule is an exception to the rule of reason and ought to be applied with prudence because a. per se rule applied to a category of restraints risks prohibiting some potentially pro-competitive agreements and prevents a defendant from justifying its conduct. See e.g., Continental T. V. Inc. v GTE Sylvania, Inc., 433 U.S. 36 (1977). 3 Joined Cases T-374/94, T-375/94, T-384/94, and T-388/94, European Night Services and Others v Commission [1998] ECR H-3141. 4 The shift to the rule of reason occurred in GTE Sylvania when the Supreme Court was no longer convinced that it could generally be presumed that non-price vertical restraints harmed competition and lacked consumer benefit justifications. The per se rule then was maintained for vertical price restraints (fixed, maximum and minimum resale price maintenance). The shift to the rule of reason, however, continued and culminated in 1997 in the Supreme Court's reversal of the per se ban on maximum resale price maintenance in Kahn.

132

Working Paper V

territorial and customer restrictions.5 The underlying rationale for the Commission's overall resistance to territorial restrictions is the fundamental Community objective of market integration, a policy objective not present in U.S. competition policy. Under U.S. law, territorial restraints (preventing a dealer from selling outside an assigned territory) and customer restraints (preventing a dealer from selling to particular categories of customers) are assessed on the basis of their effect on inter-brand competition and are subject to a rule of reason, not a per se prohibition. In applying a rule of reason to territorial and customer restrictions, U.S. courts generally examine the purpose of the clauses, the market share of the supplier and their effect on intra-brand and inter-brand competition considering pro-competitive benefits (generally a harm to intra-brand competition alone is not sufficient for a Section 1 violation). In practice, applying a per se prohibition rule to a particular practice or agreement does bring the maximum degree of legal certainty. For example, once the conduct is pigeonholed as 'pricefixing,'there is no need to inquire into the following chain of elements, all of which increase legal uncertainty: actual effect on prices/output/innovation/intra or inter-brand competition, market power, market shares (and definition of a relevant market) and asserted efficiencies or other justifications. However, a multiplication of per se prohibition rules may not be advisable. There is indeed a price to pay. U.S. courts and practitioners have always recognised that a. per se rule involves a certain degree of arbitrariness and broad generalisations that may result in some incorrect rulings. This is accepted on the basis of the underlying assumption that the presumption of anti-competitive effects is so strong that the gains from the imposition of the rule will far outweigh the losses.

2. Per se rules of permission The converse of the per se prohibition rules axe per se rules of permission which also provide complete legal certainty. The underlying policy judgment here is that certain practices have so little potential competitive harm and have 5 For example, with regard to territorial and customer restrictions, the Vertical Block Exemption blacklists (i) restrictions on the territory into which, or of the customers to whom, the buyer may sell the goods (except restrictions on active sales into the exclusive territory or to an exclusive customer group reserved to the supplier or allocated by the supplier to another buyer, restrictions on sales to end users by a buyer operating at the wholesale trade level, restrictions of sales to unauthorized distributors by members of a selective distribution system, and restrictions on the buyer's ability to sell components to customers who would use them to manufacture products competing with the supplier's products); (ii) restrictions on active or passive sales to end users by members of a selective distribution system operating at the wholesale trade level; or (iii) restrictions on cross-supplies between distributors within a selective distribution system.

1 - Compatibility, Efficiency, Legal Security

133

sufficient competitive benefits that on balance they are permitted as a class without case-by-case inquiry into their actual competitive effects in a particular case (or market power, efficiencies, etc.). It is difficult to identify per se legal practices under Section 1 of the Sherman Act. Perhaps the only example is exclusive distributorships, but even here the rule is qualified by a monopoly power caveat. Under Article 81, one might consider that rules of per se permission are included in block exemptions (or at least 'white listed' clauses) and the De Minimis Notice.6 For example, the old block exemptions on vertical restraints operated much like per se permissive rules because the market power of the parties was not relevant. This explains the passionate resistance to the introduction of market share thresholds in the new Vertical Restraints Block Exemption. Opponents to the market share threshold object that the legal certainty of the old block exemptions has been removed.7 Will the new Vertical Restraints Block Exemption increase legal certainty as claimed by the Commission? It is helpful to distinguish two questions. Will legal certainty be increased because more agreements will benefit from the block exemptions? Will legal certainty be increased because the block exemptions will make cases more 'outcome predictable'? Will the new Block Exemption cover more agreements? In theory, yes. The Commission's intention was to draft the new Block Exemption much wider so as to apply to more agreements (or more restrictions contained in agreements). To this extent, .one can say that the new Block Exemption will increase legal certainty by increasing the number of agreements that could in theory benefit from the block exemption. However, the Commission appears to take away what it 6

See European Commission: Notice on agreements of minor importance which do not fall under Article 85 (1) of the Treaty establishing the European Community, OJ C 371 of 9.12.1997. It is questionable whether the De Minimis Notice really contains/w se permissive rules. The Notice establishes presumptions that in certain conditions a restriction will have 'no appreciable effect' on competition (based on market share and/or turnover thresholds) and therefore is permitted. The same practice engaged in by another firm could be prohibited and thus these rules do not constitute per se rules of permission as denned above. 7 Critics, however, have questioned the track records of the old vertical block exemptions. They have provided some degree of legal certainty, but at the price of inflexibility and with a straightjacket effect. They have also generated some degree of uncertainty, e.g., through the gray clauses, application of the opposition procedure and the Commission's right to withdraw the benefit of the block exemption. Introducing the draft of a new vertical block exemption, the Commission presented it as a new type of block exemption regulation, one that will no longer restrict exemption to certain specific agreements and clauses identified in the Regulation, but rather will provide a general exemption for all agreements and clauses in a given category (subject only to a list of prohibited restrictions, specific conditions of application and a restriction of the benefit of the general exemption through a market share criterion). The Commission argued that the new generation block exemption would provide legal certainty for a wider category of agreements.

134

Working Paper V

just gave, by introducing a market power threshold. The wider Block Exemption will be available only to firms below a market share threshold. This limits the number of agreements that will benefit from the new block exemption. More importantly, the new Block Exemption will reduce outcome predictability because of its introduction of a market share screen and the attendant vagaries of market definition. However, if competition laws are to have economic meaningfulness, and if per se rules cannot regulate all antitrust conduct or agreements, then some form of a market power (market share is a surrogate) threshold is necessary. This is the approach followed in the United States for practices not falling into per se categories.

3. The 'rule(s) of reason' If a particular practice or agreement cannot be pigeonholed into a 'per se' category (prohibitive or permissive), some inquiry into the actual competitive effects (benefits and harms) of the practice or agreement is required. In the United States this inquiry is undertaken under the rubric of the 'rule of reason.' Use of the singular 'rule of reason' grossly misrepresents the anarchic state of U.S. law. There are numerous formulations of the 'rule of reason,' culminating in the Delphic (which is unkind to the Greeks) formulation of the Supreme Court in last year's decision in California Dental* The 'legal uncertainty' generated by the U.S. 'rules of reason' dwarfs the situation under Article 81. The classical definition of the rule of reason analysis was given by the Supreme Court in 1919: The true test is whether the restraint imposed is such as merely regulates and perhaps thereby promotes competition or whether it is such as may suppress or even destroy competition. To determine that question, the court must ordinarily consider the facts peculiar to the business to which the restraint is applied, its condition before and after the restraint was imposed, the nature of the restraint and its effect, actual and probable. The history of the restraint, the evil believed to exist, the reason for adopting the particular remedy and the purpose or end sought to be achieved are all relevant facts. Commentators have criticised the classical rule of reason on the ground that it is too broad and too costly for the parties. One of the most vocal critics has been Professor, now Judge, Frank Easterbrook. He argues that judges and juries cannot realistically perform the entire evaluation entailed by the rule of reason (citing the Supreme Court's own admission that courts 'are ill-equipped and ill-suited for [economic] decision making [and cannot] analyse, interpret and evaluate the myriad of competing interests and the endless data that would surely be brought to bear on such decisions').9 Easterbrook also finds the rule of reason 'empty' on the ground that so many factors and facts are deemed 8 9

California Dental Associations v FTC, 526 U.S. 756, 119 S. Ct. 1604 (1999). United States v Topco Assocs, Inc., 405 U.S. 596, 629-12 (1972).

1 - Compatibility, Efficiency, Legal Security

135

relevant ('when everything is relevant, nothing is dispositive') that nothing helps business plan their conduct.10 These and other criticisms of the rule of reason have prompted proposals for a more structured rule of reason analysis that examines the various issues involved in the analysis in specific sequences through shifting burdens of proof. A description of the multitudinous formulations of the 'rule of reason' is well beyond the scope of this paper. Over the years courts, authorities and commentators have formulated 'quick looks' and 'truncated,' 'structured,' 'flexible' and 'abbreviated' rules of reason (to date developed primarily in horizontal cases), which are sort of a middle ground between the per se rule and a full-blown rule of reason. The most recent Supreme Court (unsuccessful) attempt to formulate an operable rule of reason is California Dental which reversed the appellate court's 'quick look' and called for a 'quicker look.'11 As is painfully obvious from California Dental, U.S. law has failed to evolve a clear operable truncated rule of reason. The FTC brought a complaint against the California Dental Association ('CDA') for unreasonably restricting price advertising by its members (discounted fees in particular), and advertising relating to the quality of dental services. An Administrative Law Judge found a violation of Section 1 of the Sherman Act, ruling that 'although there had been no proof that the CDA exerted market power, no such proof was required to establish an antitrust violation (...), since the CDA had unreasonably prevented members and potential members from using truthful, non-deceptive advertising, all to the detriment of both dentists and consumers of dental services.' On appeal the FTC found a violation and entered an order. The Commission condemned the restrictions on price advertising as per se illegal and the non-price advertising restrictions under a rule of reason which, per the Commission, was a 'quick look' at the rule of reason. CDA appealed to the Court of Appeals for the Ninth Circuit. The Ninth Circuit ruled that the Commission erred in applying the per se analysis to the price advertising restrictions, declaring that the rule of reason should have been applied to all the alleged restrictions. But the court of appeals went on to explain that the Commission had properly 'applied an abbreviated or 'quick look' rule of reason analysis designed for restraints that are not per se unlawful but are sufficiently anti-competitive on their face that they do not require a full-blown rule of reason inquiry.' The court of appeals found the quick look analysis appropriate for several reasons. First, it found that the restrictions on discount advertising amounted in practice to 'a fairly naked restraint on price competition itself and that the CDA's pro-competitive justification that the restrictions prevented false and misleading advertising carried little weight because 'the record provides |no evidence that the rule has in fact led to increased disclosure and transparency of dental pricing.' Further, 10

See Easterbrook F. (1986): 'Antitrust law enforcement in the vertical restraints area: Vertical arrangements and the rule of reason', 53 Antitrust Law Journal, 135. 1 ' See generally Calkins S. (2000): 'California Dental Ass'n: Not a quick look but not the Full Monty', 67 Antitrust Law Journal, 495.

136

Working Paper V

the court of appeals found that the non-price advertising restrictions were in effect a form of output limitations because they restricted the supply of information about individual dentists' services and thus the restrictions were sufficiently naked restraints to justify a quick look analysis. On appeal, the Supreme Court reversed, distinguishing CDA from other cases in which it had accepted a quick look analysis. In prior cases the Court held that 'naked restraint on price and output requires some competitive justification even in the absence of a detailed market analysis'12 and that 'no elaborate industry analysis is required to demonstrate the anti-competitive character of horizontal agreements among competitors to refuse to discuss prices. The Court explained these rulings as follows: 'In each of these cases, which have formed the basis for what has come to be called abbreviated or "quick look" analysis under the rule of reason, an observer with even a rudimentary understanding of economics could conclude that the arrangements in question would have an anti-competitive effect on customers and markets.'

The Court found that the CDA case, on the other hand, failed to present a situation in which the likelihood of anti-competitive effects is comparably obvious on the reasoning that 'the restrictions on both discount and non-discount advertising are at least on their face designed to avoid false or deceptive advertising in a market characterised by striking disparities between the information available to the professional and the patient.' The plausibility of procompetitive claims about the effects of the restrictions 'rules out the indulgently abbreviated review to which the Commission's order was treated. The obvious anti-competitive effect that triggers abbreviated analysis has not been shown.' Justice Breyer in his dissenting opinion in CDA offered another alternative to the quick look review, which may be an example of a sliding scale test. He proposes to break the question of whether a restraint is anti-competitive into four questions: (a) (b) (c) (d)

What is the specific restraint at issue? What are its likely anti-competitive effects? Are there offsetting pro-competitive justifications? Do the parties have sufficient market power to make a difference?

An analysis of the competitive effects (benefits and harms) of a practice necessarily introduces some legal uncertainty. It is probably fair to say that the more refined/robust the inquiry into the actual competitive effects and justifications of a practice, the greater the uncertainty. In other words, there is a trade-off between legal certainty and a robust analysis in a particular case. The logical first step of a robust analysis would be to inquire into the effects of the practice on prices or output. Effect on prices and output often is very difficult to prove, however. Market power (and market definition) frequently 12

National Collegiate Athletic Assn. v Board of Regents of Univ. of Okla., 468 U.S. 85, 104 S.Ct. 2948, 82 L.2d 70 (1984).

1 - Compatibility, Efficiency, Legal Security

137

serves as an alternative method to gauge effective prices or output. Justice Breyer in his dissenting opinion in California Dental Association explained the importance of the market power: Market power is part of the rule of reason analysis, but it is important to remember why market power is examined. We consider market power to help inform our understanding of the competitive effect of a restraint. Where the consequences of a restraint are ambiguous, or where substantial efficienciesflowfrom a restraint, a more detailed examination of market power may be needed.

Of course, market power also is not necessarily easy to measure in the litigation context. Parties typically will need to define markets, calculate market shares, analyse ease of entry or barriers to entry, calculate elasticity of supply or demand, show excess capacity in the industry, and discuss customers countervailing bargaining power and dynamics of the market, etc. Market power analysis inevitably reduces legal certainty about predictability of outcome.13 In sum, there is considerable legal uncertainty under U.S. antitrust law in those cases wherever se rules do not apply. Predictability of outcome probably is far less clear than under Article 81. Nonetheless, the European concern (obsession?) with legal certainty is largely absent in the United States where intellectual efforts are devoted toward formulations of a less costly and more predictable rule of reason rather than calls for more per se rules (of either prohibition or permission). Speaking very generally, U.S. jurists more easily accept the greater legal uncertainty resulting from the inclusion of economics in antitrust law. 4. White Paper proposals to increase outcome predictability Commentators have argued that the elimination of the notification system and the limit on the type/number of binding decisions that will be issued by the Commission introduces legal uncertainty. Moreover, the economic analysis and the decentralisation brought about by the new system will create some additional degree of uncertainty. The Commission, however, proposes several measures to limit these uncertainties. 4.1. Reliance on Commission notices and guidelines Current guidelines and notices have a mixed record. Certainly, the Merger Regulation notices have been of tremendous help to lawyers and firms by providing reliable guidance on key issues. The situation under Articles 81 and 82 cases is less clear. 13 If market share is used to measure market power, further uncertainty can result from market definition. However, this would create less uncertainty than other measurements of market power like the Lerner Index.

138

Working Paper V

The Commission started using notices in 1962 to indicate which agreements did not fall within Article 81 and therefore need not be notified for negative clearance (de minimis agreements, commercial agency, co-operation agreements, subcontracting agreements, co-operative joint ventures, etc.). The success of these notices is not easily measurable (the number of notification for negative clearance may have somewhat declined). To be successful, the proposed new notices will need to be clear,flexibleand updated regularly. One important question is to what extent the Community Courts, the national courts and the national competition authorities will follow these notices. Guidelines are also part of the U.S. antitrust practice. Guidelines currently in force include the Joint Venture Guidelines (2000), the Horizontal Merger Guidelines (1992), the Intellectual Property Guidelines (1995), the International Guidelines (1995) and the Health Care Statements (1996). Although these Guidelines are technically not binding, as a practical matter, they have an enormous influence on the authorities' activities. 4.2. Optional notification for 'gray area' agreements and conduct Some commentators have suggested that the Commission should maintain an option to notify agreements when they involve considerable investments and there is a risk that the agreement or important provisions might not be enforceable.14 Parties would have the option to notify and the Commission would be required to decide within a reasonable time frame. The Commission could send a notification to national competition authorities for treatment under Article 81 if the agreement was national in character, on the basis of well-defined jurisdiction allocation criteria. But it is not clear that firms notify only where considerable investments are at stake and where they have serious doubts about the compatibility of their agreements with Article 81. An interesting parallel is with Article 82. There are probably many more gray area issues under Article 82, resulting from the ambiguity of the legal principles, the application of economic principles and the difficulty of applying the law to complex facts. Yet, dominant firms seem to live with the uncertainty of un-enforceability of their agreements (if any), the risk of fines (which is probably higher under Article 82 than under Article 81) and the risk of damages (which again is a higher risk than with regard to agreements under Article 81). In the United States there also is a voluntary notification system. Under this procedure, parties may notify the Justice Department of a proposed business practice or agreement (providing all necessary and relevant background information) and request a Business Review Letter. The Justice Department may (1) 14 See, e.g., Siragusa M. (2000): 'A Critical Review of the White Paper on the Reform of the EC Competition Law Enforcement Rules', in Barry Hawk (ed. ) Fordham Corporate Law Institute, ch. 15.

1 - Compatibility, Efficiency, Legal Security

139

state its present enforcement intention, (2) decline to state its enforcement position, or (3) take any other position or action it considers appropriate. In over 95% of cases, the Justice Department states its intention not to challenge the notified proposed conduct. An expedited review (sixty to ninety days) is possible for agreements to form joint ventures, and agreements to collect and disseminate business information; otherwise there are no deadlines for a response. This procedure is rarely used (DOJ receives on average less than 25 requests per year), even where significant investments are involved, probably because it has some substantial downsides. The Business Review Letter as well as all the information submitted by the party are made public; the stated enforcement intention of the Justice Department is valid only as of the date of the letter (it is free to bring any action it subsequently decides is appropriate); the letter does not protect against action by a third party; and Business Review does not relieve a party from pre-merger notification obligations. These downsides appear largely similar to the disadvantages and cost of the European notification comfort letter system. One may ask whether the demand for legal certainty (and the itch to notify/run to the government to get permission to do business) comes more from EC lawyers representing the same firms who fail to seek U.S. business review letters for the same transaction. However, there is need for some procedural mechanism to permit parties to agreements involving large investments where there is uncertainty about the legal rules—perhaps some form of limited notification system with sanctions imposed on the lawyers if the Commission finds that there were no non-trivial antitrust concerns raised by the agreement! 4.3. Reliance on existing case law The Commission states in the White Paper that firms and lawyers can rely on the existing 40 years of decisions and case law in their assessment of whether an agreement or practice is likely to infringe Article 81 (1) and require exemption. The reliability of prior decisions and case law probably depends on the issues involved. The case law is well developed and, in our view, unlikely to change, for issues such as what constitutes an agreement under Article 81, the treatment of price fixing, certain abuses under Article 82 and the types of conduct warranting fines. But there are areas of uncertainty where more case law is needed. The Commission also commits to publish its decisions to continue to expand the case law and clarify its positions on specific issues. The White Paper, however, appears to propose that the Commission's decisions will be binding only if they are negative or if they impose conditions on the parties. Positive decisions that would clarify the scope of application of Articles 81(1) and 81 (3) outside of the block exemptions or confirm the approach set out in guidelines and notices would merely be of a declaratory nature and would have the same legal effects as negative clearance decisions, i.e. not binding. The result could be that positive decisions of the Commission could be disregarded by national

140

Working Paper V

competition authorities, a factor of legal uncertainty. This aspect of the White Paper appears troublesome.

III. Legal certainty about the consequences of a violation The second meaning or definition of legal certainty concerns predictability about the legal consequences of the finding of a violation of the competition rules. As to fines, we believe that there is today considerable legal certainty under EC law as to which agreements and abuses will result infines.Thus, the abolition of the notification system will not reduce certainty in this respect. Fining agreements are not notified (and post-notification immunity would be lifted even if notified) and the vast majority of notified agreements do not present a realistic risk of fines, i.e.finesare not a significant factor of legal uncertainty.15 There is, however, significant uncertainty about the amount of fines that will be imposed in a particular case. The problem lies here with the weaknesses in the Commission's leniency program under which the Commission is unable to provide assurances to co-operating firms about the exact reward (i.e. the amount of thefinereduction) for co-operation in a particular case. The quid pro quo is unknown.16 But the abolition of the notification system will not worsen the present situation. There is a stark contrast between Europe and the United States with respect to one issue of legal certainty about the legal consequences of an antitrust violation: that is the passionate (and over-inflated) European concern expressed about the enforceability of contractual provisions under Article 81. The advice to the client can go as follows: 'Although there is no risk of fines, you should notify because you might get into litigation in a national court with your partner or a third party.' Sometimes, there are pompous admonitions about 'obeying the law' of Article 81 (I).17 This concern (may we say obsession) with enforceability of contractual provisions does not exist under U.S. antitrust law. For example, there is no debate about the enforceability of vertical restraints should they be challenged as unreasonable under the antitrust laws. U.S. lawyers and businesses appear to live with the uncertainty. Why? 15

See Forrester I. (2000: 'Modernisation of EC Competition Law', in Hawk B. (ed.) Fordham Corporate Law Institute 1999, ch. 12. 16 See Hawk B.: (1999): 'Administrative/Prosecutorial Discretion of Antitrust Authorities', in Hawk B. (ed): Fordham Corporate Law Institute 1999, (B. Hawk ed. 2000), ch. 31. 17 As to the latter comment, we should remember the admonition of Professor Hart who taught that there is no moral duty to comply with legal commands. See Hart H.L.A., Raz J. and Bulloch P.A. , eds. (1994): The Concept of Law, Oxford Clarendon Press.

1 — Compatibility, Efficiency, Legal Security

141

There are several possible explanations. First, firms prefer an uncertain rule of reason to a. per se prohibition rule. For example, suppliers (generally the parties in Europe complaining the most about the loss of legal certainty that affects the enforceability of their agreements) were helped by the shift in the United States from per se prohibitions on territorial and customer restraints to a rule of reason and therefore had no incentive to complain about the lesser degree of legal uncertainty (either of the predictability of outcome or legal consequences). Second, the shift to an uncertain rule of reason has helped defendants (generally manufacturers or producers) by requiring the plaintiff to establish that the challenged practice had harmful effects (or that defendant has market power). This shift of the burden of proof puts the bulk of the litigation effort and associated costs on the plaintiff and limits the number of cases brought against defendants/suppliers. Third, U.S. parties are more concerned about the risk of treble damages than on ensuring the enforceability of their contracts.

IV. Legal Uncertainty as to the Forum Some commentators speak about the legal (un)certainty as to the relevant forum or jurisdiction to hear a case. For example, some critics assert that the White Paper's proposed decentralisation of the application of Article 81 (3) generates 'uncertainty' because the parties are not sure that the decision of one authority is binding on others and they may face risks of double jeopardy. Does the proposed decentralisation change the current relationships among the Commission, the national courts and the national competition authorities and does it reduce legal certainty? The White Paper states that the Commission's efforts to increase cooperation at both levels have largely been a failure: complainants are still reluctant to bring EC law complaints to national courts or authorities; national proceedings can be thwarted by notifications to the Commission; the Commission still has a monopoly to exempt under Article 81 (3). The Commission proposes to decentralise the application of Articles 81 (in its two prongs) and 82 to national competition authorities and national courts. Indeed, given the size of the EC and the breadth of competition law, decentralisation is inevitable and necessary. Without a substantial increase in the Competition Directorate's human andfinancialresources, the Commission can no longer play an almost exclusive role. National authorities and courts must help. Will the proposed decentralisation help or hurt legal certainty? A few tentative answers follow, mostly to stimulate discussion.

142

Working Paper V

1. National Competition Authorities The Commission states that national competition authorities are well placed to take effective action in certain types of cases: they are not there to settle a dispute between parties, but rather to ensure that competition is not distorted; they are familiar with the national markets and local operators; most have the human and legal resources needed to take action within their territory; some of them have the infrastructure needed to carry out nation-wide investigations quickly; and they are closer to complainants. 1.1. Is there a risk that legal certainty will be reduced by the power of national competition authorities to withdraw the benefit of block exemptions? The national competition authorities will be given the power to withdraw the benefit of a block exemption on their territory if that territory or part of it constitutes a separate market. The Commission finds that this is justified on the ground that national authorities are well placed to assess whether an agreement in their territory, although falling within a block exemption, does not meet the conditions of Article 81 (3). This could increase legal uncertainty to the extent that (1) the parties would fear during the life of their agreement that a national competition authority could at any time withdraw the benefit of the block exemption; or (2) the parties could end up with an agreement that benefits from the block exemption in some Member States and not others. Realistically, how often will this happen? The Commission rarely has withdrawn the benefit of a block exemption. Should it be feared that the national competition authorities will do it more regularly? 1.2. Is there a risk that legal certainty will be reduced by the power of the national competition authorities to grant individual exemptions and the resulting risk of inconsistency between the Commission and a national authority or between national authorities? An important part of legal certainty is the ability to rely on one decision by one authority. Conflicts between NCAs and the Commission could arise if a national authority exempts an agreement that is prohibited by the Commission, or where the Commission takes a favourable view of an agreement but one or several national authorities prohibit it. The risk of inconsistent rulings between several NCAs is increased by the fact that (i) not all national competition authorities currently have the power to apply Community law; (ii) not all authorities may have the adequate experience or resources to conduct the complex analysis required under Article 81(3); (iii) there are differences in the national authorities and the procedural rules to which they are subject (independent bodies v. government cabinets, administrative v. judicial appeals, etc.);

1 - Compatibility, Efficiency, Legal Security

143

and (iv) the decisions of a national authority are not recognised (under any Conventions) in another country. 1.3. Will decentralised application of Article 81 (3) lead to forum shopping by applicants and complainants? The White Paper proposes information exchange measures and co-ordination procedure to limit the risk of inconsistent outcomes between the Commission and any particular national competition authority, but nothing is envisaged to address the risk of conflicting or inconsistent decisions by several national authorities that could be examining the same case. The Commission says that it wants to establish a 'network of authorities operating on common principles in close collaboration.' The White Paper offers several measures to attempt to avoid that decentralisation result in an inconsistent application of Community competition law.18

2. National Courts Currently, because of the Commission's exclusive power to apply Article 81 (3), parties can stop national proceedings by notifying an agreement to the Commission. Granting Article 81 (3) jurisdiction to national courts should 18 First, competition law policy will continue to be determined by the Commission at Community level by legislative texts (notices, guidelines, block exemptions and regulations) and individual decisions. Second, the Commission refers to the rules that are currently in place to ensure primacy of Community law and the coherent application of Articles 81 and 82 in parallel national and Community procedures. The Commission finds that these rules have worked well so far, under the supervision of the European Court of Justice, but states that it would nevertheless be advisable to establish information and cooperation mechanisms to avoid conflicts in thefirstplace. The Commission proposes to impose a reporting obligation on national competition authorities. Specifically, competition authorities will be under an obligation to inform the Commission of cases in which Articles 81 and 82 are applied by the authority—the Commission must be informed of the initiation of proceedings and before their termination; the Commission must also be informed if an authority intends to withdraw the benefit of a block exemption. Some commentators have noted that this reporting obligation will effectively introduce a new type of notification with which the Commission will have to deal, defeating the purpose of the decentralization process. The Commission also proposes that the authority considering a case, whether at EC or national level, must be in a position to pass afileto another authority that would be best placed to handle it. If the Commission finds that the effects of a disputed practice are located essentially in one Member State, it would be entitled to send the whole file (including confidential information) to the competent national authority for it to continue the investigation. Conversely, if after a detailed investigation, a national competition authorityfindsthat the case has a Community dimension and requires Commission action, it can refer the whole case to the Commission.

144

Working Paper V

increase legal certainty because the courts will be able to hear the whole Article 81 case. Some query whether the power given to the national courts reduces legal certainty. One concern is that empowering the courts to apply Article 81 (3) will require every court to balance the economic efficiencies of an agreement with its restrictive effects and most national courts are not equipped for this analysis. I belong to the perhaps hopelessly naive school of thought that judges are just as able to determine Article 81 (3) issues as Article 81 (1) issues, at least where the inquiry is limited to competition criteria. The proposed reform may heighten obstacles to the effective application of EC law. But this is a separate and different issue from that of legal certainty. There are many obstacles today to the effective application of Community competition law by national courts, including procedural differences, different rules of evidence, etc. These obstacles, however, do not directly concern legal certainty. Procedural harmonisation would advance the effective application of Articles 81 and 82 by national courts throughout the Community. In our view, the Commission probably exaggerates the positive benefits of decentralisation to national courts for the effective application of Articles 81 and 82. The Commission in the past has been unduly optimistic about the role that national courts could play (and it recognises that in the White Paper). If the Commission decides that it wants to take measures to increase the effective application of competition law at the national level—and this will likely become a more important issue in the near future given the growth of the EC and the possible subjects to the law (and their different level of expertise with Community laws or competition law in general)—it will have to make a fundamental choice between 'decentralising' (with all its potential problems and costs) or significantly increasing the Commission's enforcement resources, notably staff. 2.1. Is there a risk that legal certainty will be reduced if courts adopt inconsistent decisions/remedies? The Commission proposes mechanisms for close co-operation with the national courts to eliminate this risk, notably that courts will be required to inform the Commission of any proceedings in which Articles 81 and/or 82 are invoked. There have been few cases where inconsistent judgments were adopted by different national courts on the same agreement, or that a national court's judgment was inconsistent with a Commission decision. The White Paper provides that the Commission could still prohibit an agreement or practice that a national court has permitted. To conclude, we do not fear more judgments by more national courts whether or not they are fully consistent either among themselves or with the Commission. We do not believe that the Commission has a monopoly of competition law wisdom. We think the benefits of the case law system outweigh the legal uncertainties. Epluribus unum.

VI Giuliano Marenco Does a Legal Exception System Require an Amendment of the Treaty?

I. Introduction The proposal in the Commission's White Paper1 to replace the present system of notification and exemption of restrictive agreements with a system whereby Article 81 EC would be directly applicable in its entirely has met inter alia with objections as to the conformity of the proposed system with the Treaty. In particular, it is argued that the proposed system (also referred to as the legal exception system, meaning that the waiver from the prohibition would stem directly from the law and not from an administrative decision) runs counter to Article 81 itself, insofar as: • paragraph 1 thereof provides for a directly applicable prohibition of restrictive agreements; • paragraph 3 states that ' The provisions of paragraph 1 may, however, be. declared inapplicable in the case of. . .' (emphasis added), thereby allegedly referring to an administrative decision; and • the conditions laid down in paragraph 3 could not be rendered directly applicable, for they would not lend themselves to judicial interpretation, thus confirming the necessity of an administrative decision. Other objections are derived from Articles 83 and 84 EC Treaty. Some of these objections derive from misinterpretations of the Commission position. It is therefore necessary to rectify these misunderstandings at the outset.

II. The proposed system is not a system of control of abuse In a control of abuse system a decision is necessary before any given behaviour can be considered illegal. Before such a decision, firms are unfettered in the exercise of their freedom of contract. Only if they persist after a finding that they have abused this freedom are they liable to pay a fine or damages and is

* Any views expressed are personal to the author, who is indebted to Wouter Wils, Emil Paulis and Karen Banks for their various valuable contributions. Special thanks are addressed to Mr Fernando Sanchez-Martin of the Council's historical archives. 1 OJC132, 12.5.1999.

146

Working Paper VI

their contract void.2 A legal exception system works exactly the other way round. From the beginning the agreement would be either legal or illegal with regard to Article 81 taken in its entirety (that is, both the prohibition laid down in paragraph 1 and the derogations provided for in paragraph 3), with the consequence that the nullity of the agreement as well as the exposure to fines or damages would run as of the moment at which the agreement was entered into. No decision would be necessary for the illegality to exist. To be sure, in order to inflict a fine, a decision must be adopted, but for the quantification of the fine the whole period since the entering into the agreement can be taken into account. The two systems differ widely as to their deterrent effect, which is nil in a control of abuse system and greatest in a legal exception system. This should reassure those critics who have analysed the proposed system as one of abuse control. The first and most vigorous of these critics in Mestmacker.3 He writes that the system put forward in the White Paper means that every cartel is legal from its inception until the Commission, a national competition authority or a national judge decides it is incompatible with Article 81; that there would be no presumption that cartels are void, that the prohibition would only bite once it has been found not only that there is a restriction of competition, but also that it does not contribute to the economic advantages mentioned in Article 81 paragraph 3; and that the principle that an agreement is legal until a decision finding an infringement is adopted would apply regardless of the kind and degree of the restriction of competition. He adds that all cartels mentioned as examples in paragraph 1 could be conducted 2

A system of control abuse existed in the Netherlands under the Statute of June 28th 1956 {Wet houdende regelen omtrent de economische mededinging, Stb. 1958, 413). Agreements and decisions regulating competition between firms {mededingingsregelingen) had to be notified. The relevant Ministers could • decide that the mededingingsregeling would be compulsory in the whole industry (algemene verbindenverklaring), • prohibit generally certain types of mededingingsregelingen (generieke algemene onverbindendverklaring), • prohibit a mededingingsregeling as contrary to the public interest or as applied in a way which is contrary to public interest (individuele onverbindendverklaring). The obligation to notify was the only directly applicable provision of the Statute. A mededingingsregeling was perfectly legal before possibly becoming the target of an onverbindendverklaring. Similarly, under the Belgian Statute of 27 May 1960 (Loisur la protection contre I'abus de la puissance economique, Moniteur beige, 22 juin 1960, p. 4674; erratum 6 juillet 1960, p. 5210) only the abuse of economic power was illegal and not the mere existence of a monopoly or a cartel. The finding of an abuse was to be decided by the Minister for economic affairs. Germany had experienced a control of abuse under the Verordnung gegen den Mifibrauch wirtschaftlicher Machtstellung, dated 2 November 1923 (RGBL I.S. 1067), which turned out to be ineffective: Moschel, Le controle des prix d'apres le droit allemand et europeen sur les restrictions de la Concurrence, Revue du Marche Commun, 1976, p. 263. 3 Mestmacker E.-J. (1999), 523.

1 - Compatibility, Efficiency, Legal Security

147

at no risk.4 According to him the proposed system corresponds to a general exemption that is valid until the adoption of a decision finding an infringement.5 This system—he concludes—would be incompatible with the direct applicability of Article 81 paragraph I.6 Mestmacker is certainly right in his view that a system of abuse control would run counter to the direct applicability of Article 81 paragraph 1 and, as will be seen below, the documents on the negotiations of the treaty confirm this view.7 The point, however, is that this is not the system proposed by the Commission. Let us read from the White Paper at point 11: 'Under a directly applicable exemption system, the prohibition on restrictive practices is not applicable to those which satisfy certain conditions denned by law. Taken as a whole, such conditions may be regarded as an exemption to the prohibition principle. Restrictive practices which satisfy the conditions are therefore valid as soon as they have been concluded.'

It results from this description that the prohibition applies unconditionally to restrictive practices that do not satisfy the conditions: in particular these practices are invalid as soon as they have been concluded. The misunderstanding about the system put forward by the White Paper probably derives from the use of the expression 'ex post control', whereby the Commission refers to the difference between the present system (in which firms can—without being obliged to— notify the Commission of an agreement before its implementation), and the proposed system (in which the enforcement authority would normally discover a cartel later on). The fact that the infringement would be discovered after its implementation—which incidentally is already the case now in most instances—must

4 'Das jetz wiederbelebte Theorie des kartellverbots mit Legalausnahme bedeutet: Alle Kartelle sind von Anfang an rechtswirksam, bis ihre Unvereinbarkeit mit Art. 81 durch eine Entscheidung der Kommission, der staatlichen Kartellbehorden oder der staatlichen gerichte festgestellt ist. Es gibt bei dieser Auslegung von Art. 81 keine Vermutung der Niehtigkeit von Kartellen. Das Verbot greift nur ein, wenn zusatzlich zu einer Wettbewerbsbeschrankenden Vereinbarung festgestellt ist, dass sie nicht zu den positiven wirtschaftlichen Vorteilen beitragt, die nach Art. 81 HI Ausnahmen vom Verbot rechtfertigen. Der Grundsatz der Rechtswirksamkeit bis zu einer den VerstoB feststellenden Entscheidung soil unabhangig von der Art und der Intensitat der Wettbewerbsbeschrankung gelten. Alle kartelle, die in Art. 81 I als Beispiele fur verbotene Wettbewerbsbeschrankungen gennant werden, diirfen zunachst risikolos praktiziert werden': p. 525. 5 'Die Auslegung als Verbot mit Legalausnahme fiihrt aber zur einen allgemeinen Freistellung bis zum ErlaB einer den VerstoB feststellenden Entscheidung': p. 525. In the same sense Monopolkommission (1999, point 16): 'Die Moglichkeit, samtliche Vereinbarungen pauschal zu erlauben, ist dagegen nicht vorgesehen.' 6 'Sollten die Gerichte der Mitgliedstaaten durch Verordnung verpflichtet werden, Art. 81 III anzuwenden, so wiirde damit die unmittelbare Anwendbarkeit von Art. 811 im Widerspruch zur Rechtsprechung des EuGH aufgehoben. Eine solche Regelung ware wegen Verletzung des EG-Vertrages auf Klage fur nichtig zu erklaren': p. 517. 7 Cf. also Hoeren T. (1999), 405.

148

Working Paper VI

not be confused with the legality of the cartel until such discovery.8 The expression 'ex post control' appears for instance at point 77 of the White Paper: 'In a system of ex post control, undertakings would have to make their own assessment of the compatibility of their restrictive practices with Community law, in the light of the legislation in force and the case-law; this would certainly lighten the administrative burden weighing on them, but it would also require them to take on added responsibility'.

The added responsibility referred to here is just the opposite of the 'no risk' situation envisaged by Mestmacker. In fact, a legal exception system implies a more stringent direct effect of the prohibition compared to the present system, where authorities can be notified of a restrictive agreement and the parties to the agreement thus enjoy immunity from fines pending its examination by the Commission. Furthermore in the present system the direct applicability of Article 81 paragraph 1 before national judges and competition authorities is more illusory than real since the Commission's exemption monopoly produces a chilling effect on the propensity of those judges and authorities to take or decide a case on the basis of paragraph 1 only. This is indeed one of the major reasons underlying the White Paper proposal. It is only by allowing national judges and authorities to apply Article 81 in its entirety that it can be hoped to associate them in a much needed effort of effectively enforcing Article 81. Far from being generally exempted until a decision finding them illegal, agreements would incur the risk of being found in breach of Article 81 from the beginning.9 Hopefully, this clarification of the Commission's intentions will remove the concerns based on a supposed equivalence of the proposed system with a system of abuse control. This should cover also the implied criticism expressed by the Monopolkommission 1999, which acknowledged that the proposed system cannot be assimilated to an abuse control system where agreements are legal until a constitutive decision makes them illegal but goes on to say that one can speak of an approximation to de facto abuse control in so far as the necessity of a prior exemption decision is eliminated.10 It must be remembered that a prior 8

The terminology 'ex post control' 'ex ante control' was already used in the 1961 Deringer Report in the same sense as in the White Paper. Cf. Deringer Report, 1961, points 13-15, where the three systems (abuse control, prohibition with legal exemption, prohibition with possibility of authorisation) are extensively described and analysed. 9 In the same sense, see Schaub and Dohms (1999), 1066. 10 'Ein solches System kann nicht mit einem MiBbrauchsprinzip gleichgesetzt werden. MiBbrauchsprinzip in seiner reinen Form bedeutet Zulassigkeit einer Wettbewerbsbeschrankung bis zu einer konstitutiven Feststellung ihrer Unzulassigkeit. Der Vorschlag der Europaischen Kommission geht dagegen formal unverandert von einem Verbotsprinzip aus. Man kann lediglich insoweit von einer Annaherung an ein faktisches MiBbrauchsprinzip sprechen, als die Notwendigkeit praventiver Erlaubnisentscheidungen generell entfallt': point 6. Similarly Deringer (2000, 5) on the one hand admits that the White Paper implicitly assumes that Art. 81 enshrines a prohibition system and not an abuse control system (p. 5), on the other states in the

1 - Compatibility, Efficiency, Legal Security

149

exemption is not necessary under the present system. It is up to the firms to decide whether they want to notify the Commission of an agreement. Moreover, even if they do, they are not prevented from implementing the agreement without waiting for the Commission's decision. In other words, firms may implement their agreements at their own risk. The proposed system would alter this situation in that the firms would have no choice other than assuming their own risks, and not because prior exemption decisions will no longer be necessary. The German government echoes the above-mentioned concerns when it states that Article 81 is based on the principle of prohibition with the possibility of exempting certain clearly circumscribed cases, and that it does not provide for the possibility of exempting (by means of a single legislative instrument working in practice as an overall block exemption regulation) agreements which meet the requirements.'' It should be clear by now that the Regulation contemplated by the White Paper would contain no assessment as to which agreements meet the requirements for escaping the prohibition; it could not, therefore, be considered an exemption regulation. It would merely require all authorities and judges dealing with the conformity of an agreement with Article 81 to apply paragraphs 1 and 3 at the same time. The assessment as to the meeting of the conditions prescribed by paragraph 3 would thus be made—with declaratory {ex tune) effect—by a national judge or authority, or by the Commission for an individual agreement, and not by the Council regulation for all agreements. In the hope that these observations will be sufficient to answer the arguments based on the direct applicability of the prohibition contained in paragraph 1, let us now examine those based on paragraph 3. Here again a misunderstanding must first be dispelled.

III. The Commission does not claim that the proposed system is the only permissible one Some critics insist that paragraph 3 of Article 81 does not provide for a legal exception system as the only possible system. Nobody would deny such a statement, least of all the Commission. It would indeed be strange for the Commission to adopt a position implying the illegality of Regulation 17 (a conclusion without an explanation that there is a danger that Art. 81 would in practice evolve as an abuse control provision (p. 11). 11 'Die Formulierung des Art. 81 abs. 3 EGV beruht auf dem Prinzip eines Verbots mit bestimmten begrenzten Ausnahmemoglichkeiten und spricht ebenfalls gegen die Moglichkeit, durch eine einzige gesetzgeberische MaBnahme quasi im Wege einer "Total-Gruppenfreistellungsverordnung" samtliche Absprachen freizustellen, die die Voraussetzungen erfullen'.

150

Working Paper VI

provision that it proposed in the early sixties and has applied for almost 40 years). In fact, according to the White Paper (point 12): 'Article 85 (3) allows the Community legislator the freedom to choose between an authorisation system and a directly applicable exemption system.' In other words, the drafting of Article 81 (3), according to the Commission, does not describe a legal exception system, it is instead broad enough to allow the Council to choose such a system when legislating under Article 83. This position of the Commission answers the argument12 that, if the authors of the Treaty had meant to provide for a legal exception system, they would have used different wording in Article 81 paragraph 3 (such as 'The provisions of paragraph 1 do not apply in the case of...', instead of 'may be declared inapplicable'). That is correct. But it is necessary to escape the frame of mind that prevailed during the discussions about the future Regulation 17 which turned on which system was prescribed in the Treaty itself, and where each opposing thesis claimed that the Treaty imposed its favoured solution. Nobody seemed to discuss whether the Treaty could have left the question open, probably because each of the two opposing factions hoped to win the battle outright on legal grounds. If it cannot be claimed that the chosen wording imposes a system of legal exemption, it remains to be seen whether it allows such a system.

IV. Paragraph 3 of Article 81 requires implementing legislation in order to be operative Although nobody contests this proposition, it is useful to stress it from the outset. Paragraph 3 will not be immediately operable. This is clear from the silence of the text as to who is empowered to declare inapplicable the prohibition contained in paragraph 1, and from the express provision in Article 83 (former 87) that the Council 'adopt any appropriate regulations or directives to give effect to the principles set out in Articles 81 and 82 . . . in particular . . . to lay down detailed rules for the application of Article 81 (3), taking into account the need to ensure effective supervision on the one hand, and to simplify administration to the greatest possible extent on the other'. The delegation to the Council and the words used tend to show that the authors of the Treaty had not resolved the problem of how the exceptions from the general prohibition should be administered. The existence of this delegation and the way it is formulated shed a priori strong doubts on a reading of paragraph 3 limiting the choice of the legislator, let alone imposing on him a particular method for applying those exemptions. The Treaty of Rome has been dubbed a traite-cadre, as opposed to 12 See Deringer (2000, 5) recalling the wording of the French legislation enshrining the legal exemption.

1 - Compatibility, Efficiency, Legal Security

151

a traite-loi. It contains the principles, not the rules of detail. The Council is a real legislator, to whom the authors of the Treaty have entrusted the elaboration of entire policies (such as the agricultural policy). These remarks should also put into perspective an argument put forward by Deringer.13 He writes that, since the equivalent provision of the Coal and Steel Treaty contains (Article 65 paragraph 2) the word authorise relating to the High Authority, the authors of the Treaty of Rome should have made it unequivocal that they wanted to introduce, or even simply not exclude, the introduction of a different system. In fact, the Coal and Steel Treaty (a traite-loi) contains an exhaustive regulation whereby the High Authority is given the power to authorise certain agreements. There is no delegation to lay down detailed rules, nor is there a need for it. On this point, any comparison between the two treaties is stretched. If anything, the comparison could easily turn against the supporters of the authorisation theory, for the choice of substantially different wording in the subsequent treaty can be said to demonstrate an intention to follow a different approach.14

V. Does the wording of paragraph 3 of Article 81 imply an administrative decision? Notwithstanding the presumption that the authors of the Treaty have left the procedural problem of the administration of paragraph 3 squarely to the legislator, it cannot be excluded that they might have decided to limit the choice of the Council. The strongest argument used to assert the incompatibility of the proposed system with the Treaty is precisely that the wording 'The provisions of paragraph 1 may, however, be declared inapplicable in the case of . . .' necessarily implies a decision by an administrative authority. In particular, the word declared is said to exclude the possibility that the exemption be administered by the judiciary.

1. Historical arguments The White Paper states (point 12) that paragraph 3 ' . . . is the result of a compromise between the delegations favouring a directly applicable exception system and those favouring a prior authorisation system. Whilst those in favour of an authorisation system proposed wording along the lines of "restrictive 13 'Wenn die Vertragspartner bei der Formulierung des EWG-Vertrages in Art. 85 HI ein anderes System einfuhren oder auch nur zulassen wollten, hatten sie das eindeutig anders formulieren miissen': p. 6. Cf. also Ehlermann (2000, 21, point 50). 14 In the same sense, see Schaub and Dohms (1996), 1066.

152

Working Paper VI

agreements may be declared valid", agreement was eventually reached on a negative wording: "the provisions of paragraph 1 may, however, be declared inapplicable". By opting for this negative approach, Art. 85 (3) allows the Community legislator the freedom to choose between an authorisation system and a directly applicable exception system.' In 1963, Deringer wrote in the same sense: 'Le rapport dit de Messine se limite au principe general de l'interdiction des conventions en vue de la limitation de la concurrence, sans aborder la question de savoir si, et dans quelle mesure, des exceptions peuvent etre autorisees. Au cours des negociations en vue du traite, le groupe de travail competent, la Delegation Francaise et la Delegation allemande ont done pris comme point de depart les reglements tels qu'ils existent dans leur legislation nationale. La Delegation Francaise a propose que les situations entrant dans le cadre des exceptions a prevoir devraient etre licites en vertu de la loi, tandis que la Delegation Allemande cherchait a obtenir une procedure d'autorisation suivant l'esprit de la loi allemande. Comme aucun accord n'a pu etre conclu, et que des propositions n'ont pas trouve de majorite, on a prefere choisir la formule suivante: «peuvent etre declares valables ...». Cette formule, suivant le desir de la Delegation Francaise, a ete integree dans le texte definitif du traite sous une forme negative: «Les dispositions du paragraphe I peuvent etre declarees comme n'etant pas susceptibles d'application».'15 In his recent article on the White Paper,16 the same author writes that the documents now put at his disposal by Mr von der Groeben (chairman of the Groupe du marche commun, which prepared the text of the EEC Treaty at the Val Duchesse conference), do not disclose either that the problem was expressly discussed or that the final wording was meant to leave to the legislator the choice between the two solutions favoured respectively by the French and the Germans. According to him, it is only during the work of the European Parliament on what would become Regulation 17—Deringer was at the time an MEP and authored the report of the relevant committee—that the issue was raised and that he proposed the compromise situation of making the exemption delivered by the Commission possibly retroactive.17

15 Deringer (1963, 257). Also the Monopolkommission (1999, point 15) admits that the introductory phrase of paragraph 3 is a drafting compromise, but it still considers the system proposed in the White Paper as incompatible with this phrase ('Die Gesetzesformulierung "konnen fur nicht anwendbar erklart werden" waren ein FormelkompromiB zwischen den Anhangern eines Genehmigungssystem und den Anhangern einer Legalausnahme. Dem Wesen eines Formelkompromisses entsprechend meinte jede Seite, sich in der beschlossenen Fassung wiederzuerkennen. Die Anh anger eines Genehmigungssystems hatten zunachst in Anlehnung an Art. 65 EGKS-Vertrag die Formulierung :konnen fur giiltig erklart werden" befurwortet. Der Ubergang von einer positiven Fassung zu einer negativen Fassung andere nichts an der Substanz, daB es einer "Erklarung" bedurfte.'). 16 Deringer (2000, 5). 17 P. 6. In the Deringer Report, 1961, the discussion between the two interpretations of the Treaty takes place in points 59-64.

1 - Compatibility, Efficiency, Legal Security

153

I have consulted the documents relating to the negotiation works on the future Article 85 that are available from the historical archives of the Council. The date, number, title and summary of the content (limited to what might be relevant to the problem discussed here) of the documents I have found are reported in Appendix I. The documents contain several drafts of the competition rules but few descriptions or explanations of the positions of the various delegations. In their first phase (June to August 1956), the drafts contain a single article declaring a number of practices and situations incompatible, including restrictive agreements and abuses of monopoly described in general terms. At this stage there is no trace of the future paragraph 3 of Article 85. In a second phase (September to beginning of November 1956), the positions of the various delegations become clearer. The German delegation insists on the prohibition principle under which some exemptions should be admitted but subjected to an authorisation system ('Toutefois, lorsque la demande en est faite, les accords ou decisions de ce genre peuvent etre autorises si le requerant est en mesure de prouver . . .'). 1 8 The Belgians and Dutch favour a 'control of abuse' system. The position of the French delegation emerges from the draft they support rather than from an expressed view. The prohibition is rendered by the word incompatible, while the exemption is introduced by the words 'Peuvent etre relevees des dispositions precedentes les situations ou pratiques dont les auteurs sont en mesure de justifier qu'elles contribuent...'. The French thus seem to agree with the Germans on the prohibition principle, although they do not use the word "authorisation" for the exceptions. We enter into a third (hot) phase with the text prepared by a group of experts between 7 and 12 November 1956. The text of paragraph 1 reflects the prohibition principle but, according to paragraph 2, agreements declared to the Commission are presumed to be legal unless they are prohibited. This sounds like subjecting the exceptions to a control of abuse process, but since all agreements can be 'declared' and thus enjoy a presumption of legality, it is closer to a generalised control of abuse. Such a draft cannot have pleased the Germans. 19 Mr von der Groeben, Chairman of the Groupe du marche commun, proposes retaining paragraph 1 only, while the provision on procedure (the future Article 18 There is in fact a discrepancy between the position taken by the German delegation in the Groupe du marche commun on 3-5 September, according to which the exemptions would be submitted to a control of abuse (MAE 252/56), position which is still taken by MAE 325/56 as describing the German proposed system and the draft actually proposed by the Germans, as reflected in the synoptic table contained in the same MAE 325/56, quoted in the text, whereby the exemptions need an authorisation. To accept that the exemptions are submitted to a control of abuse inevitably transforms the whole system in a control of abuse system, since a notified agreement must enjoy a presumption of legality. 19 Although they seem to be at least partly to blame for this draft, given the discrepancy, shown in the previous footnote, between what they actually wanted and their description of it.

154

Working Paper VI

87) would leave it to the implementing texts to rule on the exceptions and on the question whether the agreements falling under the exceptions should be declared to the Commission for registration or authorised by the Commission. At the meeting of the Groupe du marche commun on 13, 14 and 15 November, the French say that they can accept the text of the expert group if it is supplemented by a provision on procedure (the future Article 87) specifying that the implementing regulations should determine the conditions for permitting the agreements fulfilling the conditions for derogation, and taking into account the need to ensure effective supervision on the one hand and to simplify administration to the greatest possible extent on the other. The German delegation, on the contrary, voices its dissatisfaction in relation to the text of the expert group and says that—rather than accepting a text that does not reflect the prohibition principle—it prefers a general declaration against restrictive agreements leaving the Member States free to implement it in their own way. This threat prompts the Belgian and the Dutch delegations to review their positions. The next draft of the expert group (20 November) contains a virtually definitive paragraph 1, while paragraph 2 is introduced by the words 'Seront releves des interdictions precedentes les accords ...'. This is the original French formula in a more automatic format {seront releves instead of peuvent etre releves). Paragraph 3 reads: 'Les accords ou decisions interdits en vertu du paragraphe precedent [sic] sont nuls de plein droit et ne peuvent etre invoques devant aucune juridiction des Etats membres'. At this stage it looks as if the German insistence on the prohibition principle prevailed, but under the French variant of the legal exception. On 28 november 1956 the Groups du marche commun approves the text of the expert group with two changes that are highly significant for the present discussion. Paragraph 2 now starts with: 'Peuvent etre declares valables les accords ...'. In paragraph 3, paragraphe precedent has been replaced with paragraphe 1 du present article. Both changes move away from a legal exception system and towards an authorisation system. While the first change triggers a question about who can declare the validity (and thus weakens the possibility of interpreting paragraph 2 as a directly applicable provision), the second one reinforces this impression by seeming to prevent the national courts from applying the exceptions. It is now the turn of the prohibition principle under the German variant (authorisation system) to seem to triumph. The future article 87 is accepted at the same time, substantially in its final form. The next stage is the Comite des Chefs de delegation. At its meeting of 6 December 1956, one of the two changes decided by the Groupe du marche commun is undone, thus swinging the pendulum half-way back towards the legal exception option: the paragraph on nullity, which is still paragraph 3, is modified in the sense that instead of en vertu du paragraphe 1 du present article it should read en vertu du present article, the Groupe de redaction being at the same time requested to advise on the question of whether the words sont nuls de plein droit can be suppressed without weakening the prescription of the paragraph. Here ends the third phase.

1 - Compatibility, Efficiency, Legal Security

155

The next document is dated almost two months later. We are in the last phase, which starts on 14 February, a little over a month before the signing of the Treaty. The documents do not disclose any explanation for the further changes, which are all relevant to our problem. The Groupe de redaction is presented with a draft proposed by the rapporteurs du groupe marche commun. The beginning of paragraph 1 now reads: 'Sont incompatibles avec le marche commun, interdits et nuls de plein droit...'. The following sentence is added in square brackets at the end of paragraph 1: '[La nullite de plein droit visee ci-dessus doit etre constatee d'office par les juridictions des Etats membres.]'. Paragraph 2 reads: 'Toutefois, les dispositions du paragraphe 1 du present article peuvent etre declarees non applicables a ...'. There is no longer a paragraph 3. Insofar as the nullity and its finding by the courts solely concern paragraph 1, this is a move back towards an authorisation system. As to the exceptions, the text does no longer say that the agreements can be declared valid but that the provisions of paragraph 1 may be declared inapplicable. With the following document we are back to a three-paragraph structure but paragraphs 2 and 3 have been inverted. Paragraph 2 on nullity now refers to the whole article. It is drafted as in the final form, the words '. . . et ne peuvent etre invoques devant aucune juridiction des Etats membres' being added in square brackets. The file ends here. Deringer is right: the documents contain no express discussion about how the exceptions to the prohibition of restrictive agreements should be administered. Nor do they contain an express confirmation that the final wording was meant to leave to the legislator the choice between the solutions favoured by the French and the Germans respectively. It emerges, however, that while the rest of the future Article 85 remains virtually the same as in the final text, from 20 November 1956 the phrase which introduces the exemption undergoes several changes: —'Seront relevees des dispositions precedentes les situations ou pratiques dont les auteurs sont en mesure de justifier qu'elles contribuent...' (20 November 1956). —'Peuvent etre declares valables les accords . . .' (10 December 1956) —'Toutefois, les dispositions du paragraphe 1 peuvent etre declarees non applicables . . .' (14 February 1957) —'Toutefois, les dispositions du paragraphe 1 peuvent etre declarees inapplicables . . .' (25 February 1957 or a little later). It is difficult to doubt that a battle was being fought around this formula. Such a conclusion is all the more compelling if at the same time we observe that the nullity provision, which appears for the first time on 12 November 1956, refers to —paragraph 1 on 12 November 1956; —the previous paragraph (the one on exceptions!) on 20 November; —paragraph 1 on 28 November; —the whole article on 6 December;

156

Working Paper VI

—paragraph 1 on 14 February 1957; —the whole article on 25 February. How else can we explain this extraordinary ballet other than in terms of a tug of war between those who wanted to make it clear that the paragraph on prohibition and the one on the exceptions were subjected to different regimes (direct applicability for the former and authorisation for the latter, and those who favoured direct applicability for both)? The place of the provision on nullity also moves. Having remained at the end of the Article from the day of its inception (12 November 1956) until 14 February 1957, it is then inserted between the prohibition and the exceptions. This change is all the more strange in view of the fact that the provision reads: 'Any agreements or decisions prohibited pursuant to this Article shall be automatically void'. Its correct place was therefore at the end and not in the middle of the Article. Let us notice, incidentally, that the question put by the Chefs de delegation to the Groupe de redaction on 6 December on whether the words 'sont nuls de plein droif could be eliminated without weakening the prescription, has apparently been answered by eliminating not the words 'sont nuls de plein droif but the words 'et ne peuvent etre invoques devant aucune juridiction des Etats membres'.

Finally, the documents certify that, as from mid-November 1956, the idea that the differing views as to the way the exemptions should work could be left to the Council to dispose of was aired both by Mr von der Groeben, chairman of the Groupe du marche commun, and by the French delegation. This delegation even put forward the draft of what has become Article 87 (now 83) paragraph 2 under b. These findings make it difficult, if not impossible, to deny the correctness of the reconstruction of a conflict between the French and the Germans made by Deringer in 1963 when the memory of the negotiations was livelier. The location of the provision on nullity in the middle of the Article was no doubt meant to advance the German thesis that the prohibition only was directly applicable, while the exceptions should have been in the hands of the administration.20 However the Germans scored only half a point, or less, since the provision still refers to the whole Article. On this reconstruction the French obtained a much vaguer formula for the introduction of the exemptions that, together with the delegation to the Council, would (in their mind) have left thefinalchoice of the system to the legislator. In this regard it can be noted that referring to the inapplicability of the provisions of paragraph 1 to certain agreements, rather than dealing directly with the fate of the latter, reinforces the reading that the addressee of this text is the legislator rather than the authority dealing with a concrete case or series of cases. It might be argued that this is a mere reconstruction. However, even if the changes underlined above were disconnected from a difference of positions as to the choice between an authorisation system and a legal exception system, the 20

Cf. Ehlermann (2000, 22, point 52).

1 - Compatibility, Efficiency, Legal Security

157

theory that a legal exception system is contrary to the Treaty would not thereby be advanced. On the contrary, such a disconnection would point in the direction of the freedom of the legislator in this regard. The incompatibility theory would only be advanced on historical grounds if the negotiations showed that there was a discussion about the final system and that the preference was given to the authorisation system. This reconstruction is the only one that is certainly excluded by the documents. The draft that most closely reflects an authorisation system is the one accepted by the Groupe du marche commun on 28 November 1956, where the exceptions are introduced by 'Peuvent etre declares valables les accords...' ('The agreements may be declared valid . . .') and where the nullity refers to the agreements prohibited in paragraph 1. However, the latter formula was modified by the Chefs de delegation on 6 December, and the former was modified at the stage of the Groupe de redaction. These changes tend to show that the negotiators purposely avoided enshrining an authorisation system in the Treaty. In sum, the history of the drafting of Article 85 (now 81) can only be used to prove an intention that the legislator be free to choose between the two systems of implementation of the prohibition principle.

2. Drafting arguments According to Deringer,21 even if the text of Article 81 had been approved by the French delegation with the intention of leaving it open to the legislator to choose between the two systems, that intention has not crystallised in the final wording. In support of this conclusion he cites Advocate General Lagrange in his opinion in the Bosch case.22 This was the first reference from a national judge under what used to be numbered Article 177 (now 234) and, at the same time, the first EEC competition case to reach the Court. It turned on the application of Article 85 to a restrictive agreement entered into before the coming into effect of Regulation 17. The Lagrange opinion is generally very interesting for our subject because it discusses the competing theories of 'legal exception' and 'administrative authorisation' in a simple and clear way. That is why I think it is appropriate to quote it extensively, even if such a long quotation constitutes a digression in the discussion on the compatibility of the legal exception with the Treaty. The reader who does not want to be diverted can jump to the passage in bold characters and then skip the rest of the quotation. The passage in bold characters is the one on which Deringer relies, extended to the immediately following sentence which I think is no less relevant. I have chosen to reproduce the quotation in the original French, which is far superior in clarity and elegance to the English translation. (This translation can be found in Appendix II.) 21 22

Deringer (2000, 5 and 6). Cf. also House of Lords (2000, 27, point 104). Case 13/61 de Geus v Bosch [1962] E C R 111.

158

Working Paper VI

'Nous touchons ici au defaut essentiel du regime institue par les articles 85 et suivants du traite, c'est-a-dire l'inadaptation du systeme de controle a la legislation materielle. D'une part, en effet, cette legislation materielle repose sur l'existence d'un lien de connexite evident entre les regies definissant la nature des accords interdits, figurant a l'article 85, paragraphe 1, et celles qui definissent les conditions permettant le relevemerit de l'interdiction figurant a l'article 85, paragraphe 3: il suffit de lire les deux derniers alineas, a et b, du paragraphe 3 pour s'en rendre compte. Les memes autorites ou les memes juridictions, au cours de la meme procedure, devraient logiquement etre competentes pour statuer a propos du meme accord aussi bien sur la compatibility de cet accord avec les dispositions du paragraphe 1 que sur la "declaration d'inapplicabilite" de l'interdiction sur la base du paragraphe 3. D'ailleurs, les termes memes de l'article 88 declarant la legislation nationale applicable pendant la periode transitoire concurremment avec la legislation du traite contenue aux articles 85 et 86 demontrent bien l'existence de ce lien de connexite et la conscience qu'en avaient les auteurs du traite, puisque le texte insiste sur la necessite de tenir compte, dans cette application concurrente, notamment du paragraphe 3 de l'article 85. Au surplus, il eut ete contraire a l'equite la plus elementaire de permettre l'application de l'interdiction edictee au paragraphe 1, avec la sanction de la nullite de plein droit qui s'y attache et toutes les consequences que les tribunaux auraient pu et meme du en tirer, sans qu'il fut possible aux entreprises de se prevaloir eventuellement des dispositions du paragraphe 3. C'est pourquoi la theorie dite "de l'exception legale", qui sert de fondement au systeme francais, se trouvait fort bien adaptee aux necessites de l'application du traite deja pendant la periode transitoire. Cette theorie, en effet, comporte l'examen simultane par la meme autorite, et eventuellement la meme juridiction, du principe de l'interdiction et du relevement eventuel de celle-ci: l'application des effets de la nullite de plein droit ne souffre pas alors de difficultes. C'est pourquoi aussi cette meme theorie etait, a notre sens, la seule qui fut susceptible de justifier l'application immediate de l'article 85 dans les pays ne possedant encore aucune legislation anti-cartel: la juridiction de droit commun ayait alors naturellement competence pour l'application aussi bien du paragraphe 3 que du paragraphe 1. Dans la these contraire, celle qui exige une decision ayant effet constitutif pour l'application du paragraphe 3, theorie qui a ete defendue surtout en Allemagne et a toujours ete soutenue par la Commission, il est evident que l'application immediate de l'article 85 devenait impossible tant qu'une autorite nationale qualifiee n'etait pas habilitee a prendre les decisions requises, selon cette theorie, pour appliquer le paragraphe 3. C'est d'ailleurs ce qui reconnait expressement le gouvernement allemand dans son memoire au paragraphe IV: "Conformement a l'article 88 du traite, l'application de l'article 85, paragraphe 3, est attribute provisoirement aux autorites des Etats membres de la Communaute economique europeenne. Ce faisant, c'est le droit interne de chaque Etat qui determine quelles sont les autorites competentes pour cela." II appartenait done aux autorites des Etats depourvus de legislation a cet egard d'edicter les prescriptions necessaires, mais, lorsque cella n'a pas ete fait, il ne pouvait plus etre question, toujours selon cette theorie, d'appliquer l'article 85 pendant la periode transitoire dans ces Etats, l'application du paragraphe 1 ne pouvant, encore une fois, se concevoir sans une application correlative eventuelle du paragraphe 3.

1 - Compatibility, Efficiency, Legal Security

159

Certains ont meme soutenu que l'impossibilite d'appliquer le traite dans certains des Etats membres, faute d'une reglementation appropriee dans ces Etats, s'opposait a cette application dans toute la Communaute: ceci allait trop loin a notre avis. II etait conforme a l'esprit du traite que celui-ci commencat a entrer en application la oil la chose etait possible, en meme temps que la legislation interne. Actuellement, la controverse est close, puisque le reglement a consacre en droit la deuxieme these, dont il faut d'ailleurs reconnaitre qu'elle paraissait plus compatible que l'autre avec la redaction du paragraphs 3: "Toutefois les dispositions du paragraphe 1 peuvent etre declarees inapplicables a tout accord, etc." La these de Pexception legale eut exige une formulation differente telle que: "Les dispositions du paragraphe 1 sont reconnues inapplicables ou tout simplement "ne sont pas applicables". II ne peut etre question, dans ces conditions, de soutenir que le reglement serait illegal comme contraire au traite a cet egard: soulever l'illegalite d'un tel reglement serait si grave que la Cour ne devrait, a notre avis, le faire que si cette illegalite etait certaine, ce qui n'est nullement le cas, loin de la.

Mais alors, la contradiction que nous avons relevee subsiste, de fait que, contrairement a ce qui en est dans le traite C.E.C.A. (article 65), la competence exclusive n'est attribute que pour le relevement d'interdiction et non pour la constatation d'incompatibilite entramant la nullite de plein droit. C'est pourquoi le reglement a edicte toute une serie de dispositions destinees a assurer une application aussi equilibree que possible a l'article 85 dans son ensemble, et aussi de l'article 86. La piece maitresse du systeme est l'obligation pour les entreprises qui entendent obtenir une "declaration d'inapplicabilite" au titre de l'article 85, paragraphe 3, de notifier leurs accords a la Commission qui peut faire retroagir les effets d'une decision favorable jusqu'a la date de la notification (art. 4, § 1, et 6, § 1). Cette "notification" ressemble d'ailleurs fort a une demande d'autorisation, et la "declaration d'inapplicabilite" a une autorisation ("les dispositions du paragraphe 1 peuvent etre declarees inapplicables"): malgre des termes differents, le systeme est visiblement inspire de l'article 65, paragraphe 2, du traite C.E.C.A. Dans ce systeme (nous ne parlons jusqu'ici que du regime definitif, concernant les ententes conclues apres l'entree en vigueur du reglement, et du regime de droit commun, car il existe, vous le savez, un regime transitoire et il existe aussi un regime de faveur plus souple pour certaines categories d'ententes), dans ce systeme done, l'equilibre parait correctement assure. La nullite d'un accord contraire a l'article 85, paragraphe 1, peut toujours etre invoquee devant les tribunaux nationaux et, meme si une "notification" a lieu en cours de procedure, le tribunal peut statuer et tirer eventuellement les consequences de la nullite de plein droit s'attachant a un accord qu'il reconnaitrait contraire a l'article 85, paragraphe 1, au moins pour ce qui concerne la periode anterieure a la notification, puisqu'il est certain qu'une "declaration d'inapplicabilite" ulterieure ne pourra pas, en tout cas, retroagir au dela. En revanche, ce meme tribunal serait bien avise d'attendre le resultat de la procedure devant la Commission pour faire sortir les effets de la nullite de plein droit posterieurement a la notification. Y est-il juridiquement oblige? Cela ne nous parait pas possible a admettre en l'absence de toute disposition formelle en ce sens. En particulier, les dispositions de l'article 9, paragraphe 3 du reglement, que nous avons deja citees, ne le permettent pas. Au surplus, il peut paraitre opportun de s'en remettre dans un cas de ce genre a la sagesse des juges, car il peut se presenter des hypotheses ou il apparaitrait avec evidence que les dispositions du paragraphe 3 ne sont pas applicables et que

160

Working Paper VI

la "notification", effectuee apres la decouverte des faits et l'action portee devant le tribunal, n'a qu'un caractere dilatoire. Au point de vue juridique, le veritable inconvenient proviendrait plutot du cas ou le tribunal admettrait l'incompatibilite d'un accord avec les dispositions de l'article 85, paragraphe 1, tandis que la Commission (ou eventuellement la Cour), statuant ulterieurement, aurait sur ce point une opinion contraire et estimerait, en consequence, comme sans objet, et done impossible, une "declaration d'inapplicabilite" au titre du paragraphe 3; mais e'est la une consequence inevitable de l'existence des competentes concurrentes. En ce qui concerne les ententes beneficiant du "regime de faveur" prevu a l'article 4, paragraphs 2, une prudence particuliere du juge national sera egalement souhaitable au cas ou ces ententes ont ete notifees a la Commission: en effet, pour ces ententes, une declaration d'inapplicabilite de la Commission, au titre de l'article 85, paragraphe 3, peut retroagir a une date librementfixeepar la Commission et qui peut etre anterieure a la notification (art. 6, § 2, du reglement). Reste, enfin, le cas des ententes existantes a la date d'entree en vigueur du reglement, qui fait l'objet de dispositions speciales aux articles 5 et 7 de ce reglement. Pour ces ententes, des lors qu'elles ont ete notifiees avant le ler aout 1962 (ou le ler Janvier 1964 pour les ententes beneficiant du regime de faveur), "l'interdiction edictee a l'article 85, paragraphe 1, ne s'applique que pour la periodefixeepar la Commission", des lors que les entreprises mettent fin aux accords "ou les modifient de telle sorte qu'ils ne tombent plus sous l'interdiction edictee par l'article 85, paragraphe 1, ou qu'ils remplissent les conditions d'application de l'application de l'article 85, paragraphe 3". Ceci veut dire que la decision de la Commission pourra comporter l'absence de tout effet retroactif pour l'interdiction, e'est-a-dire le maintien integral des effets de l'accord dans le passe, voire dans l'avenir, si des delais sont accordes pour permettre aux interesses de se mettre en regie: en pareil cas, les effets de la nullite de plein droit, laquelle a par nature un effet retroactif (ex tune), se trouvent entierement paralyses. Ici encore, aucune disposition du traite ni du reglement n'oblige le tribunal a surseoir, meme si la notification a deja eu lieu ou est faite en cours de procedure, mais il va de soi qu'en pareil cas les tribunaux nationaux doivent etre amenes a faire preuve de la plus grande prudence etant donnees les consequences que peut avoir une decision ulterieure favorable de la Commission. Cependant, les juridictions nationales devront tenir compte, le cas echeant, de la disposition suivante du reglement (art. 7, § 1, in fine): "Une decision de la Commission en application de la phrase precedente (qui lui donne le pouvoir de fixer la periode d'application de l'interdiction) ne peut etre opposee aux entreprises et associations d'entreprises qui n'ont pas donne leur accord expres a la notification." Voila done, Messieurs, comment a notre sens doit s'articuler la competence concurrente des juridictions nationales et de la Commission pour l'application de l'article 85, paragraphes 1 et 2, et la competence exclusive de la Commission pour l'application de l'article 85, paragraphe 3. Sans doute, le resultat n'est pas pleinement satisfaisant, mais e'est la la consequence du double compromis juridique qui est a la base du traite, interprets comme nous l'avons dit: 1° compromis entre la these de "l'exception legale" seule pleinement compatible avec la notion de nullite de plein droit contenue a l'article 85, paragraphe 2, et la these de "l'effet constitutif" qui logiquement devrait s'accompagner, comme dans le regime allemand, d'un pouvoir donne aux autorites de

1 - Compatibility, Efficiency, Legal Security

161

cartel de "declarer inefficaces" (fur unwirksam erklaren") les accords contraires a la loi, ce qui correspond a une notion bien differente de la nullite de plein droit; 2° compromis sur la competence, que le traite ne regie pas et que le reglement n'a pas voulu, ou a estime n'avoir pas pu confier exclusivement aux autorites et a la juridiction communautaires, comme dans le traite C.E.C.A.' (emphasis added) Let us now examine closely what the Advocate General has said in the passage reproduced in bold. The last sentence of this passage shows that the defects of the system adopted in Regulation 17 were, in his mind, so profound that he wondered whether the regulation could have been incompatible with the Treaty and whether the Court should raise this question ex officio. At the same time he rejected this idea because raising this question would have been such a serious matter that it should have been done only if the illegality were certain. The illegality was far from certain as the text of Article 85 paragraph 3 was closer to an authorisation system than to a legal exception system, which would have required more straightforward drafting. If it is understandable that Deringer cites this passage in support of his thesis, this support is far from unequivocal. First, the Advocate General, who was mulling over the incompatibility of the authorisation system with the Treaty, thought that the Court should exercise considerable self-restraint before annulling a legislative choice. Second, the greater closeness of the text of Article 85 paragraph 3 to an authorisation system than to the legal exemption was, in his mind, the reason why it was difficult to say that the authorisation system was incompatible with the Treaty. In this context it made no sense to ask whether the text was open to both. But if we ask how a text that is meant to leave the choice to the legislator should be written, it is difficult to conceive of something better than what we have. Does the word declare point to an administrative decision?23 If anything, it points to a declaratory decision. 24 Is the word declare inappropriate to describe judicial activity? Of course not. A court interprets the law; it declares what the law means in a concrete case. This argument deserves to be stressed. The essence of an authorisation system is that, before the authorisation, the derogating provisions cannot be applied by the judge because they are not sufficient in themselves. They have to be narrowed down by an administrative 23

Monopolkommission (1999, point 16): 'Fiir diese Auffassung laBt sich ferner der Wortlaut des Art. 81 Abs. 3 EGV anfiihren, wonach eine Freistellung "erklart" werden kann. Eine Erklarung in diesem Sinn setzt eine rechtsgestaltende Wirkung voraus und kann sich gemaB Art. 81 Abs. 3 EGV auf Einzelvereinbarungen oder auf Gruppen von solchen beziehen. So bewirkt etwa eine erteilte Einzelfreistellung, daB die wettbewerbsbeschrankende MaBnahme ab dem Zeitpunkt, den die Europaische Kommission bezeichnet, voile Gultigkeit erlangt. In einem System der Legalausnahme gibt es keine "Erklarung" einer Freistellung mehr'. See also German Government (1999, point 1.1); Austrian Government (1999, point I.I). 24 The Monopolkommission (1999) turns the difficulty by assuming the desired result rather than demonstrating that it flows from the text (see quotation in the previous footnote).

162

Working Paper VI

authority whose decision is of a constitutive nature. Only after this decision has come into effect is the law complete. By contrast, the word declare seems to assume that the law is already complete and only needs to be applied to a concrete case or set of cases. That is to say that the wording of paragraph 3 is far from imposing a constitutive decision, even if one cannot come to the opposite conclusion that it imposes a legal exception system. In the same sense the expression relating to the inapplicability of the prohibition seens to point to a notion different from that of authorisation. While an authorisation assumes the applicability of the prohibition until the latter is lifted, the inapplicability appears to exclude the prohibition from the start. It is interesting to observe that, at the time the EEC Treaty was negotiated, the High Authority used precisely this terminology to describe the situation in which Article 65 ECSC did not apply at all to an agreement entered into before the date at which that provision came into effect.25 More importantly, one should not lose sight of the fact that this is language addressed to the legislator and it must therefore be assumed that, if it is shrouded in a certain degree of vagueness, it is in order not to pre-empt the legislator's choices.

3. Systematic arguments We thus come back to the point that the interpretation of the phrase in question must be seen in context. This is not a provision meant to be applicable by itself from a procedural point of view. Article 83 (former 87) contemplates that the Council should lay down detailed rules for the application of Article 81 (3), taking into account the need to ensure effective supervision on the one hand, and to simplify administration to the greatest possible extent on the other. It is hard to conceive that, in the presence of this delegation of powers, Article 81 paragraph 3 would have already determined the way in which the exemptions must be administered. If the authors of the Treaty had already made their choice, little would be left to legislators. In this case the Treaty ought to have been extremely clear, which is manifestly not the case. The language adopted by the Treaty does not in itself describe either a legal exception system or an authorisation system because it addresses itself to the legislator. If one reads paragraph 3 this way, everything becomes clear. All past claims to derive the final solution from the Treaty itself, as if legislators had in fact no other choice than to extract from the Treaty what its authors had already decided themselves, turn out to be in vain. Leaving open the way in which the exemption would be administered was all the more necessary in view of the fact that paragraph 3 contemplates two different possibilities: the inapplicability of paragraph 1 may concern either an 25

Cf. the decision challenged before the Court in Case 1/58 Stork [1959] ECR 17.

1 - Compatibility, Efficiency, Legal Security

163

individual agreement or a category of agreements. Each of these possibilities necessarily involves a different technique and different authorities. When applied to a category of agreements, the declaration of inapplicability is a regulatory activity that is certainly not of a judicial nature, that does not easily lend itself to be devolved to the national competition authorities, that might be entrusted directly to the Commission, but that the legislator might also (as the Council has done so far) reserve to itself to decide at least in principle, the Commission being responsible for the delegated legislation. How could one use this technique if paragraph 3 had established once and for all that the declaration of inapplicability was to take the form of an authorisation on the part of an administrative authority? The authors of the Treaty had to use a formula capable of encompassing a number of techniques in order to allow the Council to deal satisfactorily with a variety of situations. In addition, in order to be credible, the thesis that the Treaty has chosen the administrative authorisation system should point to a strong constitutional value enshrined in such a system. However, if we look at the arguments made by the supporters of this thesis, we find that they are about the alleged relative advantages of the authorisation system or the alleged relative disadvantages of the legal exception system. Such arguments do not disclose the existence of a superior principle of constitutional importance that could be infringed by the adoption of the legal exception system. Thus, according to those claiming the incompatibility of a system based on a legal exception with the Treaty: • in a legal exception system the competition authorities would, in practice, rarely take up horizontal agreements other than hard-core cartels, with the result that hard-core cartels tend to be severely repressed while infringements in the grey zone are rarely stopped; 26 • in a legal exception system the elimination of notifications causes the competition authority to lose an invaluable source of information and the overview of the market; 27 • legal certainty is better ensured in an authorisation system where firms can found on the authority to state what the legal position is and enjoy immunity from fines in the meantime, while the self-assessment required by a legal exception system is too hard for firms which also risk an aggravation of fines for hard-core cartels.28 Of course, these statements deserve examination and discussion. But the answer to the problems they pose is far from evident and it might change over time with the evolution of competition culture and the development of the 26

Monopolkommission (1999, points 26 and 27 and 36-37); G e r m a n Government (1999, points 1.2 and 1.5). 27 Monopolkommission (1999, points 28-32); G e r m a n Government (1999, point 1.5); Austrian Government (1999, point 1.3); Deringer (2000, point IV.3.i). 28 Monopolkommission (1999, points 33-35); G e r m a n Government (1999, points 1.3 and 1.5); Austrian Government (1999, point 1.2); Deringer (2000, points IV.3.i and IV.5).

164

Working Paper VI

markets. In other words, they are problems of policy that the legislator should be able to decide according to circumstances. By contrast, these statements do not point to any constitutional value that ought to be safeguarded come what may. The thesis—according to which the Treaty imposes the authorisation system on the legislator—is not only unwarranted by the text of Article 81, but it is also not credible in terms of systematic construction.

4. Interpretation by the Court Some critics take the view that the Court has already interpreted Article 81 paragraph 3 as implying the existence of an authorisation system and excluding the possibility of one based on legal exceptions. It goes without saying that the Court has never had an occasion to rule on this point. The arguments of these critics are therefore indirect and accordingly require an inordinate amount of space in order to be presented and discussed, and the judgments relied on need to be extensively quoted. In order not to interrupt the thread of reasoning, the presentation and discussion of these arguments takes place in Appendix III (Bosch judgment), IV (Portelange judgment) and V (Lancome judgment). Thefirsttwo of these judgments are put forward by Mok, the third by the German Government.

VI. Is paragraph 3 of Article 31 justiciable? The argument about the alleged insusceptibility of paragraph 3 to judicial application is used by the opponents of the legal exception system to support the position that paragraph 3 requires an administrative decision for its implementation, as well as the position that the proposed system would undermine the direct applicability of paragraph 1,29 The argument is that paragraph 3 includes undefined notions and requires a complex economic assessment not suitable to the judicial function. It cannot be denied that paragraph 3 contains notions that require a complex economic assessment to be conducted in order to be applied to a concrete case. However, this circumstance is not sufficient to conclude that paragraph 3 does not lend itself to judicial application. Paragraph 1 requires a no less complex economic assessment, as the Court of Justice has recognised.30 The same can 29 Monopolkommission (1999, point 18): 'Vor diesem Hintergrund bliebe die unmittelbare Anwendbarkeit des Verbotstatbestands nur erhalten, wenn auch Art. 81 Abs.3.EGV justitiabel ware'. Cf. also House of Lords (2000, 34-35, point 149). 30 See judgments in Case 42/84 Remia [1985] E C R 2566, Joint Cases 142/84 and 156/84 BAT and Reynolds [1987] ECR 4487, point 62 and C-215/96, and C-216/96 Bagnasco [1999] ECr 1-179, point 50.

1 - Compatibility, Efficiency, Legal Security

165

undoubtedly be said about Article 82, an article that nobody denies can be applied by national courts. In fact, the analyses under Articles 81 and 82 are closely intertwined in some cases, and it is artificial and uncomfortable to separate them from a procedural point of view.31 It can be added that while the national courts now dispose of no regulation or notice to guide them as to the manner of interpreting Article 82, a long set of regulations, guidelines and notices assist them in the application of Article 81. It is true that the Courts have formally displayed a degree of self-restraint. In Groupement des cartes bancaires32 the CFI has thus summarised the case-law: 'The review undertaken by the Court of the complex economic appraisals made by the Commission when it makes use of the discretion conferred on it by Article 85 (3) of the Treaty, with regard to each of the four conditions laid down in that provision, is necessarily limited to verifying whether the rules on procedure and on the statement of reasons have been complied with, whether the facts have been accurately stated and whether there has been any manifest error of appraisal or a misuse of powers (see the judgments of the Court of Justice in Remia, cited above, and in Joined Cases 142/84 and 156/84 BAT and Reynolds v Commission [1987] ECR 4487)" (point 109). A first observation will bear on a terminological question. The CFI uses the word discretion here. However, the French version uses the expression marge d'appreciation. The same can be observed in some judgments of the Court of Justice: in Metro I33 at points 45 and 50 respectively the Court refers to the Commission's margin of discretion and discretionary power. In the French version, the equivalent expression is marge d'appreciation in both cases; an expression similar to the one used at point 30 where, referring to Article 81 paragraph 1, the Court speaks of pouvoir d'appreciation in French and of power of appraisal in English. In substance, in almost 40 years of application of Regulation 17, the Commission has never refused an exemption when the conditions in paragraph 31 The Irish ice cream case is an example. In its decision of 11 March 1998 (Cases IV/34.073, IV/34.395 and IV/34.35.436— Van den Bergh Foods Limited, OJ L 246 1998, p. 1) the Commission found that the practice whereby Unilever provides an ice-cream freezer to a retailer for no direct payment, or induces the retailer to accept the freezer subject to the condition that no non-Unilever ice-cream be stocked in it constituted a non-exemptable infringement of Article 81 as well as an infringement of Article 82 EC. In a judgment of 28 May 1992 the Irish High Court had reached the opposite conclusion regarding both provisions. The matter is currently pending before the Court of First Instance, on application for annulment of the Commission decision (Case T-65/98 Van den Bergh Foods v Commission) as well as before the Court of Justice, on preliminary reference from the Supreme Court of Ireland (Case C-344/98 Masterfoods). The question of the legality of Unilever's distribution practice implies a complex economic assessment, with regard to the exclusionary effects of the practice on the one hand and Unilever's investment on the other hand. This analysis appears simultaneously relevant under Article 81 and under Article 82. 32 [1994] E C R 11-49. 33 Case 26/76 Metro / [1977] E C R 1875.

166

Working Paper VI

3 were met, nor has it granted an exemption for reasons other than the fulfilment of those conditions. As to the Courts, the underlying conception of the case law is that fulfilment of the conditions entitles an enterprise to the benefit of paragraph 3,34 even if this point has never been expressly decided because a firm requesting an exemption has never brought an action for failure to act against the Commission.35 Above all, the case law shows how punctiliously—despite the proclaimed selfrestraint—the Courts control the assessment made by the Commission in accepting or refusing the fulfilment of a condition.36 It is possible to conclude that paragraph 3 implies a margin of economic assessment, not a margin of discretion, with the consequence that it is a provision susceptible to judicial application. Neither the word may in paragraph 3, which indicates an attribution of power with the consequent obligation to exercise that power (and not discretion), nor the complexity of the assessment, stands in the way of this conclusion. In any event, even on the basis that the Courts actually exercise a limited power of review over the implementation of paragraph^, a legitimate question seems to be whether this is the result of the text of the Treaty, or rather of the system set up by Regulation 17. In such a system, where the implementation of the provision in question is entrusted to the executive (as opposed to the judiciary), it is a logical consequence that the judicial review is of an administrative nature and therefore more limited than the control that the ordinary courts 34

In Case T - 1 7 / 9 3 Matra Hachette v Commission [1994] E C R 11-595, the C F I said: '. . . the C o u r t considers t h a t , i n principle, n o anti-competitive practice can exist which, whatever the extent of its effects o n a given market, c a n n o t be exempted, provided that all the conditions laid d o w n in Article 85 (3) of the Treaty are satisfied and the practice in question has been properly notified to the Commission' (point 85). 35 T h e case-law i n s t e a d s h o w s consistently t h a t failure t o meet o n e condition is sufficient t o disqualify from t h e benefit of p a r a g r a p h 3: Joint Cases C-65/64 and 58/64 Cons ten and Grundig v Commission [1966] E C R 209; Case 26/76 Metro I [1977] E C R 1875' Joint Cases 209/78 a n d 215/78 Van Landewyck and Others v Commission (Fedetab) [1980] E C R 3125; C a s e 258/78 Nungesser [1982] E C R 2015; Joint Cases 43.82 and 63/82 VBVB and VBBB v Commission [1984] E C R 19; Case 42/84 Remia [1985] E C R 2545; C a s e Metro II [1986] E C R 3 0 2 1 ; Case 45/85 Verband der Sachversicherer [1987] E C R 405; Case Groupement des cartes bancaires [1994] E C R 11-49; Case T 13/23 Matra Hachette v Commission [1994] E C R 11-595; C a s e C - l 37/95 SPO and Others v Commission [1995] E C R 11-289; Joint C a s e s T - 5 2 8 / 9 3 , T - 5 4 2 / 9 3 , T - 5 4 3 / 9 3 a n d T - 5 4 6 / 9 3 Metropole television and Others v Commission [1996] E C R 11-649. 36

In the s a m e sense see K o n (1982), according to whom, despite warnings of selfrestraint, 'the C o u r t h a s always proved itself willing to enter into considerable economic detail when considering a n d o n occasion annulling Commission decisions in application of Article 85, p a r a . 3.' H e cites a s examples Case 71/74 Frubo v Commission [1975] E C R 563, Kali and Sah [1975] E C R 499 a n d Fedetab. By contrast, Wesseling (1999, 420) seems to overlook the reality o f the judicial review exercised by the European Courts when he writes that ' t h e E C J h a s repeatedly endorsed Commission decisions which took

into account non-competition concerns when determining whether an agreement fulfilled all four conditions' and that 'there is a large degree of administrative, or policy, discretion involved in the implementation of Article 81 (3)'.

1 - Compatibility, Efficiency, Legal Security

167

would exercise if they had jurisdiction. The review may tend to be even more limited if, in addition, the task is also entrusted to a single administration—the Commission—to the exclusion of national authorities, for such a monopoly can be understood as a wish on the part of legislators that a consistent policy be developed. If these remarks are correct, nothing prevents the legislator from triggering a new, more pervasive type of judicial involvement by opting for another system. More generally, concerns have been voiced that the courts would not be equipped to apply paragraph 3, this task being of a specialised economic nature and thus more suitable for an administration. These concerns lie at the level of expediency rather than on legal grounds, but they nonetheless require careful examination. In this regard, one should not lose sight of the fact that paragraph 3 is the European codified form of the American 'rule of reason', and that in America the 'rule of reason' is a judicial invention. It is not credible that the European courts would not be capable of applying what their American counterparts have succeeded in reading into the statute book despite the Sherman Act's silence on the subject. Paragraph 3 is, in fact, a less complicated text than it appears. Its purpose is to save from prohibition those contractual restraints of trade that in the long run have a pro-competitive function. The conditions it lays down are criteria intended to guide the interpreter in finding the right solution to this question. Their implementation is no doubt a delicate operation, but—as demonstrated by the American experience—it is by no means one that defies judicial capacities.

VII. Other objections to compatibility The major objections to compatibility are, it seems to me, those that were discussed so far. However, in the interests of completeness, mention should be made of other arguments that, at least in part, constitute a different presentation of the main ones.

1. Paragraphs 1 and 3 cannot be subjected to the same regime (Articles 81 and 83) It is argued that it is not in conformity with the Treaty for the two paragraphs to be implemented through the same procedure. The Treaty is said to impose that, in implementing Article 83, the legislator chooses for paragraph 3 a different procedure from the one applicable to paragraph 1. Since paragraph 1 is directly applicable, it would not be open to the legislator to render paragraph 3 equally directly applicable. To do so would inter alia prejudice the direct

168

Working Paper VI

applicability of paragraph 1. The necessary separation {Trennung) in the treatment between the two paragraphs is alleged to derive from the difference in the wording of the prohibition provision as opposed to the provision on exceptions. It is also said to flow from the different treatment reserved to the two paragraphs by Article 83 (former 87) paragraph 2 under (a) and (b) respectively.37 In particular, if paragraph 3 were rendered directly applicable, it would become impossible to inflict a fine for an agreement incompatible with paragraph 1 without first ensuring that the conditions laid down in paragraph 3 are not met. This situation is said to run counter to the provision of Article 83 paragraph 2 under (a). The provision under (b) would become superfluous, as with the direct applicability of Article 81 paragraph 3 there would no longer be any need for detailed rules for the application thereof.38 The idea of Trennung between the two paragraphs is of course undeniable. The Treaty does not put them on the same footing. While the prohibition is directly applicable, the regime of the exceptions is not determined by the treaty and is left to the legislator to decide. That does not mean, however, that the legislator is necessarily prevented from rendering the exceptions equally directly applicable. That conclusion is a non sequitur. The insistence on wanting to be able to apply the prohibition without examining whether the conditions laid down in paragraph 3 are met, it defies common sense and also the case-law that has underlined the link between the two paragraphs.39 37

Article 83 reads: ' 1 . Within three years of the entry into force of this Treaty the Council shall, acting unanimously on a proposal from the Commission and after consulting the European Parliament, adopt any appropriate regulations or directives to give effect to the principles set out in articles 81 and 82. If such provisions have not been adopted within the period mentioned, they shall be laid down by the Council, acting by a qualified majority on a proposal from the Commission and after consulting the European Parliament. 2. The regulations or directives referred to in paragraph 1 shall be designed in particular: a. to ensure compliance with the prohibitions laid down in Article 81 (1) and in article 82 by making provision for fines and periodic penalty payments; b. to lay down detailed rules for the application of Article 81 (3), taking into account the need to ensure effective supervision o n the one hand, and to simplify administration to the greatest possible extent on the other; c. to define, if need be, in the various branches of the economy, the scope of the provisions of Articles 81 and 82; d. to define the respective functions of the Commission and of the Court of Justice in applying the provisions laid down in this paragraph; e. to determine the relationship between national laws and the provisions contained in this Section or adopted pursuant to this Article.' 38 Mestmacker (1999, 525 and 526); Monopolkommission (1999, points 17 and 18); G e r m a n G o v e r n m e n t (1999, point I.I); Austrian Government (1999, point I.I). 39 Reference is m a d e t o the case law where the C o u r t has considered the applicability of the prohibition in a situation where there is legally no authority to examine whether the conditions for the application of the exemptions are met. The Court has consistently ruled that the prohibition does not apply in such a situation. Already under the Coal and

1 - Compatibility, Efficiency, Legal Security

169

Finally, the fear that Article 83 paragraph 2 (b) would become superfluous must be a relic of the time when the supporters of the two theories based themselves on the Treaty. In fact, even at that time nobody to my knowledge ever Steel Treaty, this situation was examined in the Stork case. While the ECSC Treaty came into force on 25 July 1952, the High Authority was only empowered to authorise the restrictive agreements in question as from 10 February 1953, the date of the establishment of the common market for coal. Was the prohibition already applicable during the six months between these two dates? No, answered the Court: 'The system established by Article 65 of the Treaty is based not only upon the prohibition of agreements laid down in paragraph 1 of that article but at the same time on the possibility of authorizing useful and necessary agreements contained in paragraph 2 .. . It is inconceivable that the Contracting Parties accepted a situation whereby, for a period of uncertain duration following the entry into force of the Treaty . .. the prohibition in Article 65 (1) was applicable, whilst the power of authorization provided for in paragraph 2 of that article and closely connected with the prohibition was inapplicable . . . That interpretation alone avoids the unsatisfactory situation [le resultat choquant] described above, that is to say, the arbitrary separation of the various connected elements of Article 65 into those which are immediately applicable and those which would only become applicable after an indefinite period' (1959 ECR 17, point 5). A similar question was raised in the Bosch case concerning an agreement pre-existing to regulation 17. Despite the existence of transitional provisions in the EEC Treaty (Articles 88 and 89), the Court ruled that'. .. it would be contrary to the general principle of legal certainty. . . to render agreements automatically void before it is even possible to tell which are the agreements to which Article 85 as a whole applies'. The same problem

was raised in Asjes in thefieldof air transport, not yet covered at that time by any implementing regulation under Article 87 (now 83). The Court said (points 61-64): 'It should be borne in mind that, as the Court held in its judgment of 6 April 1962 (Case 13/61 Bosch v. Van Rijn [1962] ECR 45), "Articles 88 and 89 are, however, not of such a nature as to ensure a complete and consistent application of Article 85 so that their mere existence would permit the assumption that Article 85 had been fully effective from the date of entry into force of the Treaty". In fact Article 88 envisages a decision by the authorities of Member States on the admissibility of agreements, decisions and concerted practices only when these are submitted for their approval within the framework of the laws relating to competition in their countries. Under Article 89 the Commission is empowered to record any infringements of Articles 85 and 86 but it does not have the power to declare Article 85 (1) inapplicable within the meaning of Article 85 (3). In those circumstances the fact that an agreement, decision or concerted practice may fall within the ambit of Article 85 does not suffice for it to be immediately considered to be prohibited by Article 85 (1) and consequently automatically void under Article 85 (2). Such a conclusion would be contrary to the general principle of legal certainty, which, as the Court held in its said judgment of 6 April 1962, is a rule of law that must be upheld in the application of the Treaty, since it would have the effect of prohibiting and rendering automatically void certain agreements, even before it is possible to ascertain whether Article 85 as a whole is applicable to those agreements.' See also Portelange [1969] ECR 309; Case 14/76 De Bloos v Bouyer [1977] ECR 2359; Case 48/72 Brasserie De Haecht [1973] ECR 77 (where the Court speaks of indivisibility of paragraphs 1 and 3). Another case-law which emphasizes the connection between paragraphs 1 and 3 is the one about the mechanisms (in particular suspension of proceedings in national courts) to be used in order to avoid conflict between national courts applying paragraph 1 and the Commission disposing of monopoly for exemption: see Case 234/89 Delimitis [1991] ECR 1-935.

170

Working Paper VI

maintained that paragraph 3 not only established the legal exception system but was also directly applicable so that an implementing regulation to that effect was unnecessary. In any event, the White Paper does not claim that the legal exception systemflowsfrom the Treaty itself. In order to achieve that result, legislative intervention is necessary, precisely on the basis of Article 83 paragraph 2 (b): a new Regulation replacing Regulation 17 would have to lay down that the application of paragraph 3 takes place at the same time as the application of paragraph 1.40 2. Article 84 Article 84 (former 88) is said to have already indicated that paragraph 3 is to be applied by an administrative authority.41 Article 84 reads: 'Until the entry into force of the provisions adopted in pursuance of Article 83, the authorities in Member States shall rule on the admissibility of agreements, decisions and concerted practices and on abuse of a dominant position in the common market in accordance with the law of their country and with the provisions of Article 81, in particular paragraph 3, and of Article 82.'

According to this argument, by referring to the authorities in Member States, the Treaty made clear that paragraph 3 cannot be applied by the courts. According to this approach, not only would Article 84 itself prevent that application during the transitional period preceding the entry into force of the Council regulation, but it would also shed light on the interpretation of paragraph 3 so as to guide the legislator as to which type of system can be chosen. In this context it might be remembered that the Court has interpreted the word 'authorities' as including in certain Member States courts especially entrusted with the task of applying domestic legislation on competition or that of ensuring the legality of that application by the administrative authorities, but excluding the other courts.42 40 E h l e r m a n n (2000, 22, point 54) argues that even in a regime of a directly effective exemption Article 83 paragraph 2 b remains useful in order to a d o p t accompanying regulatory mechanisms. A s explained in the text, there is more t h a n that: the power granted t o the Council by this provision must be exercised to render Article 81 paragraph 3 directly applicable in the first place. 41 M o n o p o l k o m m i s s i o n (1999, point 17 in fine); G e r m a n G o v e r n m e n t (1999, point 1.1). 42 30 J a n u a r y 1974, Case 127/73 BRTvSABAM[1914] E C R 51. In Asjes(Joint Cases 299/84 a n d 213/84 Ministere Public v Asjes [1986] E C R 1425), though citing SABAM with approval, the C o u r t seems t o have further circumscribed the notion of authorities: ' . . . the t e r m " a u t h o r i t i e s " in M e m b e r States in Article 88 refers to either the administrative authorities entrusted, in most M e m b e r States, with the task of applying domestic legislation o n c o m p e t i t i o n subject to the review of legality carried o u t by the c o m p e t e n t courts, o r else the courts to which, in other M e m b e r States, that task has been especially entrusted. Construed in that way, the term "authorities in M e m b e r States"

1 - Compatibility, Efficiency, Legal Security

171

In any event, I do not think that much can be drawn from this judgment in one sense or the other. What counts is that Article 84 is a transitional provision destined to lose its purpose with the entry into force of the Council regulations adopted under Article 83. By definition, therefore, it cannot prejudge the content of these regulations.

VIII. Questions of internal consistency Certain criticisms of the internal consistency of the proposals contained in the White Paper do not properly relate to the issue of the compatibility with the Treaty of a legal exception system. However, they do pose connected problems that make it advisable to discuss them in this context. In a legal exception system there are no exemption decisions. The application of paragraph 3 does not give rise to specific decisions. Administrative decisions applying Article 81 only have a negative content for companies, that is inflicting penalties or enjoining given behaviour. Such decisions imply that paragraph 1 is applicable and paragraph 3 is not. Similarly, whenever one party relies on it, the court decides whether Article 81 as a whole is applicable or not. However, the White Paper suggests the possibility that the Commission could make positive decisions. Moreover, the power of exemption would survive in the traditional form of a block exemption regulation. Finally, a new form of administrative decision inspired by the consent decree—a judicial instrument of American law—is advocated. These three instruments have been criticised as inconsistent with a legal exception system. 1. Positive decisions According to the White Paper (points 88 and 89): 'It is true that the Commission would no longer adopt exemption decisions under Article 85 (3) as it does not, but it should nevertheless be able to adopt individual decisions that are not prohibition decisions. Where a transaction raises a question that is new, it may be necessary to provide the market with guidance regarding the Commission's approach to certain restrictions in it. Positive decisions of this kind would therefore be taken in exceptional cases, on grounds of general interest. These positive decisions would confine themselves to a finding that an agreement is compatible with Article 85 as a whole, whether because it falls outside Article 85 (1), or because it satisfies the tests of Article 85 (3). They would be of a declaratory nature, and would have the same legal effect as negative clearance decisions have at present.'

within the meaning of Article 88 does not include the criminal courts whose task is to punish breaches of the law.'

172

Working Paper VI

Is it possible to reconcile a legal exception system with the proposal to grant to the Commission a monopoly over the adoption of positive decisions?43 Will these decisions bind the national courts and authorities?44 These questions are of great importance because it is one thing to say that the treaty allows the legislator to choose either an authorisation system or a legal exception system, and it is another to say that the same legislator can combine the two systems as he likes. In a legal exception system there is concurrent jurisdiction of administrative authorities and national courts in the application of Article 81 to a concrete case. Is it imaginable that an administrative authority like the Commission would retain the ability to bind the courts in such a system? I think that such a possibility would be contrary to the separation of powers. In the present authorisation system the monopoly granted to an administrative authority excludes the courts altogether from the application of Article 81. What is not possible is to give the courts a power and, at the same time, to reserve the possibility that an administrative authority could dictate to them the way it should be used.45 However, that is not the effect that the positive decisions envisaged in the White Paper would have: 'They would be of a declaratory nature, and would have the same legal effect as negative clearance decisions have at present.' Their purpose would be to make the position of the Commission known in order to try to persuade the courts, without claiming to bind them.46 In this context it is also important to stress that companies should not be entitled to a positive decision. The Commission would use this possibility in the public interest only in exceptional cases. Were it otherwise, the whole point of the reform—that is, freeing the Commission from the bureaucratic obligation of vetting all agreements about which companies decide to notify it—would be lost.

43

Contra German Government (1999, point 1.3.2). Mestmacker (1999, 527). 45 For this reason the proposal to combine a legal exception system with the possibility to notify and claim a constitutive decision—cf. Siragusa (2000)—seems to me to be legally impossible. In Siragusa's system the national courts can apply Article 81 (3) but must stay proceedings if the agreement under dispute is notified to the Commission. Apart from the perpetuation of the defects of the present regulation, this system would thus place the courts in a subordinate position vis-a-vis the Commission. 46 Deringer (2000) is therefore quite right when he writes: 'Bei einem System der Legalausnahme ist auch logisch eine "Freistellung" nicht moglich, denn die Freistellung ist ein konstitutiver Verwaltungsakt, wahrend bei dem System der Legalausnahme nur die negative Feststellung moglich ist, dass kein Grund bestehe, gegen eine Vereinbarung vorzugehen' (p. 7) and 'EinEin Negativattest der Kommission dagegen wiirde vielleicht die Beurteilung durch die nationalen Behorden oder Gerichte beeinflussen, es konnte diese aber nicht daran hindern, die betreffende Vereinbarung doch zu untersagen' (p. 9). 44

1 - Compatibility, Efficiency, Legal Security

173

2. Block exemptions If it is not possible to combine a legal exception system with the power of an administrative authority to make constitutive decisions, is it not contradictory to contemplate that the Commission would still adopt block exemption regulations?47 There is a fundamental difference between constitutive individual decisions and block exemption regulations in their respective interplay with a legal exception system. As has been seen, the problem with constitutive individual decisions is that they interfere with, and condition, the courts'jurisdictions, if they were allowed, the principle of separation of powers would be wounded. This is not the case with block exemption regulations; they legislatively circumscribe a portion of the field where Article 81 is not applicable. Once they have been adopted, there is no possible interference between courts and administrative authorities as to whether Article 81 would be applicable to the agreements described. The applicable law is the regulation itself and the only residual question is whether a concrete agreement meets the legislative description or not.

3. Decisions accepting commitments These decisions are contemplated in point 90 of the White Paper: 'In the course of procedures that might otherwise end with a prohibition, it can happen that the undertakings concerned propose to give the Commission commitments that would overcome the objections raised against their agreement. It is useful that the Commission should be able to make such commitments binding, both in order to oblige the undertakings to comply with them and to enable the parties and others to rely on them before their national courts. In the new Regulation applying Articles 85 and 86, therefore, the Commission intends to make provision for a new kind of individual decision, subject to the ordinary publication requirements, in which the Commission would take note of the commitments entered into by the parties and render them binding. Such a decision would allow the procedure to be terminated while ensuring that the commitments were respected. As a corollary to this change, a clause would be included in the system of penalties in the Regulation providing forfinesand periodic penalty payments in the event of failure on the part of undertakings to meet such commitments.'

47 Deringer (2000, 7 and 8) gives an affirmative answer to this question. According to him the Commission should only be able to adopt block negative clearance regulations. Mestmacker 91999, 525) has a different problem. He takes the view that in a legal exception system there is no room for block exemption regulations, because the legal exception is tantamount to a general exemption. It has already been answered to this objection above (see text corresponding to footnotes 5 and 10) by denying that a legal exception system means a general exemption.

174

Working Paper VI

In this type of decision there would be no finding of infringement, which is precisely the interest that the firm concerned might attach to it. The recitals would certainly explain that the Commission has started proceedings to ascertain whether a certain infringement has been committed. They would continue by saying that, before the Commission had come to a conclusion, the firm, without accepting responsibility, has offered commitments capable of ending any interest for the Commission to continue the proceedings. While there would be nofindingof infringement, nor would there be of course anyfindingeither that the situation after the commitments is legal. Are these decisions of a constitutive kind that would be irreconcilable with a legal exception system? In particular, would a company to which such a decision was addressed become immune from damages?48 The answer is 'no'. When the White Paper says that the commitments would be made binding by the decision (point 90), it does not intend to state that the national courts would be prevented from analysing agreements differently with regard to Article 81. The binding character of the commitments means that: • the Commission can oblige a company to respect its commitments by imposing fines and penalty payments; • third parties can ask a national court to force a company to respect the commitments; • parties to restrictive agreements can no longer rely on the original contracts once they have been modified through their commitments. By contrast, provided they do not interfere with the binding character of the commitments, the courts remain free to make their own analysis about the application of Article 81 to the agreement in question.

IX. Conclusion The prohibition principle is enshrined in the Treaty, and the documents on the negotiations of the Treaty show that this result was deliberate. It represents a constitutional value that reflects the commitment of the authors of the Treaty to the preservation of firms' freedom of action. To organise a control of abuse would not respect this value. Among the three basic enforcement methods of an anti-cartel provision, 'control of abuse' is the least effective and a 'legal exception system' is the most rigorous from the point of view of protecting competition. The place of an 'authorisation system' lies between the other two. Though respectful of the prohibition principle, an authorisation system is less rigorous than 'legal exemption' as it gives firms the possibility of benefiting (for a time) from immunity 48

Mestmacker (1999, 527 in fine).

1 - Compatibility, Efficiency, Legal Security

175

from fines and, practically even if not theoretically, from actions in courts. Only in a legal exception system does the prohibition apply fully. For the Community to have started with the authorisation system and to evolve towards a legal exception system would be a natural and historically justified crescendo. When it was grafted onto European law in the fifties, competition law was foreign to the continental tradition. Not only in the former officially corporatist countries, but in all six original member States, and even in unsuspected countries like the Netherlands, the entrenched economic culture was closer to ideas of order, agreement and organisation among suppliers than to competition. It was wise to centralise competition policy in the Commission's hands: the grafting needed a dedicated gardener to take care of it in a greenhouse. Now that the grafting has succeeded, Europe is mature enough for the upgrading of its competition law in a way that would result from exposing it to the full participation of all interested players in its enforcement. Once it is clear that the White Paper aims at a legal exception system and not to control of abuse, criticism based on a concern that the prohibition principle would be watered down should lapse and its authors should welcome the project. More difficult to handle is the criticism that the authorisation system is enshrined in Article 81 paragraph 3 of the Treaty. For, as Lagrange AG observed, the text of paragraph 3 accords better with such an authorisation system than with a legal exception one. This is because—at first reading—it seems that an administrative decision is required. On closer examination, however, the text of paragraph 3 does not suit either solution very well. The verb to declare is not what one would use to refer to an authorisation. And the text is incomplete. If the treaty-makers had wanted an authorisation, why did they not specify that it would be granted by the Commission (thus avoiding a Council regulation to that effect)? Far from it, a Council regulation was specifically contemplated in Article 87 (now 83) in order to lay down rules for the application of Article 85 (now 81) paragraph 3, taking into account the need to ensure effective supervision on the one hand, and to simplify administration to the greatest possible extent on the other. In view of this legislative programme can one doubt that paragraph 3 did not in itself resolve the problem of how it was to be enforced and that the authors of the Treaty left this task to the legislator? The negotiation documents point decisively, albeit not expressly, towards this conclusion. If an express confirmation cannot be found, it is only because most of those documents contain successive drafts, rather than the reasons for the solutions agreed upon. It remains to be seen whether an administrative decision is not in any event required on the ground that the conditions laid down by paragraph 3 are such—so runs the last major argument for incompatibility—that they imply the exercise of discretion and they are therefore not justiciable. A complex power of appraisal is in fact inherent in the application of these conditions, but no discretion. If anything, the discretion is the result of the system set up by Regulation 17; it is not inherent in the text of the Treaty. The opposite would

176

Working Paper VI

be strange if one only remembers that paragraph 3 is no more than a codified European version of the judicially-invented American rule of reason. European judges would not be able to follow in the steps of their American colleagues? A negative answer would be an expression of the traditional European lack of self-confidence. Of course, not everything will be perfect from the first day but, if we never start, we will never be ready.

1 - Compatibility, Efficiency, Legal Security

177

APPENDIX I: DOCUMENTS FOUND AT THE HISTORICAL ARCHIVES OF THE COUNCIL 27.6.1956

MAE 139/56

Propositions en vue de la redaction du traite One article covers agreements, abuses of monopoly, discriminations and mergers: paragraph 1 is about what is incompatible; paragraph 2 is about regulations that will have to specify mechanisms and derogations, such as agriculture, energy, water; paragraph 3 anticipates some of the mechanisms.

4.7.1956

MAE 144/56

6.7.1956

MAE 148/56

Suggestions de M. von der Groeben The text on competition is the same, or very close, to the previous document. Projet d 'articles One single article (II-3) declares that a number of practices and situations are incompatible with the common market, to the extent that trade between Member States is affected, including those which prefigure the infringements mentioned in Articles 85 and 86 (today 81 and 82). There is no mention of exemptions or derogations. For the implementation of these principles the Commission is entrusted in the same article with the adoption of regulations and with the instruction of individual cases of infringement. Decisions in these cases fall to the Court on the initiative of the Commission or a Member State.

17.7.1956

MAE 175/56

4.9.1956

MAE 233/56

Projet d 'articles No significant change Proposition de la delegation frangaise This French proposal divides the substance and the procedure between two different articles. The first (X + 2) continues to prefigure both Article 85 and 86. Paragraph 1 declares the incompatibility of certain practices and situations with the common market but without the condition of an effect on trade. The second states that: 'Peuvent etre relevees des dispositions precedentes les situations ou pratiques dont les auteurs sont a mesure dejustifier ...'. What follows prefigures the first two conditions of Article 85 paragraph 3. A paragraph 3, not drafted, is destined to prescribe rules concerning state monopolies and public services. In the article on procedure (X + 3), the Member States are required to make the legislation necessary to implement the previous provisions, while the Commission can encourage the Member States to apply certain remedies in individual cases where there is an effect on trade.

178

5.9.1956?

Working Paper VI

MAE 252/56

(extrait)

Proces-verbal des reunions des 3-5 septembre 1956 du

Groupe du marche commun In the meeting of the Groupe du Marche Commun of 3-5 September 1956, the German delegation insists that the provisions on restrictive agreements must be based on the prohibition principle, while some exemptions should be admitted but subject to a control of abuse regime. The Belgian delegation favours a control of abuse approach rather than an absolute prohibition. The French delegation refers to its proposal contained in document MAE 233/56 (see above). On procedure, the Germans consider that the implementation should be left to the Member States; the Belgians agree but envisage that Community organs would assume responsibility in the case of failure to act by the Member States; the French refer to their proposal; the Italians favour the competence of the Community's organs; and the Dutch seem to be close to the Belgians.

7.9.1956

MAE/Sec. 29/56

10.9.1956

24.9.1956

Premiere lecture des articles 40 a 43 According to the German delegation (Mr MiillerArmack), too strict a competition regulation would be counterproductive. The rules should not be supranational but should only be the starting point for the harmonisation of national legislation. The French delegation (M. Donnedieu de Vabres) considers it is necessary to have a precise and complete regulation. In his summary, the Chairman thinks that all delegations agree that the treaty should limit itself to fundamental principles but that they disagree about the question of whether these principles should be supranational in character. Projet, remis par M. Thiesing, de nouveau texte des articles 42 et 43 Article 42a concerns agreements. The first period is an embryo of the future paragraph 1. The beginning of the second period reads: 'Toutefois, lorsque la demande en estfaite, les accords ou decisions de ce genre sont autorises, si le requerant est en mesure deprouver. . .'

MAE 325/56

Tableau synoptique is a synoptic table of the drafts put forward by the delegations. There are three drafts a German one, a French one and a Belgian/Dutch one. One column of the table summarises the divergences by saying that the French and the Belgian/Dutch drafts deal with agreements and monopolies (dominant positions) in the same provision but while the French text forbids them both in principle, the Belgian/Dutch text submits them to a control

1 - Compatibility, Efficiency, Legal Security

179

of abuse; the German draft deals with agreements and monopolies separately by prohibiting the first and subjecting the second to a control of abuse. Looking at the texts in more detail, we observe that the German text expresses the difference between outright prohibition (for the agreements) and control of abuses (for monopolies) by using in the first case the wording Sont incompatibles avec le marche commun et interdits and, in the second case, the wording est incompatible avec le marche commun et doit etre interdit, while the exception to the prohibition of agreements is worded Toutefois, lorsque la demande en estfaite, les accords ou decisions de ce genre peuvent etre autorises si le requerant est en mesure deprouver qu'ils contribuent... The French text expresses the prohibition of both agreements and monopolies simply with the word incompatible and formulates the exception with the words Peuvent etre relevees des dispositions prededentes les situations ou pratiques dont les auteurs sont en mesure dejustifier qu'elles contribuent.... The Belgian/Dutch text expresses the subjection of both agreements and monopolies to a control of abuse with the words Sont incompatibles avec le marche commun et a interdire. As to procedure, the synoptic table summarises the divergences in the following way. The legislation necessary to implement the rules of the treaty would be taken by the Member States according to the French, while the Germans favour a further international convention and contemplate a Council regulation only if no agreement is reached; the Dutch prefer a Community regulation from the start. As to the individual cases the Germans want the Commission to set up an agency to investigate alleged infringements: if the parties (undertakings opposing each other) do not accept the amicable solution proposed by this agency, a Member State or the Commission can go to Court. The French proposal (far from clear) the Member States are competent but the Commission can make proposals and in some cases take measures; if the Member States fail to take measures, the Commission can go to the Council. According to the Dutch, only the Commission should investigate and decide. 26.10.1956

MAE 468/56

Note proposee par le President du Groupe du marche commun comme base de discussion pour la deuxieme lecture In addition to interesting reflections on the scope of the rules that do not touch our problem, the Chairman proposes a text for the future Article 85 in which paragraph 1 is similar to the final text. The

180

Working Paper VI

beginning of paragraph 2 reads: 'Sont toutefois autorises [et valables] les accords entre entreprises et les decisions d associations d'entreprises qui ont ete declares a la Commission [en vue de leur enregistrement] et au sujet desquels le declarant petit fournir lapreuve qu'ils contribuent...' (what follows is in substance the final text). 6.11.1956?

Proposition neerlandaise amendee au cours de la reunion restreinte du 6 novembre Paragraph 2 reads: 'Sont toutefois admis les accords entre entreprises et les decisions d'associations d'entreprises qui ont ete declarees a la Commission en vue de leur enregistrement et qui ne sont pas incompatibles avec les objectifs du present Traite, notamment en entravant ['amelioration de la production ou de la distribution des produits, ou a la promotion duprogres techniques ou economique, a condition qu'ils riimposent aux entreprises interessees que des restrictions indispensables a la realisation des objectifs precith.'

7.11.1956?

MAE 525/56 (extrait)

Proces-verbal reunion 5-7 novembre 1956 du Groupe du marche commun The group examines the various text proposals on the basis of a note prepared by the Chairman (MAE 468/56 above). It is decided to entrust a restricted group with the drafting of a new text taking into account the discussion and including as few alternatives as possible.

12.11.1956

MAE 527/56

Projet etabli les 7 et 12 nov. 1956 par un groupe d'experls a I'intention du Groupe du marche commun Paragraph 1 has the structure of the final text. Paragraph 2 has a common trunk: 'Ne tombent pas sous I'interdiction edictee au paragraphe precedent les accords . . . qui ont ete declares a la Commission en vue de leur enregistrement', which bifurcates at this point. The first variant goes on: 'et au sujet desquels lapreuve peut etrefourniepar le declarant qu'il contribuent. . .' (what follows are, in substance, the four conditions of the final text). The second variant reads: 'et qui ne sont pas incompatibles avec les objectifs du present Traite, notamment en entravant I 'amelioration de la production ou de la distribution des produits, ou la promotion du progres technique ou economique, a condition qu'ils riimposent aux entreprises interessees que des restrictions indispensables a la realisation des objectifs precites'. A footnote explains that the agreement is presumed to be legal if it is declared to the Commission, but that the Commission may prohibit it. Paragraph 3 states that: 'Les accords ou decisions

1 - Compatibility, Efficiency, Legal Security

181

interdits en vertu du paragraphe 1 ci-dessus sont nuls de plein droit et ne peuvent etre invoques devant aucune juridiction des Etats membres. [La mime nullite frappe les accords declares au terme du § 2 ci-dessus mais interdits par la Commission.]' A footnote says that the Belgian expert prefers that the prohibition be 'declared' by a court. On procedure, within three years Member States can enter into an international convention on the way the rules in the treaty should be implemented. After this time, if the convention has not come into effect, Community instruments would be adopted. 14.11.1956

MAE 541/56

Proposition soumise par le President du Groupe du marche commun Article 42 (agreements) is limited to paragraph 1, while Article 43 (3) (on procedure) states that the international convention or the Community regulations: 'doivent notamment regler les cas d'exemption de {'article 42 ainsi que la question de savoir si les accords... qui entrent dans le cadre de ces exemptions doivent etre declares a la Commission en vue de leur enregistrement, ou s'ils doivent etre autorises par la Commission'.

14.11.1956

MAE 547/56

Redaction proposee par la delegation neerlandaise 'Peuvent etre interdits tous accords qui sont susceptibles . . . et qui ont pour effet. . . dont I 'application est incompatible avec les objectifs . . . notamment en entravant. . . Les accords, decisions ou pratiques interdits en vertu du paragraphe 1 sont nuls de plein droit et ne peuvent etre invoques devant aucune juridiction de Etats membres'. This Dutch proposal thus makes the nonexistence of what we call the 'conditions for exemption' a requirement for the prohibition. Proposition francaise This French proposal concerns only the Article on procedure and prefigures the future Article 87. One of the subjects to be dealt with in the implementing regulations consists of: 'determiner les conditions dans lesquelles pourront etre etablis les accords autorises en vertu du § 2 de Varticle 42 en tenant compte de la necessite, d'une part, d'assurer une surveillance efficace, et, d'autre part, de simplifier dans toute la mesure possible le controle administratif.'

20.11.1956

20.11.1956

MAE/Sec. 70

Groupe du marche commun, 13,14 et 15 novembre 1956—Examen en seconds lecture des regies de concurrence The chairman invites the Group to discuss the draft prepared on 7 and 12 November by a group of experts. Mr Muller-Armack (D) reserves his position

182

Working Paper VI

on the whole draft and voices concern about the quick pace of the works. As to agreements, he prefers the prohibition principle as in the German legislation. But the draft does not seem to reflect it because, while paragraph 1 contemplates prohibition, paragraph 2 contains too many exceptions. It is better, then, that the treaty be restricted to a general declaration of the incompatibility of restrictive agreements and leave the Member States to implement it during the first stage. After four years the Commission would make proposals for common legislation. Mr Van Tichelen (B) is concerned about the consequences of this proposal for the treaty objectives and asks Mr Muller-Armack whether acceptance of the prohibition principle by the other delegations would prompt a less backward position on his part. The answer to this question is intriguing: 'M. Muller-Armack repond en disant que sa position a I'egard de ce probleme est suggeree par le souci que le projet de traite, au cas oil il enonce le principe de I'interdiction des ententes, nefasse I'objet d'une forte opposition venant de certains milieux impresses'. Mr Hyzen (NL?) finds that the draft mixes up two different systems, the prohibition system and the abuse control system. Only if a homogeneous solution were impossible to reach, should the German proposal be taken into consideration. Mr Donnedieu de Vabres (F) supports the draft as well as the proposal put forward by his delegation for Article 43 (future 87). He takes the view that the criterion should be as flexible as possible, without renouncing to state that restrictive agreements and monopolies are incompatible and prohibited. 20.11.1956

MAE 602/56

Projet de redaction etabli le 20 novembre par un groupe d'experts compte tenu des echanges de vues intervenue le 19 novembre 1956 en groupe restreint Paragraph 1 differs from the final text only in insignificant formal details. Paragraph 2 says: 'Seront releves des interdictionsprecedentes les accords ...' (' Von den vorstehenden Verboten werden ... befreit.. .') (what follows is in substance the final text, apart from the specification that the burden of proof lies with the firms). Paragraph 3 reads: 'Les accords ou decisions interdits en vertu du paragraphe [sic] precedent sont nuls deplein droit et ne peuvent etre invoques devant aucune juridiction des Etats membres'. Article 43 (future 87) contemplates an international convention (paragraph 1) or, in its absence, Community instruments (paragraph 2). Paragraph 3 (b) is virtually identical to the final text of paragraph 2.

I - Compatibility, Efficiency, Legal Security

28.11.1956

MAE 657/56

183

Projet de redaction etablipar le Groupe du marches commun au cours de la seance du 28 novembre 1956 Paragraph 1 is the same as in the previous document. The beginning of paragraph 2 is now: 'Peuvent etre declares valables les accords . . . " (in the following text

the specification about the burden of proof has been dropped). In paragraph 3 (nullity), 'paragraphe precedent' has been replaced with 'paragraphe 1 du present article'. In Article 43 the international convention has been eliminated; the text is in substance the final text. 29.11.1956?

MAE 785/56

Extrait proces-verbal de la reunion des 27-29 novembre 1956 du Groupe du marche commun The text in the previous document is submitted for approval to the Comite des chefs de delegation. .

6.12.1956?

MAE 781/56

Extrait proces-verbal de la reunion du 6 decembre 1956 du Comite des chefs de delegation The Comite approves the text including the corrigendum [not found] provided that Article 42 paragraph 3 (nullity) is modified so that, instead of en vertu du paragraphe 1 du present article, it should read en vertu du present article; at the same time, the Groupe de redaction is requested to advise on the question whether the words sont nuls deplein droit can be suppressed without weakening the prescription of the paragraph.

10.12.1956

MAE 788/56

Redaction approuvee par le Comite des chefs de delegation le 6 decembre 1956 No change. In the German version the beginning of Paragraph 2 reads: 'Es konnenfur gtiltig erkldrt werden: . . .'

14.2.1957

MAE 517/57

lire lecture redactionsproposeespar les rapporteurs du Groupe du marche commun The beginning of paragraph 1 now reads: 'Sont incompatibles avec le marche commun, interdits et nuls de plein droit..." The following sentence is added at the end of paragraph 1: '[La nullite de plein droit visee cidessus doit etre constatee d'office par les iuridictions des Etats membres.]' Paragraph 2 reads: 'Toutefois, les dispositions du paragraphe 1 du present article peuvent etre declarees non applicables a ...'. There is no longer a paragraph 3 (emphasis in the original).

25.2.1957

MAE 648/57

Groupe de redaction—Premiere lecture Paragraph 1 is the final text. Paragraphs 2 and 3 have exchanged positions, as in the final text. Paragraph 2 reads as in the final text with the addition of a sentence in square brackets: 'et ne peuvent etre invoques devant aucune juridiction des Etats membres'.

184

Working Paper VI

Paragraph 3 retains the change introduced in the previous document, that is: 'Toutefois, les dispositions du § 1 du present article peuvent etre declarees non applicables ..." MAE 648(rev.)/57

' Toutefois, les dispositions du paragraphe 1 du present article peuvent etre declarees inapplicables:...'

1 - Compatibility, Efficiency, Legal Security

185

APPENDIX II: EXTRACT (ENGLISH VERSION) FROM THE OPINION OF LAGRANGE, AG, IN THE BOSCH CASE49 'This concerns a basic fault in the system brought about by Articles 85 et seq. of the Treaty, namely, the unsuitability of the administrative provisions for the application of the substantive rules. Underlying the substantive law is a clear connection between the rules denning the types of prohibited agreement, in Article 85 (1), and those which enumerate the conditions in which the prohibition may be declared inapplicable in Article 85 (3); this emerges clearly from a reading of subparagraphs (a) and (b) of paragraph (3). Logically, the same authorities or courts, in the course of the same proceedings, should have jurisdiction to decree in respect of one particular agreement, both on the compatibility of the agreement with Article 85 (1) and on the declaration of inapplicability of the prohibition under Article 85 (3). Moreover, the existence of this connecting link, and its presence in the minds of the authors of the Treaty, is clearly demonstrated by the terms of Article 88, which provide that, during the period of transition, relevant municipal legislation shall apply concurrently with the Treaty legislation contained in Articles 85 and 86, since the text underlines the necessity in this concurrent application of paying due regard to Article 85 (3) in particular. Furthermore, it would have been contrary to the most elementary considerations of justice to allow the application of the prohibition contained in Article 85 (1), with the sanction of automatic nullity which attaches to it, with all the consequences which the courts could, and perhaps should have drawn therefrom, without allowing the enterprises concerned to avail themselves of the provisions of Article 85 (3). It is for this reason that the doctrine known as 'de I 'exception legate', a basic principle of French law, was particularly well suited to the needs of the application of the Treaty, even during the transitional period. It would, in fact, ensure that the question of prohibition and that of any possible exemption from prohibition would be examined at the same time by the same authority, and if necessary by the same court; there would be no difficulty in respect of the effects of the automatic nullity. The same doctrine, in my view, was the only one capable of justifying the immediate application of Article 85 in countries which otherwise have no anticartel law; the ordinary law courts must be as competent to apply Article 85 (3) as they are to apply Article 85 (1). According to the opposite point of view, i.e. that which would require a decision having constitutive effect for the application of Article 85 (3)—which has been particularly current in Germany and has always been the view of the 49

Case 13/61 De Geus v Bosch [1962] ECR-111.

186

Working Paper VI

Commission—it was clear that the immediate application of Article 85 was impossible as long as the competent national authorities were not empowered to take the decisions required, according to this theory, to give effect to Article 85 (3). This is the argument put forward by the German Government in paragraph IV of its written observations: 'In accordance with Article 88 of the Treaty, the application of Article 85 (3) is entrusted provisionally to the authorities of Member States of the European Economic Community. This being the case, the national law of each State must decide which is the competent authority.' It was thus for Member States which had no appropriate legislation to set up the necessary procedures themselves, but when this was not done, there could be no question, according to this theory, of applying Article 85 in the transitional period in such states, since the operation of Article 85 (1) without the corresponding operation of paragraph (3) is inconceivable. It has even been said that the impossibility of applying this part of the Treaty in those Member States which have no appropriate laws makes it impossible to apply it in the Community at all; this, however, seems to me to go too far. This principle most in conformity with the spirit of the Treaty is that it should come into operation, as and when possible, at the same time as the national legislation. In fact, the controversy is now over since the Regulation has given the second theory the force of law, and it must be recognised that this accords best with the text of Article 85 (3). 'The provisions of paragraph (1) may however be declared inapplicable in the case of any agreement, etc' The theory depending on 'I'exception Idgale' would have required a different text, for example: ' "The provisions of paragraph (1) shall be deemed not to apply..." or simply "shall not apply . . . " It cannot be contended that the Regulation is illegal as being contrary to the Treaty in this respect; to declare such a Regulation illegal would be a step of such gravity that this Court, in my view, ought not to contemplate it except when the illegality is quite plain, which in this case is far from being so.' Nonetheless, the paradox to which I have drawn attention remains, since, unlike the position under Article 65 of the ECSC Treaty, exclusive power is given only in respect of the exemption from prohibition, and not in respect of the decision as to the incompatibility of agreements with the Treaty and their consequent nullity. It is for this reason that the Regulation has laid down a series of provisions designed to ensure that Articles 85 and 86 are applied as a whole and as consistently as possible. The key to the system is the obligation on the part of undertakings wishing to obtain a declaration of inapplicability under Article 85 (3), to notify the Commission of their agreements. The Commission may make a favourable decision with an effect retroactive to the date of notification; Article 4 (1) and Article 6 (1). 'Notification' in this instance comes close to

1 - Compatibility, Efficiency, Legal Security

187

'request for authorisation', and 'declaration of inapplicability' to 'authorisation'. ('The provisions of paragraph (1) may be declared inapplicable . . .'.); despite different terminology, the system plainly draws inspiration from Article 65 (2) of the ECSC Treaty. Under this system—I have been speaking up to now only of the definitive system applying to agreements made after the entry into force of the Regulation, and of the system under the national law, since there is both a transitional system and a moreflexiblesystem designed to favour certain types of agreement—the consistency of application seems to be adequately safeguarded. Nullity of an agreement which is contrary to article 85 (1) may always be pleaded before national courts, and, even if a 'notification' is in progress, the courts can rule on the compatibility of the agreement with Article 85 (1), and if necessary, can characterise it as automatically void, at least for the period prior to notification since it is quite plain that a subsequent 'declaration of inapplicability' can never be retroactive beyond that date. Nonetheless, the national court would do well to await the result of the proceedings before the Commission in order to ascertain whether or not the nullity might be effective after the date of notification. Is the court legally bound to do so? In my view this cannot be the case in the absence of clear provision. In particular, Article 9 (3) of the Regulation, cited above, seems not to allow such a conclusion. Moreover, there is a strong case for leaving such a decision to the discretion of the national judge, since it is not hard to imagine cases where it is clear that Article 85 (3) is not applicable, and where, after the discovery of the facts and the commencement of the action, 'notification' is made purely with a view to delay. From a legal point of view, the real difficulty would occur when the national court decides that an agreement is prohibited under Article 85 (1) while the Commission (or, possibly, this Court), giving a later decision takes a different view and as a result regards a declaration of inapplicability under Article 85 (3) as being without object and thus nugatory. This, however, is an inevitable result of a system of concurrent jurisdictions. As regards those agreements which benefit from the specially favourable system provided for by Article 4 92) of the Regulation, the national court should show particular caution when the Commission has been notified; a declaration of inapplicability under Article 85 (3) by the Commission in respect of these agreements may be made retroactive to any date at the discretion of the Commission, and accordingly it may antedate the notification (Article 6 (2) of the Regulation). We are left with those agreements which were in existence at the date of entry into force of the Regulation, for which special provision was made in Articles 5 and 7. Provided they are notified before 1 August 1962 (or 1 January 1964 in the case of agreements benefiting from Article 4), 'the prohibition contained in Article 85 (1) shall apply only for a period fixed by the Commission' to such agreements, provided that the undertakings concerned put an end to agreements 'or modify them in such manner that they no longer fall within the prohibition contained in Article 85 (1) or that they satisfy the requirements of

188

Working Paper VI

Article 85 (3)'. This means that the Commission;s decision may deprive the prohibition of any retroactive effect, preserving the full effect of the agreement in the past, and even in the future, if the parties are given time to bring their agreement within the law; in such a case, the provisions with regard to nullity, which ordinarily have a retroactive effect (ex tune) are completely ineffective. Here again, no express provision of the Treaty or of the Regulation requires the court to stay proceedings even when notification has taken place, or takes place during the proceedings, but clearly in such cases national courts would have to show great caution in view of the possible consequences of a subsequent favourable decision on the Commission's part. However, national courts are required to take account of the following provision of the Regulation (Article 7(1) infine)should such a case arise: 'A decision by the Commission pursuant to the foregoing sentence shall not apply as against undertakings and associations of undertakings which did not expressly consent to the notification.' This, then, is the way in which, in my submission, the concurrent jurisdiction of the national courts and the Commission in the application of Article 85 (1) and (2), and the exclusive jurisdiction of the Commission in respect of Article 85 (3) must be interpreted. No doubt the result is not altogether satisfactory, but that is the consequence of the double legal compromise which underlies the Treaty, as I interpret it: (i) a compromise between the doctrine of 'Pexception legale', which is the only one wholly compatible with the principle of automatic nullity contained in Article 85 (2), and the doctrine of 'constitutive effect' which as in German law, should logically be accompanied by a power given to the cartel authorities to 'declare ineffective' ('fur unwirksam erklaren') agreements contrary to the law—which involve a concept very different from automatic nullity; (ii) a compromise on jurisdiction, which the Treaty does not settle and which the Regulation, either by design or from supposed lack of power, did not confer exclusively on the Community authorities and Court as is done in the ECSC Treaty.'

1 - Compatibility, Efficiency, Legal Security

189

APPENDIX III: THE BOSCH JUDGMENT50 In this judgment, according to Professor Mok, the Court has taken the view that the authorisation system flows directly from the Treaty.51 In order to understand the arguments pro and contra it is necessary to first present the judgment. Bosch had entered into a number of exclusive contracts within and outside Germany for the distribution of its products. Besides the exclusivity binding the manufacturer, the contracts contained a clause prohibiting the distributor from exporting. One such distribution contract had been concluded in 1903 between Bosch and Van Rijn with respect to the Dutch market. In 1959 and 1960 de Geus, a third party, imported into the Netherlands Bosch refrigerators purchased from German distributors bound by the export ban. Bosch and Van Rijn brought an action in the Rotterdam Court against de Geus applying for a declaration that the dealings by de Geus were illegal, an injunction and damages. Contrary to de Geus' pleadings, the Rotterdam Court held that Article 85 was not effective to annul agreements incompatible with it. The Hague Court of Appeal referred the case to the Court of Justice, which was asked to rule whether the export ban was void by virtue of Article 85 (2). The Court ruled: 'The judgment of the Court of Appeal of The Hague raises the question whether Article 85 has been applicable from the time of entry into force of the Treaty. T h e answer of this question must in principle be in the affirmative. Articles 88 and 89 of the Treaty, which confer powers on the national authorities and on the Commission respectively for the application of Article 85, presuppose its applicability from the time of entry into force of the Treaty. Articles 88 and 89 are, however, not of such a nature as to ensure a complete and consistent application of Article 85 so that their mere existence would permit the assumption that Article 85 had been fully effective from the date of entry into force of the Treaty and in particular that the annulment envisaged by article 85 (2) would have taken effect in all those cases falling under the definition of Article 85 (1) and in respect of which a declaration under Article 85 (3) had not yet been made. In fact, Article 88 envisages a decision by the authorities of member States on the admissibility of agreements only when the latter are submitted for their approval within the framework of the laws relating to competition in their respective countries. Article 89, while conferring on the Commission a general power of surveillance and control, enables it t o take note only of possible violations of Articles 85 and 86 without clothing the Commission with power to grant declarations in the sense of Article 85 (3). Neither of these two Articles, moreover, contains transitional provisions dealing with agreements existing at the moment the Treaty came into force. It may be noted, furthermore, that the authors of the First Regulation Implementing Articles 85 and 86 of the Treaty (OJ 204/62) took the same view.

50 51

Case 13/61 De Geus v Bosch [1962] E C R 45. M o k (1999, 314).

190

Working Paper VI

The combined effect of Article 6 (2) and Article 5(1) of this Regulation is that the Commission, first, is able to make declarations under Article 85 (3) in respect of agreements subsisting at the time of entry into force of the said Regulation; and secondly is empowered, to give such declarations a retroactive effect, even to a date prior to that on which a particular agreement has been notified to the Commission. It follows that the authors of the Regulation seem to have envisaged also that at the date of its entry into force there would be subsisting agreements to which Article 85 (1) applied but in respect of which decisions under Article 85 (3) had not yet been taken, without such agreements thereby being automatically void. The opposite interpretation would lead to the inadmissible result that some agreements would already have been automatically void for several years without having been so declared by any authority, and even though they might ultimately be validated subsequently with retroactive effect. In general it would be contrary to the general principle of legal certainty—a rule of law to be upheld in the application of the Treaty—to render agreements automatically void before it is even possible to tell which are the agreements to which Article 85 as a whole applies. Moreover, in accordance with the text of Article 85 (2), which in referring to agreements or decisions 'prohibited pursuant to this Article' seems to regard Articles 85 (1) and (3) as forming an indivisible whole, this Court is bound to admit that up to the time of entry into force of the First Regulation Implementing Articles 85 and 86, the nullifying provisions had operated only in respect of agreements and decisions which the authorities of the Member States, on the basis of Article 88, have expressly held to fall under Article 85(1), and not to qualify for exemption under 85 (3), or in respect of which the Commission has taken the decision envisaged by Article 89 (2).' Mok's argument is that by stating in relation to a situation preceding and therefore not covered by Regulation 17, that the authors of the First Regulation

Implementing Articles 85 and 86 of the Treaty took the same view, the Court has made it clear that the authorisation system stems directly from the Treaty. The problem with this argument is that the same view in question was not about the authorisation system at all. The Court was confronted with a question whether, in the absence of an implementing regulation, the agreements contrary to article 85 paragraph 1 were void under paragraph 2. The Court considered that nullity could only be declared when there is the possibility of benefiting from the exemptions, if they are applicable. In this regard, because these articles are not of such a nature as to ensure a complete and consistent application of Article 85, the mere existence of Articles 88 and 89 could not lead, according to the Court, to the conclusion that Article 85 had been fully effective from the day of the entry into force of the Treaty. In these conditions, to accept that an agreement contrary to Article 85 paragraph 1 was void implied the possibility that the nullity could last for an indefinite period of time and that the agreement could subsequently be validated with retroactive effect. This result was held inadmissible. According to the Court, the authors of Regulation 17, took the same view on the inadmissibility of this result. The Court derives this commonality of view from the regime reserved by Regulation 17 to pre-existing agreements. If these agreements are 'notified' before a certain date, they can be exempted by the Commission with effect even prior to the date of

1 — Compatibility, Efficiency, Legal Security

notification. That must mean, according to the Court, that those agreements were not void. In relation to this reasoning, one can argue about whether Regulation 17 had actually assumed that pre-existing agreements were not void, or whether the Court was building the foundations of its doctrine of provisional validity as set forth in the second part of the judgment (not quoted above) concerning the period after the coming into effect of Regulation 17. In any event, I fail to see the relevance of this judgment to the question of whether the authorisation system is in-built in the Treaty.52

52

Cf. Ehlermann (2000, 23, point 55).

191

192

Working Paper VI

APPENDIX IV: THE PORTELANGE JUDGMENT53 Professor Mok is reported (after the publication of his article quoted in the previous appendix) to attach even more importance to the judgment in Portelange than to the Bosch judgment. An exclusive distribution agreement, entered into before Regulation 17 came into effect, had been correctly 'notified' to the Commission. Because the manufacturer had breached the agreement, the distributor sought damages before a national court. The latter referred to the Court of Justice a question about whether the provisional validity established in Bosch included the right to sue for damages. The Court answered in the affirmative. In reaffirming the provisional validity doctrine it did not mention the fact that this was a preexisting agreement, thus giving the impression that any 'notified' agreement, regardless of the date of closing, benefited from provisional validity.54 The Court said: '9. Article 85 of the Treaty is arranged in the form of a rule imposing a prohibition (paragraph (1)) with a statement of its effects (paragraph (2)) mitigated by the exercise o f a power to grant exceptions to that rule (paragraph (3)). 10. To treat a given agreement, or certain of its clauses, as automatically void presupposes that that agreement falls within the prohibition of paragraph (1) of the said article and that it may not benefit from the provisions of paragraph (3). 11. In order t o permit the parties to invoke the provisions of Article 85 (3), regulation N o 17 lays down that the agreements or decisions mentioned in Article 85 (1) must be notified t o the Commission. 12. In the case of an agreement notified under Regulation N o 17, the mere fact of its notification cannot imply that the agreement falls under the prohibition imposed by Article 85(1). 13. The question whether such an agreement is in fact prohibited depends on the appraisal of economic and legal factors which cannot be assumed to be present in the absence of an explicit finding that the individual agreement in question not only contains all the factors mentioned in Article 85 (1), but does not qualify for the exemption provided by Article 85 (3). 14. So long as such a finding has not been made, every agreement duly notified must be considered valid. 15. Although the fact that such agreements are fully valid may possibly give rise to practical disadvantages, the difficulties which might arise from uncertainty in legal relationships based on the agreements notified would be still more harmful. 16. If the Commission considers that the implementation of an agreement notified infringes the competition rules, it is open to it to adopt, within the appropriate time, a decision either under Article 85 (3) of the Treaty or under Article 15 (6) of Regulation N o 17.

53

Case 10/69 Portelange v Smith Corona [1969] E C R 309. Only in Case 48/72 Brasserie de Haecht II [1973] E C R 85, did the Court make it clear that provisional validity concerned only preexisting agreements. 54

1 - Compatibility, Efficiency, Legal Security

193

17. Where Article 15 96) abovementioned has been applied, the persons concerned are put on notice that the Commission considers that the conditions for the application of Article 85 (1) are fulfilled and that the application of Article 85 (3) is not justified. 18. Hence, any parties who proceed, from then on, with the implementation of the agreement do so at their own risk. 19. It must therefore be concluded that the agreements mentioned in Article 85 (1) of the Treaty, duly notified in accordance with Regulation No 17, are of full effect so long as the Commission has made no decision under Article 85 (3) and the provisions of the said regulation.'

One might suppose that Mok's argument is based on the description of Article 85 contained in the first passage of the quotation. The argument could be that the reference to a power to grant exemptions implies an authorisation granted by an administrative authority. Mok probably has a point there. It should be recognised that the word grant, and even more the French equivalent octroi, seem to imply an administrative decision. These words have in this respect the opposite connotation of the word used by the Treaty, that is, declared, analysed above under the heading 'Drafting arguments'. However, to draw any solid conclusion from such a choice of language seems disproportionate. First, the intention of the Court was only to describe the provision of the Treaty without drawing any legal effect from its choice of words in these circumstances: it would be difficult to infer that the word used by the Court should prevail over the word used by the Treaty. Second, that description appears in a context that excludes that the Court was paying attention to any inference on the question debated today. Last but not least, the Court was influenced in its choice of language by the existence of Regulation 17, that in 1969 had become a stable feature of the legislative environment and which was an integral part of its reasoning about provisional validity.55 As a foundation on which to base the incompatibility of a legal exception system, this judgment appears rather thin.

55

Similarly in the two Metro cases and in van Landewyck (Fedetab) (point 176) the Court speaks of the competences granted to the Commission by Article 85 paragraph 3, thus combining in one sentence the provisions of the Treaty and those of the Regulation in a practical, albeit strictly speaking incorrect, way. [Case 26/76 Metro v Commission (Metro I) [1977] ECR 1875; Case 75/84 Metro v Commission (Metro II) [1986] ECR 3021; Van Landewyck (Fedetab) [1980] ECR 3125.]

194

Working Paper VI

APPENDIX V: THE LANCOME JUDGMENT56 The German Government57 cites this judgment as evidence that the Court interprets Article 81 as providing for a prohibition with the possibility of authorisation. It does not develop the argument but the page of the Court reports to which it refers is the one in which the Portelange judgment (examined in Appendix IV) is quoted with approval. In Lancdme, the selective distribution agreements between the perfume manufacturer and its distributors that pre-existed Regulation 17 were correctly 'notified' to the Commission. In 1972, the Director General for Competition sent a letter to the firm stating that, after the agreements had been amended, the Commission considered that there was no longer a need for it to take action in respect of the agreements under Article 85 (1). Later Lancome and its Dutch subsidiary brought proceedings before the Haarlem Arrondissementsrechtbank against a chain of retail shops, asking that the court enjoin the unauthorised sale of their perfumes. The defendants relied on the nullity of the agreements under Article 85 (1). Among the several questions referred to the Court of Justice, one concerned the effect of the administrative letter about provisional validity. To this question, the Court answered that the letter had put an end to the provisional validity: '14. With a view to answering that question, the considerations which form the basis of the case-law of the Court on "provisional validity" should be recalled. 15. As the Court has observed in particular in its judgment of 9 July 1969 in Case 10/69 Portelange v Smith Corona Marchant International [1969] ECR 309, Article 85 of the Treaty is arranged in the form of a rule imposing a prohibition (paragraph (1)) with a statement of its effect (paragraph (2)) mitigated by the exercise of a power to grant exceptions to that rule (paragraph (3)). To treat a given agreement, or certain of its clauses, as automatically void presupposes that the agreement falls within the prohibition of paragraph (1) of the said article and that it may not benefit from the provisions of paragraph (3). 16. The exclusive power to apply Article 85 (3) which is conferred on the Commission by Article 9 (1) of Regulation N o 17, considered in conjunction with the provisions laid down in favour of old agreements by Article 6 (2) and 7 of that regulation, has led the Court to conclude that, in the case of those old agreements, the principle of legal certainty in contractual matters requires that, where the agreement has been notified in accordance with the provisions of Regulation N o 17, a court may only declare it to be automatically void after the Commission has taken a decision by virtue of that regulation. 17. In the light of those considerations, it appears that the maintenance of the pro-' visional protection from which notified old agreements benefit is n o longer justified from the date on which the Commission informs the parties concerned 56 57

Case 99/79 Lancdme [1980] ECR 2511. Point 1.1.

1 - Compatibility, Efficiency, Legal Security

195

that it has decided to close thefileon the case concerning them. After the adoption of such an attitude, which indicates that the Commission does not contemplate taking an individual decision on the notified agreements in question, it is unlikely that the Commission would subsequently exercise in favour of those agreements its power to apply Article 85 (3) with, where appropriate, retroactive effect for the period prior to their notification, as permitted by Article 6 (2) of Regulation No 17. There is, therefore no longer any reason to release national courts, before which the direct effect of the prohibition in Article 85 (1) is relied upon, from the duty of giving judgment.' This judgment adds nothing to Portelange. Reference is therefore made to the remarks about that judgment in Appendix IV.

VII Mitsuo Matsushita A Japanese View on the Modernisation White Paper Published by the European Commission

I. The significance of the White Paper for countries other than the Members of the European Union The proposals made in the White Paper on the modernisation of the competition law of the European Communities address issues that are unique to Europe. However, in some respects, they are also relevant to issues with which other countries, including Japan, are confronted. This paper is intended to comment on the White Paper from a Japanese viewpoint. 'Deregulation' of government control of business activities is a common phenomenon in industrialised countries including Europe, the United States, Japan and others. In earlier times, some sectors of the economy (such as, inter alia, insurance, shipping, telecommunications, gas, electricity, aviation, railways and trucking) were under tight governmental controls. Generally the business sectors that were under the governmental control were exempted from the application of competition laws. This was certainly true in Japan. In Japan, not only the areas mentioned above, but also export and import cartels, small business cartels, depression cartels and rationalisation cartels, were exempt from the application of the Antimonopoly Law (the AML). In the regulatory framework for exemption, activities such as export and import cartels and depression cartels were exempted from the application of the AML when they obtained authorisation from the regulatory agencies. The agencies in charge varied from one sector to another. In export and import cartels, it was the Ministry of International Trade and Industry (the MITI) while in depression cartels and rationalisation cartels it was the Fair Trade Commission (the FTCJ) that issued authorisations. When an application for authorisation was filed with the relevant administering authority, it reviewed the contents of the application to make certain that the requirements for exemption were satisfied. When it found that the requirements were present, an authorisation was granted. The holder of an authorisation enjoyed an exemption from the prohibition of the AML within the scope of the exemption, and it was assured that there would be no application of the prohibition of the AML as regards the activities covered by the exemption. However, deregulation of business activities has become one of the most important economic policies of the country in the last decade or two and laws vesting power in governmental agencies to provide authorisations of activities and exemptions of such activities from the AML have been largely abolished.

198

Working Paper VII

Through a recent amendment of the AML, the natural monopoly exemption from the application of the AML was abolished. Natural monopolies are said to exist in areas such as gas, electricity and railways, and there are generally some regulatory laws that provide administering authorities with the power to grant licenses. When a license is given, the application of the AML is exempted. Although there are still a few such laws, the number of authorisation laws is clearly declining. The legal development in Japan is somewhat similar to, although not exactly the same as, that with which the European Commission is confronted today. When authorisation laws granting administering authorities with the power to authorise some activities are abolished, provisions of competition laws prohibiting anti-competitive activities apply directly to them. Some such activities may be prohibited per se and some others might be approved when some conditions are met. However, whether they come under the prohibition or they are approved has to be decided ex post by courts or administrative agencies without any prior authorisation. Enterprises might face legal uncertainty due to the fact that there is no prior authorisation of their activities, and activities that have already commenced might be challenged under competition laws by an administrative agency or a private party. Comparing the legal situation in Europe (as described by the White Paper) and that in Japan, one can see that there is a similarity between the two systems in which regulatory laws are being abolished and the provisions of competition laws are becoming more and more directly applicable. Especially relevant to the situation in Japan is the way in which administering authorities such as the European Commission, the ECJ, those in Member States and national courts, deal with legal uncertainty when European Commission's monopoly of authorisation of activities which come under Article 81 (1) but meet the requirements of Article 81 (3) of the EC Treaty is abolished.

II. What is the nature of authorisation under Article 81 (3)? Article 81 (1) of the EC Treaty states that an agreement between undertakings that might affect trade between Member States and that prevents, restricts or distorts competition within the common market is prohibited. On the other hand, Article 81 (3) states that the prohibition of Article 81 (1) is inapplicable to an agreement that comes under Article 81 (1) if that agreement contributes to improving the production or distribution of goods, or to promoting technical or economic progress while allowing consumers a fair share of the resulting benefit. It also requires that the agreement does not: (a) impose on the undertakings concerned restrictions that are not indispensable to the attainment of these objectives, and (b) afford such undertakings the possibility of eliminating competition in respect of a substantial proportion of the products in question.

1 - Compatibility, Efficiency, Legal Security

199

Regulation 17/62 of the Council of Ministers (6 February 1962) gives the European Commission the exclusive power to authorise an agreement that satisfies the requirements of Article 81 (3). When authorisation of an agreement is granted by the Commission, that agreement is exempted from the application of the prohibition of Article 81 (1). In other words, if an agreement comes under Article 81(1), it is not exempt unless an authorisation is given by the Commission under Article 81 (3). This seems to mean that, as a matter of logical sequence, an agreement exempted from the prohibition of Article 81 (1) by virtue of authorisation under Article 81 (3) comes under Article 81(1) first and is then exempted from the application of Article 81 (3) by virtue of Article 81 (3). One question is how to conceptualise the relationship between Article 81(1) and Article 81 (3). An agreement that comes under Article 81 (3) is one that does not afford the parties the possibility of eliminating competition in respect of a substantial proportion of the products in question. This should mean that the agreement in question does not eliminate competition in respect of a substantial proportion of the products in question and the relevant market remains competitive. If the agreement does not substantially eliminate competition in a market, why is it unlawful under Article 81 (1) in the first place? Under the current system, if an agreement which satisfies the requirements of Article 81 (3) is not filed with the Commission and, therefore, no authorisation is granted, is this agreement illegal under Article 81(1) even though the agreement does not substantially restrain competition? The above are some of the questions that occur to the writer of this paper when thinking about the relationship between Article 81 (1) and Article 81 (3) together with Regulation 17. This seems to raise a question as to the nature of an exemption under Article 81 (3): does Article 81 (3) exempt from the prohibition of Article 81 (1) an agreement that restricts competition substantially and is originally unlawful under Article 81 (1) on the ground that it serves an objective different to that of competition policy, although it amounts to restraint of competition in the commerce among the Member States? If so, Article 81 (3) 'creates' a legal situation in which an agreement that would be held to be unlawful but for Article 81 (3) is exempted from the prohibition and is made lawful. This interpretation could be called 'the creation theory'. Or does an exemption under Article 81 (3) merely confirm that an agreement that come under that article does not essentially infringe upon the objective of Article 81(1) since it does not eliminate a substantial part of competition in respect of the product covered by the agreement? In this interpretation, an exemption is merely a confirmation that the agreement in question does not substantially restrain competition with respect to a product and is therefore not unlawful. This interpretation could be called 'the confirmation theory'. It seems that the relationship between Article 81 (1) and Article 81 (3) can be better explained by the confirmation theory than by the creation theory. As stated above, if an agreement does not substantially eliminate competition, it causes little problem in relation to competition policy. If an exemption creates

200

Working Paper VII

a legal situation in which an agreement that would otherwise be held unlawful is made lawful, it would mean that the agreement substantially restrains competition and is objectionable from the competition policy viewpoint but is exonerated by virtue of a policy (such as environmental protection or product safety) different from the competition policy. A natural reading of Article 81 (3) and Article 81(1) seems to lead to the conclusion that an agreement covered by Article 81 (3) is justified for the reason that it contributes to rationalisation of production or distribution, that consumers share the benefit thereof, that competition with regard to this product still remains with respect to a substantial portion of the market for this product, and that, as a consequence, there is no objection to it from a competition policy viewpoint. In Japan, there are some laws that provide for exemption from the application of the Antimonopoly Law (the AML). Although the number of such laws is decreasing in the current process of deregulation, there are still some laws providing for exemptions of some activities such as export, rate fixing by insurance companies and joint-purchasing and selling by small trade associations composed of small enterprises. There are two schools of thought regarding the nature of exemption from the AML, that is, 'the confirmation theory' school and 'the creation theory' school. Commentators belonging to the former argue that laws granting exemptions from the application of the AML merely confirm that the activities covered by such laws do not constitute violations of the AML. Meanwhile commentators who belong to the latter school argue that the nature of the exemption is the creation of a legal situation in which regulatory laws grant authorisation to activities that would constitute a violation of the AML. One of the important differences between those two schools of thought is that, according to the confirmation theory, activities that satisfy the requirements for exemption can be interpreted as not constituting a violation of the AML even if they have not beenfiledwith the administering authority. On the contrary, the creation theory holds that activities that are covered by exempting laws should be held to be unlawful unless they have been filed with the administering authority in advance and have been granted authorisation from the authority. The confirmation theory seems to apply to some regulatory laws providing for exemptions from the AML and the creation theory explains the situation better with regard to other laws. In the area of small business, the Small and Medium Business Cooperatives Law provides for the establishment of trade associations composed of small and medium enterprises. Article 24 of the AML provides for exemptions from the application of the AML for activities engaged in by such associations, provided that their activities do not substantially restrict competition and unduly raise prices. Examples of activities of such trade associations include the joint purchasing and joint selling of small enterprises through trade associations. Joint-purchasing activities of small enterprises through trade associations make it possible for small enterprises to purchase, for example, raw materials on a large scale and thereby contribute to the overall reduction of costs. As long as

1 - Compatibility, Efficiency, Legal Security

201

the activities of such trade associations do not substantially restrict competition, it is hard to see why such activities constitute a violation of the AML. Authorisation might be necessary to make sure that such activities satisfy the requirements of both the Small and Medium Enterprises Cooperatives Law and the AML. However this endorses the view that the nature of Article 24 of the AML granting exemption is a confirmation that activities covered by both laws do not constitute a violation of the AML. If activities are carried out without prior filing with the administering authority and without obtaining authorisation, it might constitute a violation of the filing requirements. However this is different from a violation of the substantive provisions of the AML. Some other laws granting exemptions from the application of the AML can be better explained by the creation theory. Depression cartels were an example. Article 24 (4) of the AML exempted depression cartels from the application of the AML when they met the requirements provided for in Article 24 (4), filed with the FTCJ, and authorisation was given by the FTCJ. The requirements for depression cartels were, inter alia, that there was a serious depression as defined in the article, the participants limited their activities within the necessary minimum, and that the depression cartel agreement did not restrict entry to and withdrawal from the agreement. To obtain an exemption, there was no requirement that such depression cartel agreements did not substantially restrict competition in a relevant market. Indeed, a depression cartel would not effectively be enforced unless the cartel covered a major portion of the product that was the subject matter of the cartel and competition in the relevant market was substantially eliminated. Therefore, this exemption envisaged that depression cartels would substantially restrict competition and would collide with the objective of promoting competition as enunciated in the AML. The nature of the exemption in this instance was that, although depression cartels were anti-competitive, they were necessary on the basis of industrial policy considerations. This should be regarded as a deviation from the principle of competition policy as incorporated in the AML. This exemption could only be explained by the creation theory. Article 24 (4) of the AML granting exemptions for depression cartels was abolished by the amendment of 1999 and no depression cartels are currently allowed. Compared with the Japanese legal situation as described above, activities covered by Article 81 (3) of the Treaty seem to fall under a category that can be explained by the confirmation theory better than by the creation theory. This is because an important requirement for the exemption provided for in that article is that competition regarding the product in question is not substantially restrained. The White Paper proposes that a prior authorisation system be abolished and replaced by a system where the application of Article 81 (3) is decided ex post by courts and agencies. The premise of this proposal is that administering agencies and courts can judge an agreement to be lawful under Article 81 (3) without requiring a prior authorisation. This change will be accomplished by abolishing or amending Regulation 17/62 without amending the basic statutory

202

Working Paper VII

mandates, that is, Articles 81 (1) and 81 (3) of the Treaty. Therefore, there will be no change in the relationship between Article 81 (1) and Article 81 (3). To the writer of this paper, this means that an authorisation under the current Regulation 17/62 should amount to nothing but recognition that an authorisation merely confirms the lawfulness of an agreement if covered by Article 81 (3). This should endorse the idea that even a current authorisation is a confirmation of lawfulness of an agreement covered by Article 81 (3), and the introduction of ex post control as opposed to a system of prior control will not change this basic structure.

III. From prior control to ex post control The abolition of the prior notification and authorisation framework under Article 81 (3) and Regulation 17/62 will introduce ex post application of Article 81 (3) by the European Commission, administering authorities of Member States and courts. It also introduces the possibility that administering authorities and courts could apply Article 81 (3) without any prior authorisation from the European Commission. In other words, a direct application of Article 81 (3) could be made without authorisation by any agency with expertise in competition policy matters. The direct applicability of Article 81 (3) as proposed by the White Paper might have several major consequences. First, it will bring into the picture a plural number of enforcement agencies deciding what Article 81 (3) should mean. Not only the European Commission, but also administrative agencies and courts of the Member States, will decide whether or not an agreement falls under Article 81 (3) and is exempted from the application of Article 81(1). This will bring about a decentralisation of the power to decide the applicability of this article. One consequence might be a risk that different courts and administrative agencies of Member States will come up with different answers with regard to the applicability of this article on the same subject matter. Compared with a situation in which the European Commission has the sole power to decide whether or not Article 81 (3) applies, the diversified system advocated in the White Paper might create a situation where there exist different answers as to whether or not Article 81 (3) applies to a certain subject matter. It seems that, for the sake of legal certainty, it is important to address this concern. However, considering experiences within other jurisdictions, this problem does not seem to be insurmountable. In the United States, for example, administering agencies (the U.S. Justice Department and the Federal Trade Commission) and federal courts apply provisions of antitrust laws. There is a possibility that different courts will come up with different interpretations with regard to, for example, Section 2 of the Clayton Act as amended by the Robinson-Patman Act. In Japan, the FTCJ is empowered to apply a provision of the AML and civil

/ - Compatibility, Efficiency, Legal Security

203

courts can apply the same provision. In fact, there have been instances where interpretations by the FTCJ and a civil court of a provision in the AML differed. It seems that this has not been a major issue in either jurisdiction. Diversity in interpretations of a legal provision and a consequent degree of legal uncertainty is, up to a certain point, an inevitable consequence of introducing a plurality of agencies with decision-making power. However, such diversity in not uncommon in other jurisdictions. As long as there exists an ultimate authority which issues a final interpretation about a provision of law, different interpretations of a provision of law will gradually be integrated into a unified interpretation. In the European situation, the role is given to the European Court of Justice as the ultimate court empowered to decide issues, as it is to the Supreme Court in the United States and the Supreme Court in Japan. In the interim period, an important task is to reduce and minimise—if not entirely eliminate—different and possibly conflicting interpretations of Article 81 (3). A block exemption published by the EU authority will certainly provide a common basis for interpreting Article 81 (3). Guidelines issued by the European Commission will also introduce authoritative interpretations of Article 81 (3), even if such guidelines are not binding on the agencies and courts of Member States. In Japan, the FTCJ has published a number of guidelines on a variety of subjects such as distribution and trade practices, patent and knowhow licensing, joint R & D research and others. Although they are not binding in the same way as a legal provision is binding, they nevertheless provide useful guides to courts and they are generally followed by courts when dealing with provisions of the AML. Under an ex post control system as proposed by the White Paper (which is opposed to a prior control system), an application of Article 81 (3) is made after an agreement has been signed and put into practice. In an ex post control system, there is always a touch of uncertainty in that an agreement is held not to come under Article 81 (3) after the agreement has already been put into effect. In order to avoid this situation, as indicated earlier, block exemptions and guidelines published by the European Commission will serve as useful guidance to judges, administrative officials and corporate lawyers. However, new issues always arise and block exemptions and guidelines can never cover the variety of situations that might occur in the future. It makes a lot of sense to establish a mechanism through which, prior to putting them into practice, enterprises can consult enforcement agencies regarding their proposed agreements to make sure that they would not infringe legal prohibitions. In Japan, the FTCJ published a number of guidelines and, in some of these, a prior consultation mechanism is established. For example, the 'Guidelines on Activities of Trade Associations' state principles that the FTCJ applies when controlling activities of trade associations. One part of the guidelines establishes a consultative mechanism. When a trade association intends to carry out a program that might have implications for competition, it engages in consultation with the FTCJ as to the lawfulness or unlawfulness of the program and requests its view. The FTCJ

204

Working Paper VII

responds to the request based on the information provided by the trade association; it states that the program is not a violation of provisions of the AML or that it is likely to be held to be violating a provision of the AML, as the case may be. When the answer is that the program in question is likely to be held to be in violation when put into effect, the trade association can either abandon the program or modify it so that there is no longer the risk of violating the law. When the FTCJ responds to a request for its view with regard to a project that it is not violating any provision of the AML, the FTCJ is obliged not to proceed against this project as long as there is no change in the facts surrounding the project. In an ex post control system regarding the applicability of Article 81 (3), as proposed by the White Paper, a consultation mechanism of a sort described above seems to be useful. In a decentralised enforcement system such as the one proposed by the White Paper, there should be a uniform format by which answers to requests for consultation will be dealt with. Agencies in charge of responding to such requests should speak in one voice, that is, answers to a question as to the applicability or non-applicability of Article 81 (3) should be the same (or at least similar) regardless of who is answering the question. Close co-operation and co-ordination among agencies of the Member States and the European Commission seems to be essential in reaching an understanding on the common format for such consultation. In a system in which Article 81 (3) applies ex post control as opposed to prior control, the role of corporate counsels advising companies regarding will increase. It is essential for enterprises planning to enter into an agreement that might be covered by Article 81 (3) to receive advice from corporate lawyers as to the applicability of this article.

IV. Conclusions If the ideas expressed in the White Paper are put into practice, the role of judiciary in the enforcement of competition laws in Europe will increase. As has been the case in Japan, it might have been true that civil courts in Member States of the European Union have played a relatively minor role in the enforcement of competition laws in the past. Under the new system, the role of courts could increase dramatically as regards Articles 81 (1) and 81 (3). It seems that the role of the European Commission as administrative agency controlling and directing business activities through advance regulation will decrease simultaneously. As far as Article 81 is concerned, its major function will primarily be that of a prosecuting agency. When the Commission decides that an agreement falls under Article 81 (1) and Article 81 (3) concurrently, the Commission will simply refrain from initiating a proceeding against that agreement by utilising its prosecution discretion.

1 - Compatibility, Efficiency, Legal Security

205

If an ex post control system is introduced—as advocated by the White Paper—this will prompt administering authorities and courts enforcing Article 81 (1) and 81 (3) to take an analytical approach in which Article 81 (1) and Article 81 (3) are looked at as an integrated whole, that is, they will consider Article 81 as an integrated whole. In fact, it might be that Article 81 (3) is an integral part of Article 81(1), and not merely an exception to Article 81(1) that should be interpreted narrowly. This change might expand the scope of application of Article 81 (3) and this might mean that there is more of a 'rule of reason' analysis under those provisions. Does this mean that the principle of control of some of agreements will change? In other words, does this mean that there will be a change from an approach that agreements are generally prohibited with some exceptions, to a new approach that whether or not agreements are prohibited or permitted should be decided on a case-by-case basis? Not all agreements are anti-competitive. This is certainly true with some vertical agreements such as exclusive dealing arrangements. They might in fact promote inter-brand competition. This can be true even with some horizontal agreements. Consider, for example, an agreement among competitors to engage in a joint R & D project where their market share is small and where such R & D is likely to make them stronger competitors vis-d-vis their dominant competitor(s). Considering all such factors, this approach might make a better sense as a matter of competition policy. The new system proposed by the White Paper will reduce interventions of public authorities into business activities in the form of prior authorisations, and it will promote the scope of freedom to choose their own course enjoyed by enterprises. Enterprises will have to assess by themselves the legality of agreements that they might like to enter into. On the other hand, enterprises will not be able to use authorisation as a shield to ward off legal challenges that might be raised against them. The lack of prior authorisation might increase the risk of running counter to the prohibition of Article 81(1). However, legal risk is an essential feature of a deregulated and decentralised economic system.

VIII Petros C. Mawoidis and Damien J. Neven* The White Paper: A Whiter Shade of Pale of Interests, and Interests

Abstract This paper does not seek to evaluate whether decentralisation of the implementation of Article 81 EC Treaty is desirable but simply analyses how the decentralisation envisaged in the White Paper would operate. We identify two potential issues associated with the sharing of enforcement among several authorities. First, we observe that simultaneous enforcement by several authorities is likely to occur in the proposed framework and that each Member State will have little incentive to take the interests of other Member States into account in its decisions. We show that such a system of enforcement can have a 'disintegrating effect', to the extent that it does not allow for a balancing between positive and negative net benefits across Member States. We suggest that, in order to align incentives, two additional measures should be considered. We (a) advocate the re-emergence in the intra-EC context of a 'positive comity' obligation and we (b) suggest that a formal procedure for arbitration between different institutions should be laid down (as in the US). Second, we observe that the main source of inconsistency across countries is associated with different institutional frameworks to the extent that authorities responding to different institutional constraints might balance interests in different ways. We suggest that institutional constraints like accountability and independence standards should be imposed on Member States. Finally, drawing on the US experience with multiple enforcement, we argue that the role of the Commission should be as much to manage regulatory innovation (arising from the enforcement activity of Member States) as to resolve conflict.

* The authors would like to express their gratitude to Claus-Dieter Ehlermann, Celine Gauer and Henrik Horn of Rantzien for numerous discussions on this issue, and particularly to Eleanor Fox and Diane P. Wood for guiding them through some of the intricacies of US antitrust procedures.

208

Working Paper VIII

I. Introduction and overview This paper focuses on the decentralisation of European Community ('EC') competition law as spelled out in the EC White Paper of April 1999 on Modernisation of EC competition law (hereinafter 'the White Paper').1 At the outset, it might be worth emphasising what the White Paper does not do. In particular, the White Paper does not put into question the allocation of competencies as laid down in Article 81 of the EC Treaty, which is an exclusive competence. The Commission, through the White Paper, simply proposes that competence with respect to competition policy which has been transferred to the Community could from now be exercised by both Community and national authorities (national competition authorities and national courts). It is consequently inappropriate to subject the White Paper to Article 5 (subsidiarity) and to its legal consequences. The White Paper is thus not concerned with the allocation of competencies. It reorganises enforcement; what is envisaged by the White paper is limited work-sharing where the Commission retains the monopoly over individual exemptions (see § 92, p. 32—see also Ehlermann (2000)). Some of the implementation2 of Article 81 is entrusted to at least3 16 players, rather than one. This paper does not seek to evaluate whether decentralisation of the implementation of Article 81 is desirable. Rather, we analyse how the decentralisation envisaged in the White Paper would operate. In particular, we identify three potential issues associated with the sharing of enforcement among several authorities. First, and most fundamentally, different authorities might not have the same incentives in exercising their power; national authorities will indeed not have an incentive to take into account effects taking place outside their own territories. Second, different authorities will respond to different institutional constraints and might balance interests in different ways in response to these constraints. Third, different authorities might have different levels of competence and some might accordingly make more numerous mistakes than others.4 1

See, European Commission, The White Paper on Modernisation of the Rules Implementing Articles 85 and 86 of the EC Treaty, Commission Programme No 99/027, 28 April 1999. We recognise that there is some complementarity between decentralisation of enforcement and other aspects of the modernisation contained in the White Paper, in particular the treatment of vertical restraints. In a sense both endeavours aim to take off some of the burden from DG Comp's shoulders. More fundamentally, the issue of 'who will be applying EC competition law from now on?' is not independent of the design of the laws themselves. In what follows, we abstract from this issue. 2 Arguably, rather limited responsibilities are delegated to Member States under the current proposal. 3 In each Member State, both the antitrust agency and courts could implement Art 81 EC Treaty. 4 There may be an additional concern, namely that regulatory diversity among Member States with respect to a number of ancillary but influential tools for antitrust enforcement might again introduce some inconsistencies across countries. For a

1 - Compatibility, Efficiency, Legal Security

209

We argue that the first concern is potentially quite serious and we show that divergent incentives can have a 'disintegrating effect', particularly given the scope for simultaneous rulings on single cases offered by the framework currently envisaged and despite the fact that rulings from individual Member States only apply within the confines of their own territories. In the proposed framework, it is unlikely that Member States will routinely opt for a balancing test whereby they will internalise interests of other Member States (be they foreign consumers or producers as the case may be) in their decisions. By contrast, such balancing is guaranteed when the Commission is in charge. In our view, the existing legal tools are inadequate to ensure that incentives will be aligned. In particular, the duty to co-operate, as laid down in the EC Treaty and Article 9 (3) of Regulation 17/62, does not sufficiently curtail national discretion. We suggest that, in order to align incentives, two additional measures should be considered. First, we advocate the re-emergence in the intra-EC context of the 'positive comity' obligation, as we know it from the field of co-operation in the international realm. Second, we suggest that a procedure for arbitration between different institutions should be laid down. As in the US Supreme Court system, we suggest that the European Court of Justice (ECJ) should be the ultimate arbitrator with respect to enforcement of EC competition law. As in the US, the circumstances where the ECJ will have the right or the obligation to arbitrate should also be negotiated ex ante. The second concern, namely that institutions will differ across countries, is also one that we believe should be taken seriously. Besides incentive issues, differences in institutional constraints are probably the main source of 'inconsistencies' in the implementation of competition law. Unlike other areas where the law is codified in details, competition law is formulated in general terms and accordingly leaves a lot of discretion to the implementing authorities. 5 However discretion is associated with reduced accountability that in turn enlarges the scope for various forms of capture. In our view, the case law of the Community does not, as the Commission suggests in the White Paper, provide enough discipline in this respect.6 As a result of different institutional frameworks, different Member States could thus strike different balances between the interests of their various constituencies. In order to reduce this source of inconsistency, it is necessary to impose institutional constraints on the Member States. These may take the form, for instance, of accountability standards (like common publication requirements) and standards of independence (for discussion of this issue and whether, as a consequence, some minimum harmonisation is warranted in this area, the reader is referred to Kon (1999), Mersing (1999) and Forrester (1999). 5 From this perspective, the area of competition is probably not the prime candidate for delegation to Member States. 6 It is not clear either, as discussed in Section 3, that case law should act as strong discipline.

210

Working Paper VIII

instance, the status of civil servants or the nomination of competition Commissioners). The third concern, that of competence, is not one that we emphasise. After all, incompetence is hardly a structural issue. Its effects are likely to be transitory and they do not have a 'disintegrating effect' (incompetence is, in its crude form, origin-blind). To sum up, if the Commission has recognised some of the concerns just discussed, the positive proposals that it has put forward to deal with what it terms 'risk of inconsistencies' are in our view insufficient (they are essentially limited to information sharing). We believe the White Paper falls short of providing a comprehensive framework to analyse the underlying issues. The sharing of responsibilities that it envisages is unlikely to work well without important accompanying measures. In addition to information sharing, positive comity obligations, the establishment of procedures for arbitrage by the ECJ and institutional constraints should also be considered. The rest of the paper is divided as follows: Section 2 deals with the issue of multiple enforcement. Focusing on the legal aspects of the White Paper, we conclude that the occurrence of simultaneous enforcement should not be dismissed. In Section 3, we examine the consequences of such simultaneous enforcement and show that, under the current proposal, it has disintegrating effects. In Section 4, we provide a positive account of the US' experience with respect to multiple enforcement, from which useful lessons can be drawn. Finally, in Section 5, we argue that the existing framework—as well as the Commission's proposals—do not suffice to avoid undesirable outcomes associated with multiple enforcement, and we put forward a few proposals that should improve matters.

II. Multiple enforcement (the 15+1 scenario) This Section focuses on the legal parameters that the White Paper envisages for competition enforcement in the EC. Wefirstobserve that the White Paper is not subjected to the constraints imposed by subsidiarity. Second, we show that the White Paper leads to multiple enforcement.

1. The White Paper and Subsidiarity As we mentioned earlier, the White Paper should not be viewed as a proposal within the context of subsidiarity. Nowhere does the Commission submit that the question of competence is prejudged by the White Paper. The Commission simply adds new partners to exercise an exclusive Community competence. This is confirmed by the fact that it is still the Commission alone that will define the

1 - Compatibility, Efficiency, Legal Security

211

EC's competition policy (and not in some form of co-operation with the Member States).7 Hence, competition policy post-White Paper is a Community competence to be co-exercised by DG COMP and by national competition authorities (NCAs) as well. This is a substantive innovation in the sense that institutions (the NCAs) are called to apply a legal instrument over which they have no formal jurisdiction. Effectively, under the White Paper, the NCAs will operate as administrative units of an EC-wide executive. This is the first instance, to the best of our knowledge, in which Community administration is meant to operate within the Member States. Thus the importance of this reform cannot be overstated. The White Paper, by being insulated from subsidiarity-type considerations (as laid down in Article 5), still ensures that Community action will not be subjected to the constraints laid down in Article 5. Hence, the Community is unrestrained and can intervene when it deems it necessary.8 Finally, the White Paper does not deal at all with issues regarding application of national competition law. Indeed, this remains a question of national competence and the White Paper does not alter the pre-existing paradigm of allocation of competencies.

2. EC competition law: here, there and everywhere As is well known, Article 81 applies to the extent that an agreement 'may affect trade between Member States'. The ECJ has traditionally interpreted this term lato sensu.9 Voices arguing in favour of a restrictive construction of the clause have remained largely in the minority,10 whereas the majority of doctrine seems to endorse the ECJ's conclusions in this respect.'' As long as the Community is the sole enforcer, the evaluation of whether trade among Member States is affected is purely a Community-wide issue. For the purpose of exercising jurisdiction, the question of which EC countries are involved in intra-community trade is simply irrelevant. However, when national enforcers are involved, the issue is more intricate. Indeed, some countries might be unaffected (in terms of trade) and the question arises as to whether they should still be allowed to exercise jurisdiction. For example, could the Portuguese NCA assert competence over an agreement

7

See the White Paper at p. 31. On whether this conclusion is in itself a problem, see infra Section 4. 9 It should be noted that in US law, the interstate commerce clause has also been interpreted in a wide sense, the leading Supreme Court decisions being Summit Health v Pinhas, 500 US 322 and McLain v Real Estate Board, 444US 232. 10 See, for example, Wesserling (1997). 1 ' Ehlermann (2000) is concurring. 8

212

Working Paper VIII

between an Italian and a Danish undertaking that only affects trade between those countries? In our view, the answer to this question must be 'yes'. The Portuguese NCA should not be asked to establish some 'minimum contacts' other than having satisfied the 'may affect trade between Member States' requirement. The reason is simply that the White Paper only enlarges the set of potential enforcers without altering the mechanism that triggers jurisdiction. Hence, in principle, we can expect frequent multiple enforcement as several NCAs can simultaneously assert their rights even with respect to events occurring outside their territories.12 In turn, one could expect interested parties to select enforcers strategically in order to advance their own interests. This is sometimes referred to as 'forum shopping', an issue that has been raised but not fully explored in the literature so far.13 Instances of multiple enforcement are easy to illustrate. Imagine a case where country A and country B assert jurisdiction over the same practice and jointly decide not to intervene. This could be, for example, a case involving an alleged horizontal agreement between Greek and Italian carriers operating in the sea transport sector on the routes between Greece and Italy.14 Independently of whether the Greek and Italian authorities intervene or not, Portugal could assert jurisdiction, either because our preferred interpretation is correct—that is, that any NCA can intervene to the extent that trade among Member States is affected—or because Portuguese tourists were charged monopoly prices by the Greek and Italian undertakings concerned. Of course, multiple enforcement could lead to conflicting decisions.15 But even in the absence of conflicts, it is not clear that multiple enforcement will be appropriate as the involvement of several agencies might lead to an outcome that is undesirable. These issues are further analysed in Section III. The extent of conflict and the consequences of multiple enforcement are also affected by the scope of the decisions that NCAs can take. According to the White Paper (§ 60), decisions by NCAs will have a limited territorial effect in the sense that they are deprived of any legal effect beyond national boundaries. This approach can lead to the following oxymoron: a transaction that, by definition, affects trade among Member States (otherwise EC law does not come into play at all) must be submitted to one NCA knowing ex ante that, no matter what the decision is, it is binding only within a part of the common market. 12

Even if our conclusion is contested, it is legally impossible to doubt the validity of the statement that in our example, Italy and Ireland can, in principle, assert jurisdiction based on the 'nationality' clause and eventually any other Member State whose consumers' interests are affected based on the 'effects' doctrine. 13 Compare Ehlermann (2000) and Siragusa (1999). 14 Any resemblance to actual cases is completely unintentional. 15 It is irrelevant for the purposes of this exercise whether NCAs and the Commission are on an equal footing with respect to the kind of decisions they can take post-White Paper. As Ehlermann (2000, p. 30) observes, it is most likely that NCAs can adopt positive decisions. What we care about here is quid in case where two NCAs reach divergent decisions independently of how much they can decide.

1 - Compatibility, Efficiency, Legal Security

213

As Nehl (1999) points out, in the name of decentralisation the Commission effectively puts into question the network concept that it wants to create. Territorial limitation can lead to a series of perverse incentives: not to submit to NCAs if EC-wide protection is sought; outlaw an otherwise valid transaction only within the four corners of a particular sovereignty; raise transaction costs to the maximum extent possible since, conceivably, fifteen different outcomes are possible in a particular case.16

III. The consequences of multiple enforcement In the previous section, we argued that (i) Member States will not be seriously constrained in asserting their enforcement rights so that multiple simultaneous enforcement can be (widely) expected, but (ii) that the scope of their decisions will be limited to their own territories (at least according to what appears to be the working hypothesis of the White Paper). In this Section we analyse the consequences of such a framework for enforcement. Wefirstconcentrate on incentive problems before turning to institutional issues.

1. Simultaneous multiple enforcement and incentives To the extent that national authorities are accountable to national constituencies, it is natural to assume that they will only consider the interests of those constituencies. In other words, national authorities cannot be expected to take into account the effects that are taking place outside their territory in their evaluations of the cases that they handle. Each country will thus consider both the competitive effects and the potential efficiency benefits that accrue within its own territory. Since, as discussed above, a number of countries can be expected to assert their enforcement rights, a likely outcome is one where a number of countries will simultaneously assess the competitive effects and the efficiency benefits that accrue within their own territories, and each will adopt a ruling that has effects in its own territory. One can then wonder whether such an outcome would significantly differ from that obtained if there was a single enforcer at the EU level. For the sake of argument, assume that each country can impose remedies meeting its own concern without affecting the agreement under review in other countries (in terms of competition or efficiency benefits). In such a hypothetical world, the deal under review is effectively 'separable' across countries and it 16

Ehlermann (2000), Nehl (1999) and Siragusa (1999) have identified some of the externalities mentioned here.

214

Working Paper VIII

is not clear that a single enforcer would achieve an outcome that is fundamentally different from that arising from simultaneous enforcement across countries.17 Such an authority would indeed recognise that each country can be assessed separately and would impose appropriate remedies in each country. When remedies in one country affect other countries, the matter is altogether different. For instance, assume that a particular agreement only makes business sense if it can be implemented in all countries. In such a case there is indeed a potential external effect across countries because a negative decision in one country will effectively prevent the deal from being implemented in other countries where the deal might possibly bring positive net benefits (that is, where efficiency benefits dominate potential anti-competitive effects). The outcome of simultaneous multiple enforcement might then differ significantly from that arising from a single EC-wide enforcer. Consider, for instance, a deal that brings positive net benefits at the EC level but is such that the balance between anti-competitive effects and efficiency is unfavourable in one country. An ECwide enforcer would, in all likelihood (that is, observing the Community Interest Clause), allow (or fail to sanction) the deal. By contrast, under multiple enforcement the deal will be banned (or sanctioned) by the country which suffers. Hence, whereas an EC-wide single enforcement allows for a balancing between positive and negative net benefits across countries, multiple enforcement does not. In general, simultaneous multiple enforcement also imposes more numerous constraints than single EC-wide enforcement (net benefits have to be positive in 15 subsets of the EC and not only at the level of the EC as a whole). As a consequence, simultaneous multiple enforcement should be expected to lead to more prohibitions (or sanctions) than EC-wide single enforcement. It will also be biased against deals that have an EC-wide scope but that have uneven consequences across the EC. From that perspective, simultaneous multiple enforcement will thus have a 'disintegrating effect' relative to the current situation. So far, we have considered—following the White paper—that rulings by NCAs only have effects within their own territories. It is worth noting that our main conclusion (namely, that simultaneous multiple enforcement can have a disintegrating effect), remains valid even if this assumption is relaxed. Indeed, if it assumed that rulings by NCAs have effects throughout the EC (in a modified White Paper scenario), even external effects across countries will be reinforced. In those circumstances, a negative decision by one authority will always prevent other countries from realising the net benefits that would potentially accrue to them. To sum up, it appears that, except in rather exceptional circumstances, simultaneous multiple enforcement should be more restrictive than single enforce17

In the absence of efficiency benefits, the scope for conflicts would also be reduced. Indeed, when the relevant antitrust market extends to a few countries, multiple enforcement by any subset of the countries concerned will yield the same outcome as single enforcement by a central authority (see Neven and Roller, 2000).

/ - Compatibility, Efficiency, Legal Security

215

ment at the EU level, and it will be biased against deals which foster EU-wide market integration.

2. Multiple enforcement and institutions As discussed above, allowing several institutions to implement the same rule might give rise to inconsistencies, even if we abstract from the issue of incentives and simultaneous enforcement just discussed. That is, even if only one institution is in charge of a particular case and all institutions have the same objectives, the identity of the institution that is actually in charge might affect the outcome. If we neglect systematic differences in the level of competence across the 16 players, the institutional constraints faced by these institutions will be the main source of divergence. Indeed, it has long been recognised that institutions should not be seen as benevolent and omniscient agents following the mandate that has been assigned to them. Civil servants will take decisions in terms of their own objectives—which should include the objective assigned by the law, but might also encompass others (like career motives)—and third parties will naturally seek to exert influence on the decision. The extent to which civil servants will actually deviate from pursuing the objectives that have been assigned to them (which is usually referred to as 'capture'), will also depend on the institutional framework. For instance, greater accountability should generally reduce the extent of capture. Other features, like independence, will involve more delicate trade-offs (see, for instance, Neven et al., 1992 for a discussion). In turn, greater accountability will be easier to achieve if the mandate given to civil servants is precisely codified. The implementation of rules can indeed be relatively easily verified ex post. By contrast, the implementation of general principles that allow for wide discretion is harder to monitor. Competition statutes in general, and Article 81 EC Treaty in particular, are formulated in very general terms and leave a lot of discretion to the agency in charge of implementing them. As a result, accountability is difficult to achieve in the area of competition and it will be difficult to effectively monitor the operation of several agencies. Accountability is further weakened if NCAs' decisions have territorial limitations, since national civil servants will then be more sheltered from EC-wide scrutiny. From that perspective, it is rather odd that competition is the first area where multiple implementation of Community policy is envisaged. Other areas, like agricultural policy or research and development policy, are substantially more codified than competition and should a priori be better suited to multiple implementation. However, if multiple implementation is considered in the area of competition, the accountability of NCAs should be enhanced to the maximum extent possible. This could take the form of harmonised publication requirements. It

216

Working Paper VIII

could also take the form of a regular review of competition decisions by an independent body. But harmonised accountability standards might not suffice because, as discussed above, only limited accountability {ex post) can be achieved in an area like competition. Hence, it would seem necessary to impose standards ex ante on particular features of the national institutions.

IV. Inside the US: the 50+1 laboratory Parallels with the US experience are sometimes striking and sometimes less so. In what follows we do not recommend an institutional transplant (a 'lock, stock and barrel' approach). However some aspects of the US experience are quite relevant for the purposes of our present discussion and they should not be lightly overlooked. In the United States, multiple enforcement occurs across different circuits of the Federal Courts.18 To illustrate the US approach, we will focus on a simulation whereby two federal courts are called to judge, first, on the same issue and, then, on a comparable issue.19 Imagine that an undertaking sues another undertaking before two Federal Courts. Both suits can proceed simultaneously. However, one of the parties can ask to transfer the case [using 28 USC 1404(a) or 1406(a)] in which case the two suits will be consolidated. If no such request is tabled, the possibility still exists for a party to request that the Court stay proceedings until the other suit is resolved. In such a case, it would be up to the Court to decide whether it will do so or not. If nothing results from the above, judgment in one of the two suits will the have force of resjudicata and, on proper motion, the winning party can dismiss the second suit. This is so essentially because judgments govern the actions of the parties in general, not where they are acting. At least for Federal Courts, there is nothing that limits the force of, say, a 7th Circuit judgment to the 7th Circuit. As mentioned above, the working hypothesis of the White Paper stands in contrast with this approach. A similar result could stem from a case where multiple plaintiffs sue the same defendant before various circuits. The Judicial Panel of Multidistrict Litigation (composed of federal judges, see 28 USC 1407) can order that all the cases be considered in one district for pre-trial proceedings. When litigation reaches the 18

Our discussion here does not focus on state law in the same way that the White Paper is not concerned about national competition law. As in the EC architecture, there are State Agencies and a Federal Agency. No formal links are established between the two and co-operation is on a voluntary basis. Notwithstanding this, some rather spectacular outcomes are the result of such co-operation (the Microsoft litigation being probably the best illustration of the sort). 19 This part of the paper is largely based on discussions with Eleanor Fox and Diane P. Wood.

1 - Compatibility, Efficiency, Legal Security

217

trial stage, it can be transferred back to various districts. There is a complicated doctrine, called non-mutual offensive issue preclusion, under which it is possible that particular facts found in one case against one party can be taken as established in a later case against the same party. If comparable cases reach various districts, there is absolutely no guarantee that they will all end up with the same result. In fact, Federal Circuits often disagree on issues of law. Perhaps the most famous and long-standing conflict was between the 9th Circuit (alone) and everyone else on the question of whether market power had to be shown by the Sherman Act Section 2 to attempt to monopolise a case. The Supreme Court finally granted certiorari and resolved this question in Spectrum Sports v McQuillan, 506 US 447 (answer—'yes'). As is explained infra, conflicting jurisprudence is top of the list for the Supreme Court to grant certiorari. What are the lessons to be drawn from this parallel? First, multiple enforcement occurs in the US. It arguably encourages innovation in interpretation of US antitrust law and this feature is much valued in US practice. Second, there are a number of procedural devices designed to encourage co-ordination between parallel enforcers. And, third, there is a final umpire who plays the important role of providing future guidance on the basis of diverse and arguably enriched experience. One should, however, be careful in applying lessons from the US to the European context. The main difference between the US and the EC lies with incentives. Whereas different circuits in the US do not have obvious incentives to concentrate on effects taking place within the territory of their jurisdiction, the same is not true for European NCAs that respond to domestic constituencies.

V. To Revamp the Post-Modern So far we have established that multiple enforcement is a real possibility in the post-White Paper era, and that it could lead to substantial adverse consequences. In what follows wefirstconsider whether the existing legal framework and proposals in the White Paper will themselves be sufficient to address the problem. We conclude that this is not the case and we advance a few suggestions to address the matter. The existing framework is in full mutation. The duty to co-operate imposed on Member States, and Article 9 (3) of Regulation 17/62 that provides the Commission with the opportunity to intervene when warranted, constitute the existing framework. In the White Paper, the Commission proposes an information-sharing system between NCAs and national courts on the one hand and the Community institutions on the other. We examine each component in turn.

218

Working Paper VIII

1. The duty to co-operate (Article 10 EC Treaty) One way to avoid non-co-operative outcomes associated with multiple enforcement is provided by Article 10 (previously Article 5). This instrument imposes a duty to co-operate on Member States; however, it appears to be rather ineffective. Article 10 imposes a double obligation on Member States: a positive obligation (to ensure fulfilment of the obligations arising out from the Treaty) and a negative one (to abstain from any measure that could jeopardise the attainment of the objectives of the Treaty). Lenaerts and van Nuffel (1999, at p. 419) have appropriately dubbed the duty to co-operate as the 'federal good faith'. It is well established that the duty to co-operate contains both a vertical angle (Member State to Community) as well as a horizontal one (Member State to Member State). Except in cases where a Member State fails to comply with specific EC obligations, the duty to co-operate has been interpreted as an obligation imposed on Member States 'to take all appropriate measures to guarantee the full scope of Community law.' (Lenaerts and van Nuffel,.1999 at p. 421) This means that Member States (NCAs for the purposes of the present paper) must not only co-operate with the EC institutions responsible for implementing EC law (NCA to DG COMP), 20 but also with institutions of other Member States (NCA to NCA).21 How far can we construe the extent of this obligation? We should keep in mind that Article 10 is normally invoked as an auxiliary basis to any given claim. By itself, it is thus a rather weak basis to carry a claim. That is, Article 10 offers a good argument when the violation of another specific obligation is alleged. But there is no such other alleged violation in the context of our discussion. More specifically, there is no obligation at all for a Member State to desist when another Member State has decided to exercise jurisdiction. Moreover, the White Paper itself acknowledges the potential for 'forum shopping', which means that the Commission does not construe Article 10 EC Treaty as imposing a duty to desist when another NCA has been requested to intervene. It seems fair to conclude that, with respect to the horizontal angle of the duty to co-operate, we should not expect too much when it is applied in the context of decentralised antitrust enforcement. Article 10 could be helpful when the Commission decides to intervene in a particular case (vertical angle). It is now settled case law that, once the Commission has initiated procedures and a fortiori when it has adopted a final decision, national courts are bound to avoid conflictive decisions (if necessary by suspending proceedings before them).22 As the White Paper notes (p. 35) the same principle could mutatis mutandis apply to NCAs as well. The duty to cooperate could serve as an argument for such an endeavour. 20 21 22

See Case 217/88 Commission v Germany [1990] E C R 1-2879 at §3 3 p. 1-2907. See for example, Case 165/91 Van Munster [1994] ECR l-A66\ at § 32 p. 1^1697. See C-234/89 Sergios Delimitis v Henninger Brau AG [1991] ECR 1-935.

1 - Compatibility, Efficiency, Legal Security

219

2. When the Commission intervenes (Article 9(3) of Regulation 17/62) According to Article 9(3) of Regulation 17/62 'As long as the Commission has not initiated any procedure under Articles 2, 3 or 6, the authorities of the Member States shall remain competent to apply Article 85 (1) and Article 86 in accordance with Article 88 of the Treaty.' The White Paper suggests that, based on this provision and in the case of a final decision by a national institution and subject to res judicata, the Commission can prohibit an agreement allowed by national courts (or NCAs); moreover, in the case of a non-final decision by a national authority, the Commission can intervene and propose its own preferred solution (White Paper pp. 35-6). However the Commission has not bound its discretion in this respect. PostWhite Paper, no one knows if, when and under what circumstances the Commission would make use of its discretion. This stands in stark contrast with the US approach where the Supreme Court has indicated the circumstances under which it is likely to intervene. The main drawback of unlimited discretion is that private parties are likely to prefer the security of a Commission decision. Ultimately this would defeat the whole purpose of the White Paper in terms of decentralisation. For decentralisation to succeed, guidance on the role of the umpire seems necessary. In other words, the institutional credibility of the NCAs largely depends on the extent to which they follow the Commission's legacy, and companies' anticipation of this will be greatly enhanced by guidelines on the circumstances in which the Commission would intervene. Rule 10 of the US Supreme Court addresses the subject of 'considerations governing review of certiorari' and provides a source of inspiration for future EC guidelines in this respect. It reads as follows: ' . . . Review on a writ of certiorari is not a matter of right, but of judicial discretion. A petition for a writ of certiorari will be granted only for compelling reasons. The following although neither controlling nor fully measuring the Court's discretion, indicate the character of the reasons the Court considers: (a) a United States court of appeals has entered a decision in conflict with the decision of another United States court of appeals on the same matter; has decided an important federal question in a way that conflicts with a decision by a state court of last resort; or has so far departed from the accepted and usual course of judicial proceedings, or sanctioned such a departure by a lower court, as to call for an exercise of this Court's supervisory power; (b) a state court of last resort has decided an important federal question in a way that conflicts with the decision of another state court of last resort or of a United States court of appeals; (c) a state court or a United States court of appeals has decided an important question of federal law that has not been, but should be, settled by this Court, or has decided an important federal question in a way that conflicts with relevant decisions of this Court.

220

Working Paper VIII

A petition for a writ of certiorari is rarely granted when the asserted error consists of erroneous factualfindingsor the misapplication of a properly stated rule of law.'

To the extent that the Commission is not the ultimate umpire (as the Commission's decisions can be appealed before the ECJ), the US institutional framework still differs from that found in the EC. It would seem desirable to enhance the legal security of Commission's rulings in its role as umpire. This could be realised by ensuring that cases comparable to those mentioned in the US legislation cited here could find their way before the ECJ. One possible solution would be for the Commission to request advisory opinions from the ECJ about such cases. What stems from the analysis above is that the existing framework is too weak and too imprecise. The EC Treaty was based on the assumption that competition law is enforced in a centralised manner. Hence, it did not pay much attention to the vertical relationship (NCAs to DG Comp). More precisely, there is nothing like an equivalent to ex-Article 177 EC Treaty to govern the relationship between national administrative authorities and European institutions.23

3. Information sharing In addition to existing instruments, the White Paper envisages the adoption of a wide information-sharing scheme. This would certainly help to reduce unintended inconsistencies in NCAs' decisions, but it certainly does not address the underlying issue of divergent incentives. The argument could, of course, be advanced that it is in the interest of NCAs anyway (even absent legal compulsion) to 'internalise' foreign interests in their decisions, since they are in some sort of repeated interaction with each other. However, practice offers examples where co-operation has broken down. BoeinglMcDonnell Douglas might be a case in point, where the EC moved to assert jurisdiction without paying much attention to the stated wish of its partner to exclusively decide the case. When stakes are high, the incentive to deviate might be hard to resist. Hence, it appears that the current legal framework, even in conjunction with information sharing, will not suffice to address the undesirable features of multiple enforcement.

4. Revamping the post-modern: taking care of some rough edges Since divergent incentives for NCAs constitute the main issue at stake, it is natural to try to align them better. Positive comity obligations on NCAs would certainly contribute to this, and we would advocate their consideration. 23

Arguments in this direction have been advanced by Siragusa (1999).

1 - Compatibility, Efficiency, Legal Security

221

In addition, as discussed above, differences in institutional constraints will be an important source of inconsistency across NCAs. Hence we would advocate the introduction of institutional constraints on NCAs. These might take the form, for instance, of accountability standards (like common publication requirements) and standards of independence (for instance with respect to the status of civil servants or the nomination of competition Commissioners). Positive comity obligations and institutional constraints are, however, unlikely to prove sufficient to avoid the negative consequences of multiple enforcement. Ultimately either an umpire or a formal rule to allocate jurisdiction will be necessary. Unfortunately, the second option does not completely address the fundamental issue of incentives. It does provide some legal security, especially if national decisions are given an EC-wide effect and parties to the dispute will avoid forum shopping. But it does not guarantee that the forum will provide a proper balancing between (eventually) conflicting interests. Such a decision would designate competence. Designated and demonstrated competence, however, do not overlap. Hence it seems that the role of umpire currently envisaged for the Commission should be formalised. In particular, the circumstances in which the Commission will intervene should be clarified ex ante. In our view, the US experience certainly suggests that the role of umpire is as much to distil diversity as to resolve conflict. Hence the focus of the White Paper must shift: instead of focusing on ex ante instruments like information sharing, the Commission should pay more attention to ex post instruments designed to ensure that gains from innovation are properly realised. Divergence is a risk, but diversity is an opportunity and the White Paper should recognise this.

IX Ernst-Joachim Mestmdcker The Modernisation of EC Antitrust Policy: Constitutional Challenge or Administrative Convenience?

In the present system of enforcement, Articles 81 (1) and 82 of the EC Treaty are administered by the Commission, national competition authorities and the courts of Member States. The Commission has exclusive power to grant Art. 81 (3) exemptions for agreements of which it has been notified. Without notification, the prohibition in Art. 81 (1) applies without consideration of Art. 81 (3). The Commission proposes to abolish the detailed rules for the implementation of Art. 81 (3) and its own exclusive power to grant exemptions.1 In the future, Art. 81 is to be applied as a whole by the Commission, national competition authorities and national courts.2 The prohibition rule and the exemption rule are to be tied together. On the proposal of the Commission, a Council Regulation is to extend to Art. 81 (3) the direct applicability formerly limited to Art. 81 (1).

I. Art. 81 (3) as a legal exception and the effet utile of competition rules 1. Control of abuse? Shorthand descriptions of the new system are bound to cause misunderstandings. High on the agenda of misunderstandings is the juxtaposition of legal exemptions and the control of abuse of restrictive agreements. Some endorse the legal exception system as being the opposite to a control of abuse system. Others criticise the legal exception system as having effects similar to a control of abuse system. A third view approves the new system of ex post deterrence because it is similar to the control of abuse by dominant enterprises (Wils 1999). Further uncertainties are inevitable when these concepts are used to examine the compatibility of the new system with the EC Treaty. The White Paper deals with horizontal restrictions of competition. They are usually referred to as 'cartels' and their treatment is a cornerstone of competition law and competition policy. 1

European Commission: White Paper on Modernization of the Rules Implementing Art. 85 and 86 of the EC Treaty, OJ C 132 of 12.5.1999, p. 1 (hereafter the White Paper). 2 White Paper no. 12.

224

Working Paper IX

For a comparison of different systems for the control of cartels it is helpful to distinguish their legal elements from their expected impact upon conduct and markets. The essential legal elements are the rules of substance, the sanctions and the remedies. The control of abuse system is predicated upon the principle that restrictive agreements are neither inherently good nor bad. They may be operated until found to be reprehensible. Sanctions apply to abusive conduct. They may include a decision declaring the underlying agreement to be void. Under Regulation 17/62, the control of abuse applies to exempted agreements (Art. 8 lit. d). Under the new system, there are no exemption decisions and, consequently, no provisions against abuse of exempted agreements. This is the consequence of the direct applicability of Art. 81 (3). The Commission proposes another kind of decision that implies a new kind of control of abuse. Its effects are compared to conditions and obligations in exemption decisions (Schaub 1999). This kind of control of abuse is conceived of as an alternative to a prohibition decision. It is to be based upon commitments of the undertakings concerned 'that would overcome the objections raised against their agreement'.3 The commitments are to be rendered binding upon participants and third parties, competition authorities and courts. Fines and periodic penalty payments are to apply if the undertakings fail to meet their commitments. This individual decision may be used, as noted earlier, as an instrument of market control based upon concerted action between undertakings and the Commission.4

2. The legal exception system Contrary to the existing system for the enforcement of Art. 81, the legal exception system is based upon ex post control of restrictive agreements.5 There is, according to the White Paper, no presumption that restrictive practices are void.6 That is, the burden of proof lies with the party who challenges the agreement. This is contrary to the jurisprudence of the European Court of Justice.7 The reach of the new Regulation is not limited to less harmful agreements. According to the text of the Treaty, and in the European Court of Justice's interpretation, there is no distortion of competition that by its nature cannot be exempted according to Art. 81 (3).8 Undertakings are said to be able to obtain 'immediate execution of their contracts before national courts with effect from the date of their conclusion provided the conditions of Art. 81 (3) are satisfied' (no. 78). This statement is correct insofar as undertakings may execute their 3 4 5 6 7 8

White Paper no. 90. Mestmacker (1999); see also Geiger (2000), at p. 165 and onwards. White Paper no. 48. White Paper no. 78. Joint Cases 43/82 and 63/82 VBVB and VBBB [1984] ECR 19, no. 52. Case T-17/93 Matra Hachette v Commission [1994] ECR 11-595, no. 85.

1 - Compatibility, Efficiency, Legal Security

225

agreements ab initio. It is misleading to relate this option to proceedings before national courts. As long as an agreement's validity is not challenged, there is no reason to go to court and not even standing for a declaratory judgment. The legal possibility of obtaining retroactive application of Art. 81 as a whole does not prevent the factual performance and operation of restrictive agreements. In the uncertain event of a challenge, the undertakings will have no difficulties in showing that they had good reason to believe that the conditions of Art. 81 (3) were satisfied. When Art. 81 (1) prohibits agreements and concerted practices that have as their purpose or effect the restriction of competition, it is apparent that implementation is to be guided by the effective preservation and maintenance of the competitive process. Concerted practices are covered in order to prevent cooperative restrictions of competition, irrespective of their legal nature. The major shortcoming of a legal exception system is the initial tolerance of restrictions of competition, empowering the undertakings concerned to decide for themselves whether these restrictions are exempted and are compatible with the objectives of the Treaty. For Art. 81 (3) to apply it is not enough that agreements promote the welfare of the participants; the agreements must be indispensable for the attainment of objective economic advantages recognised by the EC Treaty.9 When restrictions of competition are to be evaluated, the selfinterest of competing undertakings is a poor guide for the public interest. Adam Smith, who cannot be suspected of underestimating market forces or believing in regulatory intervention, pointed to the nature of the problem: 'I have never known much good done by those who affected to trade for the public good. It is an affectation, indeed, not very common among merchants, and very few words need to be employed in dissuading them from it.'10 As far as legal terminology is concerned, a prohibition of restrictive practices that applies as a matter of law without prior decision and a legal exception system that relies on ex post control may both be classified as prohibition systems. There are nevertheless fundamental legal and economic differences between the two. The European Court of Justice addressed these differences in the landmark van Gend & Loos ruling,11 establishing that the Commission's infringement procedures under Arts. 226 and 227 (previously Arts. 169 and 170) must be complemented by directly applicable individual rights. The major reason given was that infringement procedures brought after the implementation of an infringing Member State's decision were bound to be ineffective. The same rationale applies to the legal exception system. Even assuming—against realistic probabilities—that national courts will, in a number of cases and after one or more years, find an agreement void ex tune and award damages, the damage done to the competitive process cannot be remedied. More important and more 9 10 11

Cases 56 and 58/64 Grundig and Consten [1966] ECR 322, 396-7. Campbell, Skinner and Todd, eds. (1976), at p. 456. Case 26/62 van Gend & Loos [1963] ECR 5, 26.

226

Working Paper IX

realistic is that the deterrent effect of the private enforcement of Art. 81 in national courts is—as will be shown later—limited or negligible. When various commentators,12 the German Monopolies Commission13 and the former and present President of the German competition authority14 find that the new system has effects similar to a control of abuse system, they do not ignore the characteristics of the legal exception system, but find the degree to which this system tolerates restrictions of competition unacceptable. The legal exception system fails because it does not give full effect to the principles and purposes of Art. 81. Let me now turn to the related, but by no means identical issue of the new system's compatibility with the Treaty.

II. Modernisation The Commission characterises its proposals as 'modernisation of the implementation of competition rules'. Modernisation is a dialectic concept and not very enlightening as a policy guideline. It suggests both an enlightened mind and fascination with what is fashionable. The Commission's message for change refers to the growth of the Community to fifteen Member States, to economic and monetary union, to enlargement and globalisation. It is, however, less than obvious what bearing these truly global developments have upon the problem to be solved by modernisation: the Commission's administrative difficulties in handling the exemption system for notified agreements to the satisfaction of industry. The Commission's sense of being involved in a truly historical enterprise is nevertheless, even if inadvertently, justified, as it does not propose to remedy the shortcomings of the present exemption system, but rather to change the hardcore of Community competition law. The reform proposal raises important issues of Community law because it touches upon the fine borderline that separates the interpretation of the Treaty by the European Court of Justice and the implementation of the Treaty by the Commission and other Community institutions. The revision or modification of rules guaranteeing political or economic liberties point to even more profound and general issues. When such rules are adopted, their generality permits very different and frequently divergent expectations as to their interpretation and application. From the perspective of public international law, it is one of the revolutionary features of the EC Treaty that the task of interpretation has 12 Moschel (2000) pp. 61-67, Ritter (1999) at p 1485; Mestmacker (1999) pp. 523-529; Deringer (2000) pp. 5-11; Paulweber and Kogel (1999) pp. 500-515. 13 Monopolkommission (1999). 14 Wolf (1999) 'Perspektiven des europaischen Kartellrechts', available at http://www.bundeskartellamt.de/perspektiven.html, and Boge (2000) 'Das deutsche Kartellrecht in der europaischen Wettbewerbsordnung', FlW-Symposion 'Zukunft der Wettbewerbsordnung und des Kartellrechts', 9-11 March 2000, Innsbruck.

1 - Compatibility, Efficiency, Legal Security

227

been given to an independent court. Interpretation and application remove the 'veil of ignorance' that frequently made the adoption of the more fundamental rules possible in the first place. The reasons why reform and modernisation in this area are so delicate is the temptation to adjust the rules in light of past disappointments and open a Pandora's box of new interests and power balancing. The Community recognises this temptation by imposing the acquis communautaire on new Member States, and on old Member States as well. There is little doubt that the four freedoms and the rules of competition would not have been incorporated in the Treaty of Rome by the Member States, had they foreseen the pervasive effects these rules, as interpreted by the European Court of Justice, were to have on their political and economic order. Direct applicability and the supremacy of Community law over municipal law in cases of conflict are among the constitutional principles that were not foreseen by the fathers of the Treaty. At present, the reach of Community law is high on the critical agenda of constitutional reform in most Member States. So far uncontroversial is, however, the principle that the European Court of Justice's interpretation of the Treaty, including the qualification of Treaty rules as being directly applicable, is as binding upon Member States and Community institutions as the Treaty itself. When in 1991 the Court summarised its own understanding of the Treaty as a 'constitutional charter', that ruling explicitly included the system of undistorted competition as part of the Community legal order.15 And when the Court invested the justiciable state-related Treaty provisions with direct applicability in the landmark van Gend& Loos ruling,16 competition rules were the only provisions that justified the argument of the Court that the Treaty recognised 'explicitly' individual rights and duties. In line with this jurisprudence, the Court repeatedly held that not even Community Directives or Regulations can take away the rights and duties of citizens resulting from a Treaty provision of direct applicability. This has been held to be particularly true with respect to Art. 81 (1) and (2) and the implementing Regulation no. 17/62.1 quote: 'Whilst, in defining the powers of the Commission, Regulation no. 17, and in particular Art. 7 thereof enable the Commission to take into account the general principle of legal certainty, it did not modify—as indeed it could not—the effects of Art. 85 (2) [now Art. 81 (2)], but on the contrary, by Art. 1 thereof it confirmed that without prejudice to arts. 6,7, and 23 thereof agreements, decisions and concerted practices of the kind described in Art. 85 (1) [now Art. 81 (1)] shall be prohibited no prior decision to that effect being required' (emphasis added).'17 The principle that a Regulation as a matter of law cannot interfere with the direct applicability of Treaty provisions takes precedence over the issue of expediency, effectiveness and administrative convenience. By the same token, 15 16 17

Opinion 1/9 [1992] ECR 1-6079, 6102, nos. 16 and 21. Case 26/62 van Gend&Loos [1962] ECR 25. Case 48/72 Brasserie de Haecht v Wilkin [1973] ECR 77, 86, no. 6.

228

Working Paper IX

Regulations cannot confer direct applicability on Treaty provisions when the conditions for direct applicability spelled out in the Court's jurisprudence are not met.

III. The compatibility of the new system with the EC Treaty The Commission proposes the adoption under Art. 83 of a Regulation providing for the application of Art. 81 'as a whole', thus conferring direct applicability on Art. 81 (3). Such a Regulation is valid only if it is compatible with the system of undistorted competition and, more particularly, with Arts. 81 and 83. Authors who deny the compatibility of such a Regulation with the Treaty rely on the Treaty's provisions, on the history of Regulation 17/62 and on the Regulation's interference with direct applicability.18 For the history of Regulation 17/62, I rely on Deringer's authoritative account (Deringer 2000, 15). The Regulation was, it is true, the product of a political compromise. However, it does not follow that the interpretation of primary Community law was irrelevant to that compromise. On the contrary, the interpretation of Art. 81 as a provision that required no prior decision to that effect was even written into Art. 1 of Regulation 17/62. This is an important indication of the true meaning of Art. 81 without, however, being determinative and binding upon Community institutions today.

1. Giving effect to Art. 81 principles Regulations adopted on the basis of Art. 83 are intended to give effect to the competition principles enshrined in the Treaty. They cannot change these principles, nor can they modify the Treaty. The Treaty consistently distinguishes the application of the prohibition in Art. 81 (1) from the exemptions under Art. 81 (3). According to Art. 81 (3), the Art. 81 (1) prohibition 'may be declared inapplicable' in the case of any agreement or category of agreements. These words argue for a clear-cut separation of the prohibition that is directly applicable and the exemption that requires a declaration. The interpretation that Art. 81 (3) is directly applicable means that all restrictive agreements may be treated as valid until a negative decision has been issued and become definitive. In conformity with Art. 81, Art. 83 distinguishes between Regulations that are to ensure compliance with the prohibition laid down in Art. 81 (1) and 'the 18 In addition to the authors referred to above, see Wesseling (1999). Approving the Commission's proposals without going into the details of Treaty compatibility are Geiger (2000), p. 165 and onwards; Deselears and Obst (2000) at pp. 41-46; Bartosch (2000); Schaub A. and Dohms R. (1999).

I - Compatibility, Efficiency, Legal Security

229

detailed rules for the application of Art. 81 (3)'. More particularly, Art. 83 (a) provides for the adoption of Regulations making provisions for fines and periodic penalty payments to ensure compliance with Art. 81 (1) and Art. 82. Under a system of legal exemption, it is difficult to imagine how an infringement of Art. 81 (1) can be fined without a prior decision that Art. 81 (3) is inapplicable. The Commission, contrary to the text and the spirit of these Treaty provisions, argues that the prohibition rule of Art. 81 (1) may be implemented through an ex post control covering only restrictive practices that infringe Art. 81 as a whole. The risks of cartelisation are no longer the risks to cartel members but the risks to the public, to consumers as well as customers and competitors. While Art. 83 (b) empowers the Council to lay down detailed rules for the application of Art. 81 (3) providing for effective supervision and simplified administration, these implementing rules become irrelevant once the exemption is directly applicable. An agreement is either prohibited or it is found to be valid. If Art. 81 (3) is applicable, there is no opportunity to examine exempted agreements. A decision that Art. 81 as a whole is not infringed is a negative clearance. The Commission confirms that these decisions are of a declaratory nature and would have the same effect that a negative clearance has today.19

2. Direct applicability of Art. 81 (1) in the jurisprudence of the European Court of Justice The European Court of Justice interprets Art. 81 (1) as a prohibition with a proviso for an exemption. The direct applicability of Art. 81 (1) does not follow from Regulation no. 17/62, but from the Treaty itself. In the de Haecht ruling,20 the Court stressed that Regulation 17/62 did not and could not modify the effects of Art. 81 (1) and (2). The Court has consistently ruled that Art. 81 (1) and (2), as well as Art. 82, are by their very nature susceptible of having direct effects on the relationship of individuals, thus leading to individual rights that national courts must safeguard.21 In interpreting the concept of national authorities in Art. 9 (3) of Regulation 17/62, the Court distinguished between national competition authorities and national courts. Regulation no. 17/62 could not take away the competence of national courts because that would deprive individuals of rights conferred upon them by virtue of the direct applicability of Art. 81 (1) (nos. 15/17) of the Treaty. This principle has been confirmed by the Court. I quote: '... Since the prohibitions contained in Arts. 85 (1) [now 81 (1)] and 86 [now 82] tend by their very nature to produce direct effects in relations between individuals, those 19 20 21

White Paper no. 89. See above, at n. 17. Case 127/73 BRTv SABAM[\974]

ECR 51, 62.

230

Working Paper IX

articles create direct rights in respect of the individuals concerned which the national courts must safeguard. To deny, by virtue of Art. 9 of Regulation no. 17, the national court's jurisdiction to afford that safeguard would mean depriving the individuals of rights which they hold under the Treaty itself. It follows that the initiation by the Commission of a procedure under Arts. 2, 3 or 6 of that Regulation cannot exempt a national court before which the direct effect of Art. 85 (1) is prevented from giving judgment'.22

It follows that a Regulation like that contemplated by the Commission, providing for the application of Art. 81 (1) on the condition that the requirements of an exemption are present, is compatible with the Treaty only if it does not interfere with its direct applicability. That depends upon the direct applicability of Art. 81 (3). If the exemption, contrary to the Commission's assumption, does not satisfy the requirements of direct applicability, the tying of Art. 81 (1) to Art. 81 (3) would deprive Art. 81 (1) of direct applicability. The Court's jurisprudence on direct applicability of Treaty provisions is well settled. Treaty provisions are directly applicable if they are sufficiently clear and precise for judicial application. This includes interpretation. They must establish an unconditional obligation and the obligation must be complete and legally perfect, not depending upon measures to be subsequently taken by Community institutions or Member States with discretionary powers in the matter.23 Direct applicability means that the provisions of Community law must be given full effect in all Member States from their adoption and during the time of their validity.24 Provisions of national law that are incompatible with directly applicable provisions become 'inapplicable'. This effect of the supremacy of Community law over national law is binding for every court that, within its jurisdiction, has to safeguard the rights that Community law confers upon individuals. Justiciability of Community law rules is a necessary condition of direct applicability because thisfindinginvests national courts with the jurisdiction to apply Community law. If Art. 81 (3) does not satisfy the requirements of direct applicability, it is—for the reasons given—unconstitutional to tie the application of Art. 81 (1) to that of Art. 81 (3). Even a superficial reading of the exemption requirements in Art. 81 (3) suggests that courts are ill-equipped to apply them. They would have to find whether a restrictive agreement contributes to: a) improving the production or distribution of goods, or b) promoting technical or economic progress, and whether the agreements allow consumers a fair share of the resulting benefit without: a) imposing restrictions on the undertakings concerned that are not indispensable to the attainment of these objectives, and b) affording such undertakings the possibility of eliminating competition in respect of a substantial part of the products in question. 22 23 24

Case 127/73 Marty v Lauder [1980] E C R 2481, 2510, no. 13. Case 26/62 van Gend&Loos [1963] E C R 5, 25. Case 106/77 Simmenthal [1978] E C R 1978, 629, 643, nos. 17/18.

1 - Compatibility, Efficiency, Legal Security

231

These requirements must be met cumulatively. Undertakings that claim an exemption must, according to the European Court of Justice, present facts justifying the conclusion that an exemption may be granted. Undertakings claiming an exemption have an initial burden of proof that these conditions are met.25 The Commission, contrary to this jurisprudence, emphasises that under the new system there is no presumption that restrictive agreements infringe Art. 81 (I).26 The most important indications as to whether Art. 81 (3) is directly applicable are found in the cases reviewing Commission decisions granting, or refusing to grant, exemptions. The Court of Justice and the Court of First Instance have consistently held that in such cases the Commission is entitled to a wide margin of discretion. In its first leading case on the application of Art. 85 (3) [now Art. 81 (3)], the Court outlined the limits of judicial review.27 The Commission, the Court observed, must render difficult economic judgments. Judicial review of these judgments was restricted to a finding that the relevant facts are correctly ascertained and that the law has been correctly applied to these facts. In later decisions the principle of limited judicial review in Art. 81 (3) cases is spelled out in greater detail. The Court will examine only whether rules of procedure have been complied with, whether proper reasons have been provided, whether the facts have been accurately stated and whether there has been any manifest error of appraisal or misuse of powers.28 The Court, in other words, will not replace the Commission's expert judgment by its own and will not enter into the balancing of economic advantages and disadvantages of an agreement. The Commission argued that, even though Art. 81 (3) is predicated upon a margin of appreciation, it is nevertheless directly applicable because it does not provide for political discretion (Schaub and Dohms 1999). This is a distinction that has no basis in the Court's jurisprudence. In dealing with the limitations of judicial review in Art. 81 (3) cases, the Court starts with the observation that the Commission is responsible for the administration and direction of the Community's competition policy, calling for complex economic judgments. This applies to decisions in individual cases as well as to Regulations providing for exemptions under Art. 81 (3).29 In a study on behalf of the EC Commission, the authors reiterate the Court's jurisprudence as to the complexity of economic judgments and the consequently political character of such decisions.30 25 Joint Cases 43/82 and 46/82 VBVBIVBBB, ECR 1984, 19,68, no. 52; Case T 17/93 Matra Hachette [1994] ECR 11-595, 631, no. 104. 26 White Paper no. 78. 27 Cases 56 and 58/64 Grundig and Consten [1966] E C R 322, 396. 28 Case T 29/92 SPO [1995] ECR II 289, 382, no. 288; confirmed Case C-l 37/95 P [1996] ECR 1-1613, 1624, nos. 39-41. 29 Case C-234/89 Delimitis [1991] ECR 1-935, 991, no. 44. 30 European Commission (1997) Die Anwendung der A r t . 85 u n d 86 des E G Vertrages durch die Gerichte der Mitgliedstaaten, Catalogue no. CV-06-97-812-DE-C, available at http://www.europa.eu.int/comm/competition/publications/#UDIES, p. 133, no. 875.

232

Working Paper IX

Finally, the Commission's unconditional obligation to apply Art. 81 (3) when its four requirements are met is not at variance with the wide discretion the Commission enjoys, and the Court will not interfere with this discretion.31 The Court's jurisprudence shows that the conditions for exemptions under Art. 81 (3) are not justifiable. A Regulation providing for the application of Art. 81 as a whole would deprive Art. 81 (1) of its direct applicability. Such a Regulation would infringe the Treaty.

IV. Effectiveness of decentralised enforcement by national courts and national competition authorities Decentralised enforcement of competition rules in a Community of fifteen or eventually thirty Member States is an important issue by itself. If the Commission's proposals are found to be superior to the present system, or if there are other effective ways to accomplish decentralised enforcement, Member States will always have the option of amending the Treaty. The effectiveness of the new system depends primarily on the workability of decentralisation. The Commission envisages a new role for national courts and national competition authorities. The more difficult issues arise in relation to the application in national courts of Art. 81 as a whole. In this respect, there is a close interdependence between the direct applicability of Art. 81 (3) and its effective enforcement by civil law courts. The major point has been well-made by Wesseling (1999) and deserves to be quoted verbatim: 'It is very difficult to foresee how national courts will apply the administrative discretion implied in the application of Art. 81 (3). Given that the ECJ considers itself incompetent to assess the complex economic facts involved it follows that national courts are even less well-placed to balance the various interests which arise in the context of a request for an exemption from the Community antitrust rules'. This comment highlights the systematic and practical interdependence between the direct applicability of Art. 81 (3) and the effectiveness of its enforcement by national courts.

1. The application of Art. 81 as a whole by national courts The Commission expects Art. 81 to be applied as a whole in three kinds of proceedings: contractual liability proceedings (disputes between parties to an 31

This distinction has been suggested by Schaub (1999), pp. 752-756. Case T-9/93 Schoeller [1995] ECR 11-1611, no. 139/40 quoted for this proposition applies, however, both of these rules simultaneously proving that they are not at variance and not mutually exclusive.

1 - Compatibility, Efficiency, Legal Security

233

agreement), non-contractual liability proceedings (disputes between a third party and one or more parties to the agreement), and applications for injunctions (White Paper no. 99). The expected effects of the new system are summarised as follows: 'In a directly applicable exception system undertakings would be able to invoke the direct applicability of Art. 85 (3) [now Art. 81 (3)] as an argument in their defence before the courts. This new ground of defence would allow them to obtain immediate civil enforcement of those of their restrictive practices that satisfy the conditions of Art. 85 (3) [now Art. 81 (3)]. Their legal certainty would thus be strengthened. Complainants, on the other hand, would be able to obtain damages more quickly where they are victims of illegal agreements'.32

In dealing with private enforcement, the narrow limits of Community law must be observed. The only Community law sanction is Art. 81 (2), providing for the invalidation of restrictive agreements. All other sanctions and remedies are a matter of civil law in the Member States. There are nevertheless important indirect effects of Art. 81 (2) in tort and unfair competition law. For private enforcement to be effective there is obviously a trade-off between the legal certainty of the parties to an agreement and the effective protection of legitimate third parties' interests. The Commission appears to be more concerned with creating legal certainty for parties to an agreement than with the protection of potential victims.33 The implications of the initial tolerance for all restrictive agreements and the absence of a presumption that Art. 81 (3) does not apply are among the most important reasons why private enforcement is bound to be ineffective. The 'globally satisfactory legal certainty'34 applies to the parties to the agreement only, and has important indirect effects for the agreement's stability and its defence against third parties. In all tort or unfair competition cases the validity of the agreement will become relevant. Since there is, according to the Commission, no presumption under Community law that restrictive agreements are void, the general rules of evidence apply. In this respect, the position of German courts appears to be representative. Where the validity of an agreement is challenged, the plaintiff bears the burden of proof for the facts that are required to establish invalidity. The courts refuse to modify these rules, even if a Community Regulation or the jurisprudence of the Court of Justice makes it difficult or even impossible for the plaintiff to obtain and prove the relevant facts.35 The facts that are relevant for Art. 81 (3), and that the plaintiff would have to prove if he wants to challenge a cartel, are, by their very nature, known only by cartel members. The Commission's confidence in the enforcement of competition rules by national courts stands in striking contrast to the neglect of the conditions 32 33 34 35

White Paper no. 100. White Paper no. 78. White Paper no. 78. BGHZ 53, 304 308; OLG Stuttgart WuW/E OLG 5073, 5081.

234

Working Paper IX

determining the effectiveness of this remedy. One explanation is that the Commission does not look for the equivalent of the 'private attorney general' in US antitrust. Private enforcement is said to be strictly private. 'The actions of competition authorities, at both national and Community level, are guided by considerations of public policy in the economic sphere: unlike the national courts, they do not set out to decide disputes between parties, but rather to guarantee the maintenance of a system ensuring that competition is not distorted'.36 I do not agree with this position. It appears at odds with the key role individual rights play in the enforcement of directly applicable Treaty provisions, and it is hardly compatible with the Commission's own position on the direct applicability of Art. 81 as a whole. It is nevertheless important to distinguish the enforcement of competition rules by competition authorities and private plaintiffs. Independent administrative agencies are indispensable for the effective enforcement of rules against restraints of competition because there are insufficient incentives for private enforcement. Undertakings participating in a restraint will not challenge its validity as long as it is profitable to adhere to the agreement. Victims of the restraint are, of course, interested in challenging the agreement but, as a matter of self-interest and financial resources, they are frequently not in a position to do so. On the other side of a cartelised market, customers have to do business with cartel members and they are consequently reluctant to disrupt their relational contracts. Competitors who do not join a cartel will try to live in the shadow of high cartel prices by slightly underselling the cartelised competitors. But they rarely have an incentive to challenge the cartel. If, however, an outsider threatens the cartel's viability, predatory pricing is an effective instrument for bringing the outsider into line. In such a case, where a cartel set out to drive an outsider from the market, the German Reichsgericht in 1931 found a breach of unfair competition law and awarded damages.37 This landmark case, which should convince even Chicago School adherents of the possibility of predatory pricing, started a revision of the traditionally privileged status of cartels under German law. For those undertakings brave enough to go to court, the legal framework determines the odds of success or failure. The plaintiff has to take into account the rules of substance, the rules of procedure, particularly the rules of evidence and of proof, of costs, and the eventual reward, i. e. the compensation for damages and lost profits. There is a close interdependence between all these conditions. Most importantly, and least discussed in the context of Community law, is the key role of substantive rules for the prospects of a damage claim. The stricter and clearer the statutory prohibition, the better the prospects for a potential plaintiff. A comparison with private enforcement in the US legal system offers little reassurance for Community law. Pre-trial discovery and puni36

White Paper no. 91. Reichsgericht, December 12,1931, RGZ 134, 132; for details Mestmacker (1984) at p. 162. 37

7 - Compatibility, Efficiency, Legal Security

235

tive damages are unknown in the law of Member States. In the US, the most important and successful treble damage actions were based upon the per se prohibition of pricefixingor monopolisation under Section 2 of the Sherman Act, or they were related to disputes between manufacturers and dealers.38 In the Community, no case has yet been successful where a claim for damages because of breach of Art. 81 or Art. 82 was reported. This situation will not improve if the courts have to take into account the additional defence of Art. 81 (3). For obvious reasons, it is impossible to deal with the details of Member State tort and unfair competition law as instruments for applying competition rules. Among the more general issues to be taken into account is the kind of injury that is recognised as an antitrust injury. The problems of antitrust standing and antitrust injury in the US system find an equivalent in the tort law and unfair competition law of Member States. Under German tort law the test is similar to the 'direct injury test' of an early American case.39 In the context of tort law, the German Supreme Court considered a damage claim for breach of Art. 81 (1). To make a case, it was not sufficient to show the overall effects of the anticompetitive conduct; what was determinative was 'whether the plaintiff belongs to that group of persons the legislator wanted to protect and whether it was justified to allow the specific remedy as claimed by the plaintiff.'40 In the U.K. the equivalent issue becomes relevant in finding an action for breach of statutory duty. The New Competition Act 1998 appears to be silent on the point. So far there has been no case where a UK court awarded damages for breach of Arts. 81 or 82 (Whish 1994, 60-3). If the Commission wants to promote private enforcement, it would have to strengthen the prohibition of Art. 81 (1). Its policy statements, however, point in another direction. In addition to the direct applicability of Art. 81 (3), the Commission proposes a more liberal 'economic' interpretation of Art. 81 (1). The recent Draft Commission Guidelines on the applicability of Art. 81 to horizontal co-operation agreements41

emphasise the economic analysis of all factors in addition to the kind of agreement, the joint market power of the participants and other structural criteria. It is difficult to see how these criteria, patterned upon the Community merger control regulation, can be applied effectively by national courts. 2. Community law injunctions and claims for damages? So far, it is common ground that Art. 81 (2) has to be applied by national courts whenever the validity of a restrictive agreement becomes relevant. And so far it 38

For details see Jones C.A. (1999) at pp. 79-88. This author is, however, more optimistic about a transfer of U.S. experience to the Community. 39 Loeb v Eastman Kodak 183 F 704 (3rd cir. 1910); referred to by Jones (1999), at p. 160. 40 'Depotkosmetik', WuW DE-R, May 12 1998, 206-7. 41 O J C 118, 27.4.2000.

236

Working Paper IX

has been taken for granted that all other civil law sanctions and remedies are to be provided by the law of Member States. The law of torts and of unfair competition are the most important instruments for relying on Art. 81 and 82 for claims for injunctions and damages. In a recent scholarly monograph, the author maintains that there are autonomous Community law remedies for breach of competition rules. According to this carefully reasoned opinion, Arts. 81 and 82 create directly effective Community rights, the right to damages being a necessary corollary of the principle of direct effect.42 There has not yet been a case decided by the European Court of Justice or a national court recognising those rights. Nor is there an obligation of Member States to provide for sanctions or remedies unknown to their legal systems. On the contrary, the Court of Justice held that the Treaty 'was not intended to create new remedies in national courts to ensure the observance of Community law other than those already laid down by national law'.43 The great exception is, of course, the Court's jurisprudence imposing liability on Member States for loss and damages caused to individuals as a result of breach of Community law for which the state can be held responsible. In Francovich44 the Court distinguished between the right of individuals to obtain compensation, a right founded directly on Community law, and the national law provisions on liability determining that the state must make reparations for the consequences of loss and damage caused (no. 41/42). The principles underlying such rules are 'common to the legal systems of Member States' (Art. 288). Damages suffered by individuals at the hands of the state because of a breach of Community law are to be compensated in accordance with the principles applied to a breach of municipal law. It is argued that Community law remedies for the breach of competition rules are to be recognised on this basis. There are, however, quite a few bridges to be crossed before we arrive at such a conclusion. The extension of the direct applicability of state-related Treaty rules to private law relations would be incompatible with both the Community legal system and the legal orders of Member States. New Community law remedies for breaches of competition rules are neither self-explanatory nor self-executing and do not define themselves. A Community law obligation that Member States provide for such rules as part of their legal systems is unknown to the Treaty. If a uniform regime for the implementation of Art. 81 and 82 by all Member States is considered to be essential, the only way is approximation of national laws according to Arts. 94 and 95 of the EC Treaty. The result of approximation is, of course, not new Community law, but uniform national laws.

42 43 44

Jones C. A. (1999), at p. 78. Case 158/80 REWEIHauptzollamt Kiel, ECR 1981, 1805, 1838. Joint Cases C-6/90 and C-9/90 Francovich, E C R 1991 I, 5357, no. 35.

1 - Compatibility, Efficiency, Legal Security

237

V. The application of Art. 81 by national competition authorities National competition authorities are to apply Art. 81 as a whole. In addition, they are to be empowered to withdraw the benefit of a block exemption on their own territory, if that territory, or part of it, constitutes a separate market.45 This proposal relies upon the power of national competition authorities to apply Arts. 81 and 82. Not all national competition authorities have this power. If they have that competence, the objections against the direct applicability of Art. 81 (3) by courts do not apply to them. In this respect, national competition authorities are, like the Commission, guided by considerations of public policy in the economic sphere (White Paper no. 91). Under this assumption, the solution for the Commission's problems—heavy workload because of centralised notification and authorisation—is the decentralisation of notification and authorisation.46 The consistency of national exemption decisions with coherent Community competition policy can be preserved by appropriate reservations for the Commission's power to take cases out of the jurisdiction of national authorities. It would suffice to provide for an obligation of national competition authorities to keep the Commission informed of all Art. 81 (3) proceedings and to file all exemption decisions with the Commission. These decisions would become final only after a specified, relatively short waiting period. If the Commission did not agree with the national competition authority, it would set aside the decision and decide on the merits itself. Of general importance for any decentralised system of enforcement is the effect of national decisions in Community law. The Commission takes it for granted that administrative decisions by national competition authorities are effective within the territory of the respective Member State only. This territoriality is hardly compatible with the postulates of the single market and the principle of subsidiarity. If the application of Community law is to be delegated to national competition authorities, their decisions must be given the same effect as Commission decisions. Under such a system of decentralised notification and authorisation of Art. 81 (3) decisions, the Commission would not even need all the powers proposed in the White Paper for the implementation of Art. 81 as a whole. The Commission wants to be empowered to: —Take pending cases out of the jurisdiction of national competition authorities and courts (White Paper par. 105);

45

White Paper no. 95. Proposed by the German Federal Government and the German Federal Cartel Authority in Tatigkeitsbericht des Bundeskartellamts 1997/98, p. V and VII; for similar proposals see Moschel (2000) 61, 67. 46

238

Working Paper IX

—Adopt a decision contrary to a decision by a national authority or court as long as that decision is not final; —Prohibit an agreement even after positive decisions by national authorities or courts are adopted (White Paper par. 102b); and finally —Prohibit an agreement after positive action by national authorities or courts has become final with the only proviso being res judicata. These prerogatives are in harmony with the Court's jurisprudence when applied to procedures of national competition authorities. They are, however, of doubtful legitimacy, if applied to national courts. Conflicts with the principle of separation of powers appear to be inevitable both in national systems as well as in Community law.

VI. Primacy of Community law over national competition laws 1. National competition legislation In competition law cases, there are important modifications of the principle that Community law takes precedence over municipal law. In this area, Community law, in the abstract, overlaps with national legislation and administration to an unusual degree. The rule that Community law is applicable only when the restrictive conduct does 'affect trade between Member States' defines the boundary of Community law, but it does not preclude the simultaneous applicability of Community law and national law. In recognising the coexistence of Community law and national competition law, the Court relied on the different objectives of these bodies of law.47 This ruling implies the preservation of national competence for competition law and policy. The implementation of this principle is, of course, predicated upon the pre-eminence of Community law over national law in cases of conflict. The supremacy of Community law is thus not defined by its jurisdictional reach in the abstract, because the Court recognises the important complementary role of national competition law, particularly in cases where national law is stricter than Community law.48 The resultant co-existence and complementarity of Community and national law is an important manifestation of subsidiarity. The European Economic and Social Committee, in its comment on the White Paper, proposes the contrary rule to the effect that national law may not be applied to cases where Community law is applicable.49 This would amount to the factual abolition of national competition laws. 47

Case 14/68 Walt Wilhelm, E C R 1969, 1. Cases 253/78 and 1-3/79 Guerlain, ECR 1980, 2375, no. 80. 49 See Summary of the Observations Submitted, 29 February 2000, available at http://www.europa.eu.int/comm/competition/antitrust/others, no. 8.1. 48

1 - Compatibility, Efficiency, Legal Security

239

2. Primacy of Community law in cases of conflict For the reasons given, Community competition rules take precedence over national competition law in individual cases of conflicting demands on undertakings.50 There is such a conflict, if the national law authorizes an agreement that is contrary to Art. 81. However, there is no conflict if the Commission refrains from applying Art. 81 and issues a comfort letter. In the opinion of the Court, there was nothing to prevent national authorities from applying their own stricter law.51 The Court did not decide whether national competition authorities are bound by Art. 81 (3) exemption decisions. In Walt Wilhelm, the Court referred to certain 'positive though indirect actions' of Community institutions that could be authorized under Art. 85 (now 81) qualifying for precedence over national law. The Commission's traditional view is that all Art. 81 (3) decisions are 'positive', though indirect actions and cannot be disregarded by national competition authorities. The contrary view is that Art. 81 (3) decisions are essentially permissive in character and should not prevent a more stringent regime at national level.52 The White Paper proposals pose a more general and fundamental problem. If Art. 81 (3) is interpreted as a legal exemption, there are no 'positive decisions' as defined in Walt Wilhelm. For a finding of 'no infringement' of Art. 81, it is irrelevant whether there is no restriction or whether the restriction is justified because of Art. 81 (3). It follows that these decisions, as recognised by the Commission, are negative clearances within the meaning of Art. 2 Regulation 17/62. Such decisions certify that, on the basis of the facts known to the Commission, there are no grounds for action in respect of the agreement. Under present Community law, a negative clearance does not bind national competition authorities or courts in the application of Community law, nor does it prevent national competition authorities or courts from applying stricter national law (Rittner 1997, 25). These limitations apply, under the new system, to decisions of national competition authorities and courts. With respect to the Commission, the White Paper proposes another kind of individual decision that is not a prohibition decision (no. 88). Positive decisions of this kind, according to the White Paper, would be taken in exceptional cases on the grounds of general interest. But they would be of a declaratory nature and would have the same legal effect that negative clearance decisions have at present.53 By the same token, group exemptions would be of a declaratory nature and become group negative clearances (Deringer 2000,5-7; also Bartosch 2000, 462-6). 50 51 52

Case 14/68 Walt Wilhelm, ECR 1969, 1, no. 1-8. Cases 253/78 and 1-3/79 Guerlain, ECR 1980, 2375, no. 18. For details see Bellamy and Child (1993) at pp. 10-60, with references to, and preference for, the view that 81 (3) decisions are not entitled to primacy over national law. 53 White Paper no. 89.

240

Working Paper IX

If present primacy rules are applied, the new system opens a wide field to the application of stricter national competition laws. If the new system is adopted, the threshold question is whether Commission decisions and group exemptions, even though they are declaratory, are to take precedence over national laws. If new Community rules of competence are to be adopted, another key issue is whether national competition authorities should be under an obligation to apply Art. 81 as a whole, even in cases where national laws provide for different substantive or procedural rules. Such a Regulation would probably mean the eventual abolition of the independent and complementary enforcement of national competition laws. That issue is beyond the scope of this paper.

X Alexander Schaub The Reform of Regulation 17/62: The Issues of Compatibility, Effective Enforcement and Legal Certainty

I. Introduction The Commission's proposal for reform of Regulation 17/'62, presented in the White Paper on modernisation of the rules implementing Articles 81 and 82 of the EC Treaty, forms part of a comprehensive reform of EC competition law and policy, reaching from the substantive to the procedural rules applicable in the fields of antitrust, mergers and state aids. These reforms pursue the same basic objectives, namely: (a) more efficient enforcement of the EC competition rules; (b) less bureaucracy for companies; and (c) a more level playingfieldfor companies through the development of a common competition culture and an extended application of Community law. The principal task of a competition authority is to effectively protect competition. As resources are necessarily limited, efficient enforcement requires that resources are concentrated on the most important cases; in the case of the Commission that means cases that have a Community interest and that represent a real threat to competition. In the substantivefieldthe principal instrument employed is the adoption of a more economic approach. A clear example thereof is the new block exemption regulation for vertical restraints1 that focuses on the economic effect of agreements rather than their legal form. This allows the Commission and the national authorities to concentrate on the important cases, where the parties have market power, while creating legal certainty for the vast majority of companies. However, to be effective this substantive reform must be accompanied by procedural reform. The current notification and authorisation system established by Regulation 17/62 does not allow the Commission to employ its enforcement resources efficiently. Furthermore, the Commission's monopoly over Article 81 (3) prevents effective application of Article 81 at national level. The White Paper proposes to enhance enforcement by abolishing the Commission's monopoly over the application of the exemption rule of Article 81 (3) as well as the notification and authorisation system for the application of that provision. This would be accomplished by rendering Article 81 (3) directly 1 See Commission Regulation (EC) No 2790/1999 of 22.12.1999 on the application of Article 81 (3) of the Treaty to categories of vertical agreements and concerted practices, OJ 1999 L 226, page 21.

242

Working Paper X

applicable. The legal basis for the proposed reform is found in Article 83 according to which the Council is empowered, inter alia, to lay down detailed rules for the application of Article 81 (3). In the new system both national courts and national competition authorities would be empowered to apply Article 81 as a whole, article 81 (3) would constitute a legal exemption. The prohibition rule of Article 81(1) would not apply when the four conditions of Article 81 (3) are met. In such a system, agreements that are compatible with Article 81 as a whole are valid and enforceable. Conversely, agreements that are contrary to Article 81 as a whole are invalid ab initio. The prohibition rule on which Article 81 is based remains fully in place. The Commission's proposal would eliminate the current rather artificial division of power under Article 81 that has been established by Regulation 17/62. At present national courts are empowered to apply Article 81 (1), which has direct effect; another body—namely the Commission—must assess the positive effects of the very same agreement. In practice, these issues are closely related and one is not more complex than the other.

II. Compatibility with the Treaty 1. The compatibility issue Following the publication of the White Paper some commentators have argued that the proposed legal exception system is contrary to Article 81 (3), implying that the implementation of the Commission's proposal would require amendment of the Treaty.2 According to this argument the Treaty presupposes the application of Article 81 (3) by a prior administrative decision. In essence, the compatibility issue relates to the proposal to empower national courts to apply Article 81 (3). The fundamental question is therefore: is the wording and objective of Articles 81 (3) and 83 such that they preclude the Council from giving to national courts the power to apply Article 81 (3). The Commission, on its part, considers that Article 83 empowers the Council to make a choice between an authorisation system and a legal exception system when laying down the detailed rules for the application of Article 81 (3). In doing so the Council must, inter alia, take account of the need to ensure effective supervision on the one hand, and to simplify administration to the greatest possible extent on the other (cf. Article 83 (2) (b) EC Treaty).

2 This paper does not purport to deal with all the details of this issue, which are also addressed by Mr Giuliano Marenco of the Commission's Legal Service.

1 - Compatibility, Efficiency, Legal Security

243

2. The function of national courts in the Community legal order Under Community law it is the rule rather than the exemption that the law can be applied directly by national courts. The Court of Justice has considered this essential in order to ensure the effet utile of Community law.3 National courts have played an essential role in the enforcement of Community law in general. It is a unique feature of the Community legal order that it grants rights to citizens that can be invoked and enforced before national courts. The national courts' protection of individual rights has made a very significant contribution to the development of Community law. For example, in the field of the free movement of goods, persons, services and capital, private enforcement before national courts has contributed as much as enforcement by the Commission (if not more) to the development of the law. However, in the antitrust field we have—due to the Commission's monopoly on Article 81 (3) created by Regulation 17/62—become accustomed to the courts playing a rather limited role.

3. The justiciability of Article 81 (3) One of the main arguments against giving national courts the power to apply Article 81 (3) is based on a perception that the provision is not sufficiently clear and precise. Article 81 (3) contains four conditions for the application of the exemption rule: (a) the agreement must improve the production or distribution of goods or promote technical or economic progress, (b) consumers must receive a fair share of the benefits, (c) the restrictions must be indispensable for achieving the benefits, and finally (d) effective competition must not be eliminated. When examining each of these four conditions, it becomes clear that there is nothing therein to make Article 81 (3) unsuitable for application by national courts. The absence of a direct effect of Article 81 (3) does not follow from the content of that Article. It follows from the present Regulation 17/62, which grants exclusive power to the Commission. The first condition of Article 81 (3) relates to the efficiency effects of the agreement. Both horizontal and vertical agreements can create efficiencies such as economies of scale, the provision of valuable services and the supply of new or improved products. Article 81 (3) requires that any such benefit must be balanced against the negative effects on competition. The Court of Justice has expressly held that all positive effects must show appreciable objective advantages that would compensate for the disadvantages caused by restrictions of competition.4 This kind of balancing also exists in other fields of Community law. Under the free movement rules, courts are required to balance the 3 4

Joined Cases 98/83 and 230/83 Van Gend & Loos v Commission [1984] ECR-3763. Joined Cases 56/64 and 58/64 Consten and Grundig [1966] ECR-429.

244

Working Paper X

Community interest in free movement against a host of other general interests such as health protection, consumer protection and protection of the environment. Conceptually thefirstcondition is therefore by no means a novelty. The principal aim of the Community competition rules is to protect effective competition as a means of promoting consumer welfare.5 This aim is reflected in the second condition of Article 81 (3), according to which consumers must receive a fair share of the benefits. According to the Commission's existing administrative practice, the higher the degree of competition remaining in the market, the more it can be assumed that competitive pressures will ensure that a fair share of the benefits are passed on to consumers. Courts are already required to assess the impact of a practice on competition under Article 81 (1). The concept of 'fair share' implies that the benefits must clearly compensate for any likely negative impact on consumers caused by the restriction of competition found under Article 81 (1). The third condition on indispensability implies that there must not be any less restrictive means to achieve the benefits flowing from the agreement. This proportionality test is similar to that which national courts must apply in many other areas of Community law. The third condition, in most cases, has been the decisive factor in rejecting the applicability of Article 81 (3). Finally, the fourth condition concerning the elimination of competition in respect of a substantial proportion of the products in question places an upper limit on the scope of application of Article 81 (3). This condition must be read in a coherent way with the concept of dominance in Article 82. According to the existing case law, dominance is a position of economic strength that enables the firm to prevent effective competition from being maintained on the relevant market.6 Article 82 contains no exemption rule.7 Appreciable restrictions of competition implemented by a dominant firm cannot be declared compatible with Article 81 (3). The fourth condition of Article 81 (3) is therefore interpreted as referring to the concept of effective competition, which ensures consistency between the scope of the application of Articles 81 and 82. National courts must already apply Article 82 that, as mentioned, involves the same type of assessment as the last condition of Article 81 (3). In the present context, it is also useful to compare Article 81 (3) with other Treaty provisions that have been held by the Court of Justice to have direct effect. Article 86 is particularly illustrative in this respect. Article 86 (1) provides that in the case of public undertakings and undertakings holding special or exclusive rights, Member States shall neither enact nor maintain in force any measure contrary to the rules contained in the Treaty, and particularly the competition rules. Article 86 (2) contains an exemption in favour of undertakings 5

Paragraph 54 of the Green Paper on vertical restraints. See for example Case 107/76 Hoffmann La Roche v Centrafarm [1977] ECR-957, paragraph 38. 7 This statement is subject to the proviso that certain conduct may be considered objectively justified and thus not abusive. 6

1 - Compatibility, Efficiency, Legal Security

245

entrusted with the operation of services of general economic interest. Such undertakings are subject to the rules contained in the Treaty, in particular the rules on competition, in so far as the application of such rules 'does not obstruct the performance, in law and in fact, of the particular task assigned to them. The development of trade must not be affected to such an extent as would be contrary to the interests of the Community'. Finally, according to Article 86 (3), the Commission shall ensure the application of the provisions of Article 86 and shall—where necessary—address appropriate directives or decisions to Member States. Article 86—like Article 81—contains a prohibition, an exemption and a provision, providing for the adoption of decisions and instruments of a general nature. An important difference, however, is the absence of an equivalent to Article 83 (2) (b). Early on, the Court of Justice held that Article 86 (1) has' direct effect.8 More importantly, and more recently,9 the Court has implicitly held that Article 86 (2) has direct effect and can be applied by national courts.10 The assessment required under Article 86 (2) can be no less complex than that which must be carried out under article 81 (3). In Corbeau11 the national court was, inter alia, required to assess whether cross-subsidisation between different sectors of activity was necessary in order to preserve the economic equilibrium of the Belgian postal service. Article 86 (2) thus requires that the national court perform a proportionality test.12 It also requires that the national court balance the public service interest against the Community interest.13

4. Do Articles 81 and 83 require an authorisation system? When assessing the fundamental question of compatibility it is important not to lose sight of the fact that the question is not whether one system or another is reflected more in the wording of Article 81 (3), but whether the introduction of a legal exception system would be contrary to the Treaty. Such a conclusion would require that the Treaty itself clearly opted for an authorisation system.14 This is not the case. Article 81 (3) prescribes neither an authorisation system enforced by an administrative authority nor exclusive competence for the Commission. Both these essential features of the current system are set out in 8

Case 155/73 Sacchi [1974] ECR^09, paragraph 18. By doing so the Court has departed from the previous jurisprudence set in Case 172/83 Syndicat national des fabricants raffineurs d'huile de graissage [1984] ECR-555, paragraph 15, as the most recent example. 10 Case C-320/91 Corbeau [1993] ECR 1-2533. 11 See note 10. 12 Case C-209/98 Sydhavnens sten og grus at paragraphs 79 and 80 judgment of 23.5.2000, not yet reported. 13 See e.g. paragraph 74 of the judgment cited in note 12. 14 See in this respect the opinion of Advocate General Lagrange in Case 13/61 Bosch [1962] ECR-309. 9

246

Working Paper X

regulation 17/62. Article 81 (3) is also silent as to when the declaration of inapplicability may take effect (ab initio or for the future). Some commentators argue that the word 'may' in Article 81 (3) implies a margin of discretion that, by nature, must be reserved to a public authority. However, as it clear from the case law of the Community Courts15 and the administrative practice of the Commission, undertakings have a right to the benefit of Article 81 (3) when the four conditions of that article are met. Furthermore, the application of Article 81 (3) has never been refused (nor can it be) for any other reason than the non-fulfilment of one or more of the four conditions of Article 81 (3). It is not possible to pursue objectives under Article 81 (3) that cannot be subsumed under the four conditions.16 The Community Courts have recognised that there is a certain margin of appreciation in the application of Article 81 (3). Indeed, the Court of Justice has held that 'the exercise of the Commission's powers necessarily implies complex evaluations of economic matters', and that 'judicial review of these evaluations must take account of their nature.'17 The fact, however, that the application of Article 81 (3) may imply complex economic assessments does not mean that it confers discretionary powers. Nor does it constitute an obstacle to the application of Article 81 (3) by national courts. Both Articles 81 (1) and Article 82 are already applied by national courts, by virtue of their direct effect,18 notwithstanding the fact that they may equally imply complex economic assessments. In Bagnasco,19 responding to a request for a preliminary ruling from a national court, the Court of Justice expressly held that 'the answer to the question whether or not the conditions for the application of Article 81 EC are fulfilled depends on complex economic assessments which it is for the national court to undertake'. The same is true of Article 82, which involves the assessment of complex issues such as the relevant antitrust market and dominance on that market. Some commentators also argue that the words 'may be declared inapplicable' suggest positive action by an administrative authority. First of all, these words should be read in the context of the sentences that follows, namely '. . . in the case of any agreement or category of agreements between undertakings . . .' Article 81 (3) allows a choice between individual application and application to categories of agreements. Article 81 (3) together with Article 83, attributes the competence to the Council to declare itself or to empower the Commission to declare Article 81 (3) inapplicable to categories of agreements. As long as this power has not been exercised, agreements are subject to individual assessment and individual application of Article 81 (3). 15

See in particular Case T-17/93 Matra Hachette [1994] ECR11-595, paragraphs 85 and 139. 16 See to that effect the Matra judgment cited above. 17 Joint Cases 56/64 and 58/64 Consten and Grundig v Commission [1966] ECR-299. 18 Case 127/73 BRTw Sabam [1974] ECR-51, paragraph 16. 19 Joined Cases C-215/96 and 216/96 Bagnasco [1999] ECR 1-135, paragraph 50.

1 - Compatibility, Efficiency, Legal Security

247

Second, the word 'declare' says nothing about the nature of the declaration. A declaration pursuant to article 81 (3) can be either a decision that declares Article 81 (1) inapplicable to a restrictive agreement with effect ab initio, or a decision that declares Article 81 (1) only inapplicable for the future or from a given point in time fixed in the decision. In the former case it is a declaration of a declaratory nature, in the latter case it is of a constitutive nature. Unlike the word 'authorise' in Article 65 ECSC Treaty, the word 'declare' is compatible with both types of decisions (declaratory and constitutive). In a legal exception system any finding of non-infringement is of a declaratory nature; a legal exception system is therefore perfectly compatible with the term 'declare' in Article 81 (3). In fact, in the present system, national courts have the power to declare an agreement incompatible with Article 81 (1). In doing so, they must already—when an agreement has been notified—make a negative application of Article 81 (3). A national court may thus declare an agreement that falls within Article 81 (1) null and void if, having regard to existing block exemption regulations and the Commission's previous decisions, the agreement may on no account be the subject to an exemption decision under Article 81 (3).20 It should also be mentioned that the inclusion in Article 81 (3) of a power to adopt block exemption regulations does not, as some argue, exclude the introduction of a directly applicable system. Article 81 (3) contains two techniques for applying the exception rule to individual agreements and to categories of agreements. Article 83 empowers the Council to device an enforcement system that is different for each technique. There is nothing incompatible with the Treaty to give a declaratory nature to individual applications of Article 81 (3) while maintaining a constitutive nature to block exemption regulations. National courts have an obligation to apply Community Regulations unless they have been declared invalid by the Court of Justice.21 According to Article 249 EC regulations are binding and directly applicable Community acts. Block exemption regulations remain relevant in the proposed new system as a tool for ensuring the coherent application of Community law. Block exemption regulations also remain relevant for the purposes of providing legal certainty to companies.

III. Efficiency of enforcement 1. Efficient enforcement in the context of the White Paper It is clear from the very content of the Treaty that effective competition constitutes a public good that must be protected. The very foundation of the 20 21

Case C-234/89 Delimitis [1991] E C R 1-935, paragraph 50. Case 314/85 Fotofrost [1987] ECR-4199.

248

Working Paper X

European Union is an open market economy. As is now generally accepted, a competitive economy ensures the optimal functioning of the Internal Market and the competitiveness of European industry on an increasingly world wide arena, and it creates benefits for consumers and our societies as a whole. Therefore the primary reform objective is, and must be, effective protection of competition. The objective of the reform is not to create less work for the Commission but to strengthen the enforcement of EC competition law throughout the European Union.

2. The present system and its inefficiencies In the present enforcement system the prohibition rules of Article 81 (1) and Article 82 can be invoked before national courts. Similarly, national competition authorities have the power to apply these provisions. On the other hand, the Commission has the exclusive power to apply Article 81 (3). An agreement can only benefit from the exemption rule if the Commission has been notified of it. There is no obligation to notify. In this way the notification system under Regulation 17/62 differs fundamentally from the system established by the Merger Regulation, that is not affected by the White Paper.22 Agreements are only notified when the parties themselves expect that the agreement merits negative clearance or exemption. Since the adoption of Regulation 17/62 in 1962, the substantive content of Article 81 has been clarified by legislative instruments like block exemption regulations, Commission notices and guidelines, as well as a growing stock of case law and practice. Thus companies and their lawyers generally know what is prohibited and what is likely to be exempted. Only the latter agreements are 'notified'. However, we have good reason to believe that some companies also engage in practices that are prohibited by the competition rules. For obvious reasons these practices are never notified. The notification system generates a largely unproductive workload, which prevents the Commission from being a proactive and focused enforcement authority. It also causes unnecessary delays in dealing with ex officio procedures and complaints. This is most unfortunate since complaints provide a much more useful source of information about restrictive practices than do notifications. Notifications that do reveal real competition issues are virtually always the subject of a complaint. Furthermore, the Commission receives large 22

With regard to mergers, ex ante enforcement is the most effective. Once implemented, a merger is difficult and costly to unravel. Ex post control would therefore either not take place or would waste society's resources. Such considerations do not apply to Articles 81 and 82 since there is generally not the same degree of integration of activities. The situation may, however, be different with regard to production joint ventures, which is the reason why the White Paper suggests that it might be appropriate to subject production joint ventures to the Merger Regulation procedure.

1 - Compatibility, Efficiency, Legal Security

249

quantities of general information concerning market conditions and structural developments via the merger control system. Abolishment of the notification system will therefore not significantly reduce the level of market information. The Commission's monopoly under Article 81 (3) effectively prevents national competition authorities and courts from enforcing Community competition law. This blocking effect follows from the fact that companies alleged to be in breach of Article 81 (1) quite naturally claim that their agreements meet the conditions of Article 81 (3) and are therefore legal. As the national bodies have no power to apply Article 81 (3), the continuation of proceedings is made difficult. In most cases, the national action must be suspended pending the Commission's decision.23 This leads to unnecessary delays and creates a clear disincentive to invoke and apply EC competition law. Many national competition authorities have not even been empowered to apply Articles 81(1) and 82. This is still the situation in seven Member States. The result is that complainants have turned to the Commission rather than to national courts for the application of EC competition law. In the past the Commission has tried to promote the application of EC competition law by national courts and competition authorities as a means of enhancing enforcement. However, these attempts, which are spelled out in the co-operation notices from 1993 and I99724 respectively, have been largely unsuccessful.

3. The advantages of the legal exception system The legal exception system proposed in the White Paper would entail a number of enforcement gains. First, it enables the Commission to concentrate on the cases that present a real threat to competition to the detriment of consumers, which in turn strengthens the deterrent effect of the rules. The Commission would also be in a position to allocate more resources to the important task of clarifying the rules. In fact, in the proposed new system one of the main tasks of the Commission would be to develop competition policy and to issue measures that clarify the law and policy. These measures include block exemption regulations, guidelines, and opinions and decisions in individual cases including non-infringement decisions in the Community public interest. Clarification of the rules not only increases predictability for companies, it also increases the deterrent effect of the prohibition rule. Second, the proposed system would enable the national competition authorities to effectively apply EC competition law. This makes it possible to create a 23

See in this respect t h e Delimitis j u d g m e n t cited in note 20. See Commission Notice on co-operation between national courts and the Commission in applying Articles 81 and 82 of the Treaty, O J C 39 of 1993, p. 5, a n d t h e Commission Notice on co-operation between national competition authorities and the Commission in handling cases falling within the scope of Articles 81 and 82 of the Treaty, O J C 313 of 1997, p. 3. 24

250

Working Paper X

network of competition authorities within the European Union that all enforce the same body of law concentrating their enforcement resources on the most important cases. There will therefore be more enforcers and more application of Community law, increasing its deterrent effect. Third, the elimination of the current division of competence under Article 81, where one body can assess the negative effects of the agreement whereas another must assess the positive effects, would allow the national courts to play the same important role in the application of the antitrust rules as they play in all other areas of Community law. National courts are ideally suited to deal with certain types of disputes, for example, about contractual terms. Moreover, national courts possess powers beyond those of either the Commission or national competition authorities. Only national courts can make orders imputing civil law implications of compliance or non-compliance with Articles 81 and 82. This means that they alone can order the civil enforceability of agreements or alternatively find an agreement to be null and void. Again, only national courts can award damages to victims of anti-competitive practices. Competition authorities' powers are limited to ordering the termination of an infringement and to imposing fines on companies that have breached the competition rules. Finally, when it is notified of an agreement, the Commission is required to carry out a prospective analysis of, not only structural factors, but also of how the various elements of the agreement will affect competition in the future. In practice, this has led the Commission to adopt a cautious, formbased approach, insisting in many cases on the elimination of clauses in the agreements that restrict the commercial freedom of the parties. The White Paper proposal will promote a more economic approach concentrating on the current and past effects25 of the agreements on competition in the relevant market.

IV. Non-infringement decisions 1. The nature of the instrument In the White Paper the Commission proposes that it should have the power to adopt non-infringement decisions in the Community's public interest. They would only be adopted in exceptional circumstances for policy reasons. An example would be an agreement that raised new issues in respect of which it 25

Agreements that are restrictive by object, such as hard core cartels, are prohibited without it being necessary to take account of their concrete effects, see e.g. judgment of 8.7.1999, Case C-49/92 [1999] ECR \-A\25 Anic Partecipazioni, paragraph 99. As a general rule such agreements cannot benefit from the exemption of Article 81 (3).

I — Compatibility, Efficiency, Legal Security

251

would be in the Community interest to provide guidance on the Commission's approach to certain restrictions in it.26 The Commission would adopt such decisions on its own initiative and not at the request of the parties to the agreement. It is a fundamental premise of the White Paper that all forms of voluntary notification be abandoned. According to the White Paper, a non-infringement decision would find that the agreement in question is compatible with Article 81 and would be of a declaratory nature. Such decisions would therefore differ from the current exemption decisions, which are of a constitutive nature. Once final, exemption decisions have binding effect erga omnes for a determined period and can only be withdrawn by the Commission. Even if the material facts change, Article 81 (1) cannot be applied by national courts or national authorities to an agreement that benefits from an exemption decision. In contrast, non-infringement decisions would set out the Commission's assessment of the agreement based on the facts in its possession and would, as mentioned, be of a declaratory nature. National courts would not be bound to substitute such decisions for their own. However, when applying Community law, national courts are considered Community courts of general jurisdiction and are bound to co-operate with the Court of Justice and the Commission with a view to maintaining the coherence of the Community legal system. Conflicts with prior Commission decisions must therefore be avoided.27

2. Positive decisions and the position of national competition authorities The establishment of a network of competition authorities is one of the main pillars of the proposed new system. Within this network one of the main tasks of the Commission will be to define—in full co-operation with the network— Community competition policy. The main tools in this respect will be block exemption regulations, guidelines and decisions in individual cases. All members of the network will then apply this common law and policy. In this system the function of positive decisions would be to define and defend the common policy. As to the application of the common law and policy, the main instruments would be prohibition decisions, commitment decisions, and rejections of complaints. The White Paper does not envisage that the Commission would adopt non-action decisions with a view to bringing an end to an ex officio procedure in cases where there was no ground for action. Such cases would normally be closed informally, subject to the principle of good administration, requiring that the companies concerned be informed. However, in some national legal 26 27

See paragraph 88 of the White Paper. See to that effect paragraph 47 of the Delimitis judgment cited in note 20.

252

Working Paper X

systems the competition authorities are obliged to adopt a decision at the end of every procedure. National competition authorities would not be precluded from adopting such non-action decisions. The function of these decisions would be to bring an end to the procedure rather than to define Community competition policy. In addition, they would only be adopted to bring an end to ex officio procedures and not at the request of undertakings. This limitation is necessary in order to avoid the reintroduction of ex ante control. Furthermore, such non-action decisions should bind only the adopting authority and should preclude neither national courts nor another competition authority from acting in respect of the same agreement.

V. Legal certainty 1. The importance and requirements of legal certainty In an open economy, companies should be encouraged to enter into procompetitive transactions. Consequently, competition law should not only discourage anti-competitive practices, but it should also encourage those that promote competition and consumer welfare. An adequate level of legal certainty is therefore a legitimate and necessary objective. The Commission's White Paper proposal, while aimed at restoring the primary task of effective enforcement, fully recognises the need to provide companies with an appropriate and adequate degree of legal certainty. Under the proposed new system, however, legal certainty must be ensured by means other than a notification and authorisation mechanism. A system of ex ante control of individual agreements at the demand of undertakings must be excluded; otherwise the reform would not attain its main objective, namely that of effective protection of competition. In fact, a voluntary notification system, as called for by some commentators, would run counter to the very raison d'etre of the proposed reform. It would prevent the Commission from focusing on the cases that present a real threat to competition and would also perpetuate the blocking effect of the current system on the application of EC law by national courts. As under the current system, a party to litigation would have an incentive to notify the Commission of an agreement; that in turn would impede effective application of Article 81 by the courts. Companies must accept responsibility for assessing their own agreements. What companies can reasonably expect is an adequate level of predictability that allows them to properly assess how the rules will be applied. The level of predictability depends on the clarity of the rules and on the way in which they are applied. In a system of parallel competencies the issues of legal certainty and coherent application are therefore closely related. To ensure coherence within the system it is necessary that the rules are sufficiently clear and that the

1 - Compatibility, Efficiency, Legal Security

253

approaches followed in their application are similar. It must be ensured that cases are assessed on the basis of the same rules and in the same way by the bodies involved in their application. The outcome in an individual case, on the other hand, depends on the facts, and particularly, the conditions of competition prevailing within the relevant market. The fact that the outcome might differ according to particulars of the relevant market, therefore, does not itself give rise to consistency and predictability concerns. /. 1. Predictability as to fines and hard-core restrictions Some commentators on the White Paper argue that the Commission's reform proposal increases the risk of unforeseeable fines due to a lack of predictability in the new system. It should first be noted that it is a long standing practice of the Commission to impose fines only in cases where it has been clearly established, either in horizontal instruments or in prior case law or administrative practice, that certain behaviour constitutes an infringement of the antitrust rules. To give an example: in the recent decision in the 1998 World Cup case, the Commission referred expressly to the lack of clarity of the law and imposed only a symbolic fine of 1,000 Euro on the Comite frangais d'organisation de la coupe du monde de football.26 This policy will not change. Furthermore, all prohibition decisions adopted by national competition authorities will be subject to prior consultation within the network (see point 5.3.2 below). Fining policy within the network will therefore also be subject to co-ordination. As far as hardcore restrictions are concerned, the state of the law is clear and predictable. In many cases the restrictions falling into this category are expressly blacklisted in block exemption regulations. Furthermore, there is a considerable body of case law and administrative practice on the interpretation of the concept of restriction by object in Article 81 (1). After more than forty years of case law no one can reasonably doubt that horizontal pricefixingand market sharing or resale price maintenance and export bans constitute restrictions by object. In the new system, all bodies involved in the application of Community competition law must follow this settled case law when applying EC law. The real predictability issue does not concern agreements that are restrictive by object but those that might have restrictive effects. However, with regard to this category of agreements the proposed new system—when considered in its entirety—presents a number of improvements compared with the existing system. 2. The level of legal certainty provided under the present system Before dealing with the level of legal certainty in the proposed new system, it is useful to briefly examine the level of legal certainty provided under the existing 28

See OJ L 5 of 2000, p. 55.

254

Working Paper X

enforcement system. Formal exemption decisions provide a high degree of legal certainty as they have effect erga omnes for the duration of the decisions. However—as is common knowledge—very few such decisions are adopted (on average only five a year). The remaining cases are dealt with by means of nonbinding comfort letters.29 Furthermore, comfort letters are neither reasoned nor published, and they consequently do little to clarify the overall body of rules. This is not to imply that comfort letters are devoid of any significance in terms of legal certainty. Comfort letters based on Article 81 (3) indicate to national courts and national authorities that the Commission considers that the agreement merits exemption, and thus that it is covered by the Commission's monopoly. However, the new and more economic approach under Article 81 (1) significantly reduces this effect since the vast majority of comfort letters will be of the negative clearance type. Such non-action comfort letters merely indicate that—on the basis of the facts in its possession—the Commission does not intend to intervene. Account must also be taken of the fact that notification is not costless. There is a clear cost both in terms of preparing the notification and of subsequent bureaucracy. 3. Legal certainty in the proposed new system 3.1. Direct applicability of Article 81 (3) One of the immediate benefits of the Commission's proposal in the White Paper is the direct applicability of Article 81 (3). As a directly applicable norm, article 81 (3) can be invoked as a defence before national courts and national competition authorities. Agreements that are compatible with Article 81 as a whole will be valid and enforceable without any Commission intervention being required. Agreements will no longer be invalid and unenforceable merely for lack of notification. Furthermore, the direct applicability of Article 81 (3) will strengthen the primacy of Community law over national law.30 In the same way that a block exemption regulation precludes the application of stricter national laws, a company would be able to invoke the exception rule in an individual case. This enhanced primacy will contribute to the creation of a level playing field for companies that operate in the Internal Market. 3.2. The network of competition authorities One of the main pillars of the reform is, as already mentioned, the creation of a network of competition authorities that co-operate closely in the application of the common rules and policy. This network—with the supporting rules and 29 30

See most recently Case T-24/97 Stork Amsterdam [2000] ECR 11-309, paragraph 84. This issue is dealt with in m o r e detail by M r Emil Paulis in his contribution.

1 - Compatibility, Efficiency, Legal Security

255

instruments for co-operation—will in itself greatly enhance predictability compared to the present situation where 16 authorities apply 16 sometimes different sets of rules and procedures, each subject to its own autonomous system of judicial control. Taken as a whole, the present system that incorporates both Community and national elements, leaves room for significant inconsistencies and reduces the level of predictability within the Internal Market.31 The Commission will remain at the centre of the network and will play a leading role in defining policy and ensuring consistent application. The Commission will be kept informed of on-going cases via the network. In the White Paper the Commission proposes to maintain the power under the current Article 9 (3) of Regulation 17/62 to withdraw a case from a national authority. This instrument can be useful for the purposes.of resolving disputes within the network and ensuring consistency. It is assumed, however, that disputes will be satisfactorily solved by intensive consultation and discussion and that Article 9 (3) will only be applied in exceptional cases. The White Paper proposal is based on the principle of parallel competencies. Each member of the network will have the power to act on the basis of Community competition rules in respect of agreements or practices affecting trade between Member States. However, there will be guidelines on the allocation of cases within the network. The principle must be that cases are dealt with by the authority best placed to act. The aim is to avoid multiple control, duplication of work or inefficient actions. In the Community context multiple control means parallel procedures with each authority acting within its own jurisdiction. It is-not possible for several national authorities to cumulate parallel claims before the same court. Community courts are not competent to hear direct claims against private parties. Such cases are heard by national courts and reach the Court of Justice only via the reference procedure set out in Article 234 EC.32 Multiple parallel control within the same relevant market is a burden on companies and increases the risk of incoherent application; it should therefore, in general, be avoided. This is not to say that parallel action within the same relevant market should be excluded. There might be cases in which it is appropriate and effective; however, this would be the exemption rather than the rule. 31

In Case C-7/97 Oscar Bronner [1999] ECR 1-7791, the Court of Justice held that a national court was entitled to refer a question concerning the interpretation of national competition law when it considered that a conflict between Community law, in casu Article 82, and national law was capable of arising. 32 The purpose of the reference procedure is to prevent the occurrence within the Community of divergences in judicial decisions on the questions of Community law, see judgment of 6.10.1982, Case 238/81 CILFIT v Ministry of Health [1982] E C R - 3 4 1 5 , paragraph 7. The absence of appeal from the national courts to the Community Courts means that national judges have a special responsibility t o ensure that the exercise of their power t o apply Community law does n o t prejudice the uniform application of Community law, see in this respect the Fotofrost judgment at paragraphs 15 and 16, and the Delimitis judgment cited in note 20.

256

.

Working Paper X

Certain decisions to be adopted by the national competition authorities33 will also be subject to prior consultation and co-ordination. Prohibition decisions, commitment decisions and decisions withdrawing the benefit of a block exemption regulation merit special attention because, once adopted and enforced, they can only be reversed with difficulty. This is even more important when they have served as a basis for claims for damages. Consultation concerning prohibition decisions and decisions accepting commitments is also important in order to maintain a coherent competition policy since, in the new system, policy will be developed to a significant extent by means of prohibition decisions and commitment decisions. 3.3. Application by national'courts The application of Article 81 (3) by national courts will be subject to the preliminary reference procedure of Article 234. In other areas of Community law this instrument has contributed not only to the development of Community law but also to uniformity and predictability in its application. In the proposed system, more cases will become subject to Community law and the reference procedure. Furthermore, the Commission will strengthen the existing co-operation instrument with the national courts contained in the 1993 notice. In particular, it is the intention to establish a deadline for reply. It can be expected that the incentive for national courts to obtain input and opinions from the Commission will grow in the future system as they will effectively obtain the power to apply Article 81. The application of EC competition law is frequently a matter of law and fact. In such cases, the Commission might be able to provide valuable assistance to national courts, drawing on its own administrative practice. The Commission is also envisaging a mechanism whereby it could intervene more actively in court proceedings, allowing it to defend the common competition policy before national courts. One such instrument, which is mentioned in the White Paper, could be the right to intervene as amicus curiae. The Commission would then, on its own initiative, submit written observations— and possibly oral observations—to the court hearing the case. An alternative could be to grant the Commission the right to appeal in the Community interest against national court decisions. Such instruments would promote coherence and predictability. 3.4. Clarification of the rules One of the principal tasks of the Commission will be to clarify the rules. Block exemption regulations will remain important in this respect. In this field the Commission is, as indicated above, adopting a new approach based on the 33 Commission decisions are already subject to the Advisory Committee procedure and will remain so.

1 - Compatibility, Efficiency, Legal Security

257

effect of the agreement rather than its legal form. The vertical restraints block exemption regulation creates a broad safe harbour below a market share threshold of 30%, subject to a limited list of hardcore restrictions. In the area not covered by block exemptions, the Commission strives to lay down clear guidelines in order to provide predictability for a maximum number of cases. In the field of verticals the Commission has very recently adopted a comprehensive set of guidelines, and a similar set of guidelines on horizontal co-operation agreements have been published for consultation. The Commission has also been encouraged to issue guidelines in the application of Article 81 (3). Although there is a large number of Commission decisions on the application of Article 81 (3), it might be very useful to issue a set of guidelines concerning the methodology of analysis to be applied under Article 81 (3). This would no doubt assist national courts in carrying out their new responsibilities as well as assisting companies in the assessment of the compatibility of their agreements with Article 81 as a whole. Such guidelines would also allow the Commission to consolidate its more recent practice in thefield,which—like Article 81 (1)—has evolved in light of developments in economic thinking. Of course the Commission cannot address all the problems in abstract texts. It will also continue to set policy through precedents. Due to a larger number of actors applying Article 81 (3) under the new system, case law and practice on its interpretation will catch up relatively quickly with the existing practice regarding Article 81 (1). This development will take place in the framework of the above-mentioned guidelines from the Commission and in the context of mechanisms that will ensure consistent application from the outset. Predictability will therefore increase as the case law and practice develops. 3.5. Opinions The Commission accepts that there may be a need to provide guidance to companies in cases that raise genuinely new issues that are not addressed in the existing body of secondary legislation, guidelines, case law and administrative practice. What this envisages is a system of written opinions.34 Being meant to fill gaps in existing guidelines and case law, written opinions would be reasoned and published. They would therefore provide guidance, not only for the companyfilingthe request, but also others. In this respect the written opinions would be significantly different from the comfort letters which are neither reasoned nor published under the present system. Where only a limited aspect of a case presents a new feature to which the legislative framework and existing case law provide no answer, the opinion should be limited to this aspect. In order to avoid any risk of falling back into a notification type situation, the Commission should not be obliged to give an opinion. However, in order to 34

The European Parliament, in its Resolution on the White Paper, has explicitly stated that it would favour such a system.

258

Working Paper X

provide proper guidance, the Commission would set out the conditions under which it would be prepared to consider issuing opinions. It should certainly be required that the issue raised is genuinely new in the above-mentioned sense. Furthermore, it would seem appropriate to stipulate that the opinion would be issued on the basis of the information provided by the company filing the request; the Commission should not be required to carry out fact-finding measures. The purpose of the instrument would be to provide guidance on specific questions and not to make a full analysis and assessment of the case. The Commission would not consider issuing a written opinion if a case is pending before a national court or national authority. In that case co-ordination between the Commission and national authorities/courts would be ensured by the appropriate mechanisms. The procedure for the issuing of opinions should be as light and short as possible. Applying principles of good administration, DG Competition would respond within a reasonable deadline to requests from companies. However, as already indicated, the Commission would not be obliged to give a reply as to the substances.

VI. Concluding remarks The principal aim of the White Paper is to enhance the protection of competition in the European Union. The proposed reform is fundamental. It will abolish a system that, until now, has shaped the European model of antitrust enforcement. It is a clear indication that many recent national enforcement systems have adopted the Community approach. It is therefore understandable that, particularly those subject to the law, raise a number of concerns, including legal certainty and coherence. However, it is important not to lose sight of the global picture. In the same way that an agreement must be assessed in its proper legal and economic context, the Commission's proposal for reform must also be assessed and judged in its overall context. This context includes the present system as well as the proposed new system and the other elements of the Commission's policy reform of which it forms a part. The Commission's reform will lead to a more effective enforcement system based on sound substantive rules. Very importantly, it will also mark a new era of co-operation between the various enforcement bodies in the European Union, promoting the creation of a common competition culture and a more level playing field for companies.

XI Prof. Giuseppe Tesauro Some Reflections on the Commission's White Paper on the Modernisation of EC Antitrust Policy1

In the Treaty of Rome competition is among the Community's primary objectives as well as one of the major tools through which to maintain and consolidate an integrated and unified European market. In this context, which wasfinalisedby the creation of a single market, while at the same time allowing Member States to assume full responsibility in shaping their economic and monetary policies, the protection of competition provided the opportunity and the basis for developing some fragments of a common industrial policy. Such was undoubtedly the effect generated by the existing framework of competition rules, which essentially consists of a prohibition rule coupled with an exemption provision which could be exclusively granted by the Commission itself. As to the latter point, rather than sticking to the standard Community solution, i.e. granting to member States the power to enforce both rules and exceptions, as was the case with the rules on movements of goods, persons, services and capitals, the enforcement of competition rules were entrusted to the Commission. The objective was to secure the uniform application of the Treaty's antitrust provisions throughout Member States, with a view to achieving the integration of national markets, as well as precluding undertakings and governments from finding a way to preserve the divisions and closures that the four freedoms were designed to eliminate. In other words, the Treaty and also the Commission's aim was to prevent that the common market could be jeopardised because of artificial barriers being created by undertakings through anti-competitive conducts. In the light of the above, it is not a coincidence that the Community system, taken as a whole, has conferred on the Commission the task of guiding the market liberalisation process. Thus, it is with the necessary support supplied by the case law of the Court of Justice that the Commission has provided a general and sufficiently sound reference framework for all Member States, enhancing to their utmost the potentials of the rules of the Treaty and materialising its main ideas. The demand for uniformity in enforcing competition rules within the Community, which was met initially by centralising the administrative and jurisdictional regulatory and enforcement powers, has in recent times also been 1

Some of the views expressed in this paper are drawn from a previous paper of the Author published in J. Rivas and M. Horspool, Modernisation and Decentralisation of EC Competition Law, (2000).

260

Working Paper XI

pursued by alternative solutions, such as: consolidated competition policy of the Commission; case law of the Court and Court of First Instance (CFI), especially with regard to preliminary references by national judges; the direct effect of Articles 81 (1) and 82 of the Treaty, and hence the wide-ranging application of these rules by national judges and administrative bodies; the introduction of domestic antitrust provisions consistent with Community rules, with the attribution to independent authorities of the power to enforce these rules, as well as, in some States, the expressly attributed power of enforcing also Community rules. There is nothing new or peculiar in the above developments. Indeed, there are many Community provisions whose application is not entrusted solely to Community institutions but also or only to national bodies and courts, the latter being emblematically defined as common or natural judges of Community law. This is even more evident when we deal with specific Treaty provisions or Regulations, which often have direct effect, as is the case of most rules on competition: Articles 81 (1) and 82 of the Treaty, Regulation 17/62, and the regulations on block exemptions. All this has progressively led to a common approach to competition issues and problems in the application of national and Community laws, by the Community and national institutions, be they administrative or judicial. The evolution of the relationship between the "Community" and the "national" systems of competition enforcement has reduced the relevance of the issue of harmonisation, giving way to forms of co-operation, integration, and, stretching the idea somewhat further, subsidiarity, that has always characterised the Community's experience. One might say that this has been a long and constant process of Communitarisation of competition rules and of a parallel, but directly proportionate process of decentralisation in the enforcement of these rules. It goes without saying that this process has concerned competition issues related mainly to the conduct of undertakings, and less so the issue of the effects that the conduct of Governments have on competition, for which the Commission appears to be better placed and equipped to compel the Member States to put an end to the infringement of competition. The main steps in this ongoing process are the following: a) the harmonisation of national competition rules; b) the direct application of Articles 81(1) and 82 of the Treaty by national competition authorities; c) the institutional co-operation between the Commission and national competition authorities in the application of the Community competition rules; d) the idea, put into the White Paper on Modernisation, of decentralising the implementation of Article 81 (3) of the Treaty. The right to grant exemptions has so far been an exclusive prerogative of the Commission, as the latter enjoys under Regulation 17 a "natural monopoly". However, in the last few years there has been talk concerning the possibility of

1 - Compatibility, Efficiency, Legal Security

261

enabling the Member States also to apply Article 81 (3) in order to better allocate responsibilities between the Commission and national competition authorities, either on the basis of the centre of gravity of the cases, or on the basis of turnover thresholds, along the lines of Regulation (EEC) No. 4064/89. Such idea, which aims at realising a full and complete decentralisation, and which is part of a broader "modernisation" project of the rules for the application of Articles 81 and 82, has arisen first as an answer to a need that has become increasingly stronger, from the moment national authorities were expressly recognised as having the power of directly applying Articles 81 (1) and 82. At the present time, indeed, full application of such rules does not occur, also because the national authorities sometimes fear that their action may be blocked by a strategic action undertaken by the parties involved. Such a preemptive action may consist in filing a request for an exemption with Brussels and this because the possibility of granting an exemption under Article 81 (3) continues to be an exclusive power of the Commission. In this connection, it must be said that the mechanism referred by the Commission in point 57 of the Notice on co-operation with national competition authorities in order to avoid dilatory notifications, i.e. notifications aimed to thwart the actions of the body to which the matter has been referred, have not encountered with the success expected. Secondly, the decentralisation effort is also motivated by the need to relieve the Commission, given its growing difficulties, in terms of resources, in carrying out the extensive caseload resulting from the notification system. With regard to this latter remark, the advantages of the so-called modernisation would be important, precisely because the Commission could concentrate on cases of greater importance and on those for which its action would be more effective than that of the national authorities. This in turn would enable the Commission to refocus its activities on the most serious infringements, notably by detecting and stopping cross-border cartels as well as maintaining competitive market structures. A fundamental step in this direction has been made with the White Paper on the Modernisation of the rules implementing articles 85 and 86 [now 81 and 82] of the EC Treaty.2 The Commission's proposed reform amounts to a sweeping revision of Regulation 17 and, in particular, it turns upside down the current system for the application of Article 81 of the EC Treaty. In fact, after discussing several options for reform, the White Paper proposes the abolition of the notification and exemption system and its replacement by a new regulation that would render the exemption rule of Article 81 (3) directly applicable without prior decision by the Commission. This implies that Article 81 as a whole would be applied by the Commission, national competition authorities and national courts, as is already the case for Article 81(1) and 82 in the context of an ex post control system of restrictive practices. This departure from the OJ C132 of 12 May 1999.

262

Working Paper XI

existing authorisation system, in which the Article 81 prohibition can be waved only by an act of a public authority, i.e. the commission, empowered to do so, to a directly applicable exception system, whereby agreements that meet certain criteria specified by law are outside the scope of the prohibition contained in Article 81, represents a very radical reform, as it involves the abolition of the prior control for all the agreements which fall under Article 81 (1). In other words, adopting a directly applicable exception system means, actually, not only removing the exclusive power conferred on the Commission as regards the application of Article 81 (3), but also that the undertakings would have to make their own assessment of the compatibility with Community law of their restrictive practices, in the light of course of the legislation in force and the relevant case-law. The White Paper suggests to maintain the system of compulsory prior notification only as far as "partial-function production" joint ventures are concerned: the prior authorisation requirement appears desirable in this case, taking into account that operations of this kind require substantial investment and wide integration of operations, which normally makes it difficult to unravel them afterwards. In any case, a decentralised application of the competition rules would be facilitated, because any administrative authority or court endowed with the necessary powers could carry out a full assessment of restrictive practices referred to it, examining both their restrictive effects under Article 81 (1) and any economic benefits under Article 81 (3). A number of objections have been voiced against this proposal. First of all, it is argued that the legal exception system is contrary to Article 81 so that the implementation of the Commission's proposal, in any event, would require amendment of the Treaty. The major objections to compatibility are, in particular, that the proposed system amounts to a control of abuse system in violation of Article 81(2), the tying of Article 81 (1) to Article 81(3) would deprive Article 81 (3) of direct applicability, as the substantive requirements of exemptions under Article 81 (3), implying a complex economic assessment in order to be applied to a concrete case, would not bejustitiable and the wording of Article 81 (3) clearly showing that the Treaty would opt for an authorisation system, i.e. a system based on an ex ante control of restrictive practices. As to the first argument, it should be remembered that there are two possible systems of prohibition of restrictive practices: a control of abuse system, under which an agreements only becomes illegal and void after the finding of an infringement, i. e. only for the time subsequent to a prohibition decision and a prohibition system, under which an agreement is prohibited with effect ab initio. While the authors of the Treaty made a clear choice of a system based on the prohibition system (as Article 81 (2) makes it unequivocal) the wording of Article 81 (3) leaves open the question of how the exceptions to general prohibition should be administered, i.e. whether the prohibition may be lifted only by an administrative authority holding, by constitutive decision, the agreement to be valid (authorisation system), or the prohibition is directly inapplicable to those practices that fulfil the conditions defined by law (legal exception system).

1 - Compatibility, Efficiency, Legal Security

263

Unlike Article 65 (2) of the ECSC Treaty, the wording of Article 81 (3) is broad enough to be compatible with both variants of the prohibition system. Thus, according to Article 83, the Council is granted very wide powers of implementation: the only limit being to comply with the principles of a prohibition system. In this regard, it must be stressed that, unlike the doubts stated by some commentators, the Commission's proposal does not turn a prohibition system into one of control of abuse, since, as the White Paper itself and the following consultation process with Member States made it clear, an agreement that restricts competition and that does not fulfil the conditions of Article 81 (3) will remain prohibited and void with effect ex tune, according to Article 81 (2), also in the new system. The actual difference to the present system is that all agreements meeting the requirements of Article 81 (3) will be legal, and enforceable before national courts, with effect ab initio. The argument about the deprivation of the direct applicability of Article 81 (1) that would ensue from the tying of Article 81 (1) to Article 81 (3) is based on the alleged insusceptibility of Article 81 (3) to judicial application, as the provision does not fulfil the conditions to be directly applicable. In particular it is argued the unsuitability or even the impossibility of application for the rules of Articles 81 (3) by national courts as they imply a complex economic assessment and a wide margin of discretion which defy judicial capacity and are up to an administrative body. I cannot share such concerns. I think that, first of all, the assessment implied by the four conditions of Article 81 (3) is basically one of evidence and proof, which the parties have to provide to the court, or a proportionality test. The role entrusted to national courts, however, would not be substantially different from that they have already when applying Articles 81 (1) and 82 or the Treaty rules on free movement, which nobody denies are directly applicable. It would not be a new device for jurisdictions. On the other hand, it should be stressed that national courts already now have to look closely at Article 81 (3) in order to assess if the conditions for application of Article 81 (1) to an agreement are satisfied. In this respect, indeed, as the Delimitis judgment 3 made it clear, the national jurisdiction has to check whether, "regard being had to the exemption regulations and the Commission's previous decisions, the agreement may on no account be the subject of an exemption decision under Article 85 [now 81] (3)" (point 50 of the judgment). In the new system national courts will be able to apply themselves Article 81 (3), thus preventing undertakings to use delaying tactics in order to force the jurisdiction to stay its proceedings by notifying to the Commission. A kind of tactic of courts that really hinders the direct applicability of Article 81 (1) by national courts. As to the argument that the difference in the wording of Article 81 (1) as opposed to Article 81 (3) implies a different treatment between the two paragraphs, suffice it to say that the lack of direct effect of Article 81 (3) and the need of implementing legislation in order to be operative, does not follow from Case 234/89 Delimits v Henninger Brau [1991] ECR 1-935.

264

Working Paper XI

the fact that the rule includes undefined notions, but it follows from Article 83 (2), according to which the Council has to provide detailed rules for the application of Article 81 (3), and in particular in order to determine who is competent to apply the rule. In fact, enforcement by national authorities of Articles 81 (1) and 82 does not exhaust the issue of decentralisation within the Treaty of Rome. In this respect, national judges and administrative agencies are the main actors in the application of most Community rules, while the competition rules are not an exception. For instance, suffice it to mention Article 30 (now 28), along with Article 36 (now 30), as well as the other exceptions that have been spelled out by the case law as being inherent in Article 28 itself. In sum, if one considers the overall architecture of the EC system, the modernisation reform should not be perceived as a revolution, quite the reverse, it is a return to the norm. Indeed, in other crucial fields of the EC treaty, such as the four freedoms, prohibition rules together with exceptions have been equally enforced at national level by both jurisdictions and national administrations with no major problems, by virtue of two well known characteristics of the EC rules, supremacy and direct effect. Moreover, while these provisions have raised the issue of uniform application, the important case law of the EC Court of justice has sufficiently developed to make sure that any divergent application of these rules at national level remains exceptional. In this respect, I would like to stress again the fundamental function of the Court of Justice in assisting national courts in the interpretation of Community law. In particular, the reference procedure under Article 234 EC Treaty seems to be a key mechanism for ensuring a consistent application not only of the Community competition rules, but also, more generally, for strengthening the principle of consistent application of national competition rules with the application of Community law. In this context, it must be recalled the Court of Justice's judgment in the Oscar Bronner case,4 in which the Court asserted its competence to reply to the questions referred by an Austrian court, which was contended in the case, even though the dispute in the main proceedings concerned exclusively Austrian competition law and the national authority did not have jurisdiction to apply Article 81 and Article 82 of EC Treaty: this statement was based on the fact that a national court dealing with a restrictive practices dispute by applying national competition law should not be prevented from making reference to the Court on the interpretation of Community competition law in relation to that same situation, when it considers that a conflict between Community law and national law is capable of arising. Besides all this, the reform would in any case recognise the Commission as having a prominent role in shaping the Community competition policy. In this connection it is worth noting that the White paper points out the Commission's intention to adopt block exemption regulation with a wider scope of applica4

Case C-7/97 Oscar Bronner [1998] ECR 1-7791.

1 - Compatibility, Efficiency, Legal Security

265

tion, using market share thresholds to cover the vast majority of agreements, in particular those concluded by small and medium-sized undertakings. It also intends to draw up more notices and guidelines to explain and expound its policy and provide guidance for the application of the competition rules by national bodies. Furthermore, even within the new framework, the Commission intends to adopt, in exceptional cases, individual decisions, which are not prohibition decisions of a declaratory nature, to provide the market with guidance in relation to the Commission's approach to certain restrictions. Moreover, the White Paper also points out the Commission's intention to introduce a new kind of individual decision, in which it would take note of the commitments entered into by the parties in the course of the proceedings that might otherwise end with a prohibition and render them binding. A further objection voiced against the proposed reform is that the envisaged new system would weaken the level of legal certainty of undertakings that would not be able to assess by themselves the validity of their agreement under Article 81 as a whole. Moreover, it is argued, undertakings often cannot dispose of the information on the market and other factual elements necessary to make such a self-evaluation of their agreements. It should be noted that, although in abstract terms a notification and authorisation system is more suitable to provide legal certainty to undertakings, it is a fact that, at present, the Commission only on rare occasions adopts formal exemption decisions and the undertakings mostly have to be satisfied with non-binding comfort letters. So the present system does not ensure a strong legal certainty, as some commentators argue. As to the argument concerning the difficulty which the undertakings would encounter to make a self-assessment of their agreements, it is not convincing, as they normally know very well the products, the market and their position in it, probably better than the Commission or national authorities. In any event, more than 40 years of case law and Commission decision-making practice, as well as the various pieces of legislation that have been adopted or that will be adopted in the future (wide block exemption regulations, notices and guidelines) should provide a reasonable level of legal certainty to undertakings. In any event it should be outlined that, from many points of view, the direct applicability of Article 81 (3) will improve the legal certainty for undertakings, in particular as the agreements which fall under Article 81 (1), (but fulfil the conditions in Article 81 (3)) will be legally ab initio. This will cover an especially vast scope of agreements entered into by SMEs and the civil enforceability of agreements will also be improved. Moreover, in a system of multiple competence of the Commission, national and judicial authorities, protection mechanisms are, in any case, envisaged. I am referring in particular to the possibility that the Commission may reserve for itself the right to deal with a particular case and orders that a pending case before the national authorities be deferred to it, and also the duty incumbent upon national authorities and judges not to adopt decisions inconsistent with those of the Commission.

266

Working Paper XI

Ultimately, I do not believe that the decentralisation in the enforcement of Article 81 (3) could in any way jeopardise the uniform application of Community competition rules. Quite the contrary, such a reform could and should be such as to enable a greater uniformity of application of Community competition law because the very same provisions, and not fifteen different national legislations, would be applied in all Member States. If any, the problem lies elsewhere, namely that an effective and thorough decentralisation necessarily requires a strengthening of co-operation as well as exchange of information not only between Member States and the Commission, but also among the national authorities themselves. As to the former relation, the Commission's intention to strengthen co-operation mechanisms with National Competition Authorities so as to set up an effective "network" is certainly welcome. Indeed, flexible, rapid and efficient mechanisms for exchange of information and cooperation would foster a better understanding between Authorities. To this end, as proposed by the Commission in the White Paper, it is important to reinforce the role of the Advisory Committee, which should become a full-scale forum in which important cases would be discussed irrespective of the National Competition Authority dealing with them. Moreover, the reinforcement of the network would also ensure that the consistency of competition policy be preserved, while providing a forum for solutions to conflicts in the application of Community law. Incidentally, another possible way to enhance interaction between the Commission and the National Competition Authorities would be to promote co-operation mechanisms between them also, as regards the most significant national proceedings where national provisions are applied. In the White Paper the Commission seems to hint at this possibility as it is argued that National Competition Authorities should also be required to inform the Commission of any proceedings that they are conducting under national law that might have implications for Community proceedings. In fact, on this particular issue, I believe that special attention should be paid to the interaction between EC competition rules and National competition rules, as long as antitrust cases at national level will be predominantly handled under national rules. Indeed, it is a fact that to date national competition legislation plays a crucial role when it comes to the enforcement of antitrust rules by National Competition Authorities in their own countries. In the light of the above, it should therefore be stressed that, as long as national Competition Authorities will not definitely resort to EC rules whenever possible, the problem of consistency between EC rules and national rules is quite important. As then to direct co-operation among National Competition Authorities, an idea that could also solve some of the problems, at least, is to set up a sort of "Brussels" Convention on Competition, whereas to solve all of the problems time is needed, as has occurred in other areas and rules of Community law. In order to avoid multiple control or forum shopping by the parties concerned in the envisaged system of parallel competence, it is fundamental to devise few, but clear-cut, rules to allocate cases, aside from the tool chosen to provide them, i.e. a Community regulation or a convention. In this respect,

1 - Compatibility, Efficiency, Legal Security

267

although I can share the Commission's concern as to the need offlexiblecriteria to select the competent authority, that should take into account in particular the nature of the case (the fact, for instance that a case raises a new point of policy) or the scope of the relevant geographic market, there is no doubt that workable criteria may not be the authority's resources or the location of the evidence, as the Commission seems to argue in the discussions which are following the White Paper. Such criteria could not make more efficient or effective the intervention. Furthermore, it is worth noting that an important instrument of co-operation among competition authorities consists in sharing information provided to them in connection with competition procedures. However, according to the Court of Justice5 the competition authorities of the Member States are not permitted to make use, for the purposes of the application of Articles 81 and 82, of information provided to them by the Commission which the Commission has obtained in answer to a request pursuant to Article 11 of Regulation 17 or in an application for negative clearance or exemption. A similar prohibition of disclosure of information or data regarding an undertaking involved in a formal investigation by the National Competition Authority is provided for in most of the legislation of the Member States. It follows that in order to strengthen the co-operation mechanisms among the competition authorities, this legislation would have to be amended with the aim to entitle this exchange of information, with the limitation, however, as the Commission suggests in the White Paper, that the information could be used only for the purpose for which it was originally gathered and for the application of Articles 81 and 82 or for national competition law. I think that is a basic issue, which shows that some harmonisation of national procedural laws is probably suitable. To sum up, I have no doubt that we can and must devise a system which reconciles the objective of effective decentralisation with a rigorous and uniform application of Community competition law throughout the Member States. At least we have the duty to try.

5

Case C-67/91 Direction General de la Defensa de la Competencia v AEBP [1992] ECR 1-1475.

PANEL TWO COHERENCE

1 PANEL DISCUSSION

PARTICIPANTS

Christopher Bellamy Jacques H. J. Bourgeois David Edward Claus-Dieter Ehlermann Ian Forrester Eleanor Fox Dan Goyder Barry Hawk Laurence Idot Ulrich Immenga Frederic Jenny

Anne Willem Kist Petros Mavroidis Damien J. Neven Gheorghe Oprescu Emil Paulis James Rill Mario Siragusa James Venit Karel van Miert Diane Wood

2 - Coherence

271

Panel Two: Coherence • KAREL VAN MIERT—We now start the second part of the debate, and the main subject is coherence. We have eleven speakers to introduce, and afterwards we will have ample time for further discussion. I will invite now the first speaker of this session, Sir Christopher Bellamy. • CHRISTOPHER BELLAMY—As a judge responsible for the application of the new British Competition Act—introduced with the brilliant timing mentioned by Mr Van Miert at the outset—I warmly welcome the broad thrust of the White Paper. But, as it so often happens, the devil is in the detail. I would like to make two detailed comments, the first about national competition authorities, and the second about national courts. The UK experience with the application of Article 81 (1) by the national competition bodies has not been a very happy one to date. Take, for example, the proposed alliance between British Airways and American Airlines. The national competition authority took the view that the national carrier should not give up too many slots to foreign airlines at the national airport, whereas the Commission wanted, of course, a great many more slots to be surrendered—as Mr Van Miert will undoubtedly remember. Deadlock was reached very quickly. This illustrates that the national competition authority could be in difficulty when national interests are at stake. It is not a question of actual independence from government, although in some cases that may regrettably be questionable too. It is much more a question of a perception of possible unfairness or unconscious bias at the national level. To safeguard against that, it seems to me that in any case in which parties from more than one Member State are opposed, they must have the right to go to the Commission as an impartial referee. That should also be the case where the agreement has an evident Community dimension. I think that is also the view of the Office of Fair Trading, as it was expressed before the House of Lords. Second, rules are needed to determine who does what. Take the case of an agreement between a Spanish company and a Portuguese company: in due course, the Portuguese company complains to the Portuguese authority, and the Spanish one complains to the Spanish authority. Who deals with the case? That cannot be left to an informal agreement between the two authorities, if for no other reason that the question of which authority takes the case also determines the right of appeal in the national system. In my current job, I am responsible for hearing appeals fromfivedifferent competition authorities, that is to say the Office of Fair Trading and the utility regulators (telecoms, water, gas, electricity and so forth) in three different legal jurisdictions (England and Wales, Scotland and Northern Ireland). We have very clear and quite simple rules for determining which authority takes the case and which jurisdiction is seized. Similar rules are needed at the European level. Third, if the Commission's proposal is to abolish entirely one-stop shopping, I would say that this is counter-intuitive and against the trend in related legal

272

Panel Discussion

areas, such as mergers, patents, trademarks and so forth. Without getting hung up on terminology, whether we are talking about 'positive' or 'negative' decisions or something in-between, there must be the possibility that parties can obtain guidance from the Commission, and can obtain some kind of a Commission ruling in the case of agreements involving major investment, new technology, or where there is an obvious Community dimension (such as European-wide distribution networks and systems involving other kinds of networks—telecoms, energy, transport, financial markets and sport—to name but a few). Such Commission rulings should also, in principle, be susceptible to legal challenge, by either the parties themselves or third parties. In my view, the right of appeal cannot be abandoned for the sake of administrative convenience. Is the Commission excluding judicial control of such rulings by the back door? What would happen under the new system in cases like Kruidvat v Commission and European Night Services!1 The Commission cannot abandon its responsibilities. Finally, will the decisions of national competition authorities be published, and in what language? In addition, the right to address a public authority in one's own language is not to be taken lightly or disregarded. Moving briefly to national courts, as someone who has been judicially responsible for applying Article 81 (3) I can hardly say that other judges are not capable of doing the same. In any event, all judges know in their hearts that they can do everything. But, in so far as non-specialised national courts are concerned, bear in mind the limitations. First, national courts have available only what the parties put before them. They have certain possibilities to ask for expertise and so forth, but even so they will only have a limited and rather patchy vision of the problem. Second, thefinanciallystronger party generally has the advantage before the court. Do not over-invest in a system that gives the advantage to the big pocket. Third, in the parallel case of patent law, certain judges are in despair about the problem of forum shopping. I shall not go into the details of the Italian torpedo and the Dutch Rocket docket scenarios2 explored by the House of Lords, but there are enormous possibilities for playing forum-shopping games that are not dealt with by the Brussels Convention. Fourth, I would like to raise another issue, which I usually call 'the Cardiff issue'. In the case I am referring to, the Cardiff rugby football club decided not to obey by the rules of the Welsh Rugby Football Union regarding the playing of matches. The Cardiff club alleged that those rules were restrictions of competition under Article 81. Indirectly, this claim also put into question the rules 1

Case C-70/97 Kruidvat v Commission [1998] ECR 1-7183. These are expressions to describe, first, the tactics whereby proceedings may be commenced in a jurisdiction where there are lengthy delays—such as the Italian jurisdiction—in order to avoid proceedings being undertaken in a jurisdiction where a decision can be reached more quickly. Second—and conversely—proceedings may be started in a jurisdiction where orders are issued extremely quickly—such as in the Dutch jurisdiction—thus pre-empting the commencement of proceedings in another jurisdiction. 2

2 - Coherence

273

of the International Rugby Federation. The International Rugby Federation intervened in the case, which was pending before a small Welsh court in South Wales, and said: 'we have just notified our rules to the European Commission, and hope to gain an exemption. Please stay the proceedings.' The Welsh court felt obliged to stay the proceedings pending a ruling by the Commission. Therefore, my question is: how should we cope with cases that arise before national courts, where there is an evident Community interest and there are strong reasons for the decision being taken at the centre—by the Commission— but when a small national court (that might know nothing of the background and might not be able to address the issues adequately) is in fact seized with jurisdiction? Will there be some provision for the suspension of proceedings before national courts in such cases? In addition to these points, many national courts are already hopelessly blocked by numerous pending cases, sometimes for up to ten years. Decisions taken by national courts could be inaccessible for language reasons. Requests for preliminary rulings under Article 234 of the EC Treaty, whether to the ECJ or the CFI, are in fact not particularly satisfactory in time-intensive cases. The Masterfoods case3 has taken nearly seven years to reach the ECJ for a preliminary ruling. It is also probably rather difficult to involve the Commission in national proceedings. In any event, nowadays probably over half of the national disputes are dealt with through arbitration or some alternative dispute resolution procedure that does not involve the courts at all. My conclusion would be: do it (the reform), but do not expect too much of the national courts. You might have to live with some decisions that are not very convincing and with decisions that are inaccessible. One should seriously consider mechanisms for channelling important cases towards the Commission, and perhaps towards specialised jurisdictions at the national level. One last thought: perhaps one should also consider for the future the possibility of an exchange of judges of different nationalities. • JACQUES BOURGEOIS—Having followed the debates on the White Paper and the debate that took place here, I think that the choice is, in fact, between a consistent and coherent but inefficient enforcement system, and sufficient enforcement with a risk of inconsistency and incoherence. I assume that the result of the political debate on this is a foregone conclusion: there will be decentralisation along the lines of the White Paper. Basically this is the right choice under the circumstances. I am not going to go over the whole of my paper: there is no time for it and, the reform process being at the decision-making stage, a certain number of points that I made in my paper are depasses. I will only mention five brief points. First, there could have been a choice between two possible reform solutions. One was to re-define the scope of application of the Community competition Case C-344/98 Masterfoods Ltd\ HB Ice Cream Ltd, still undecided.

274

Panel Discussion

rules, narrowing them down and opting for a solution a la merger control. The other was the possibility chosen in the White Paper: splitting up the jurisdiction to enforce the Community competition rules, with the Commission on the one hand and national competition authorities on the other. I believe this is the right choice. There are many reasons for it. One of them, which I thought was interesting, is that a study carried out in 1994 on behalf of the European Commission found that one third of the cases that were handled by the European Commission over the relevant period actually concerned only one Member State. Two thirds had a competitive or economic impact in more than one Member State. Second point: coherence and consistency of what? It is striking to me is that there is not a word in the White Paper about coherence and consistency in cases of 'non-action' by national competition authorities and by the European Commission. The White Paper stresses the increasing importance of complaints in a system where there will be no notifications. The Commission is under a duty to act on a complaint, as was made clear by the Court of First Instance in Automec II4 and in subsequent cases. There should also be a duty for national competition authorities to act on complaints, and that should be made clear in a Community legal text. I know, and Laurence Idot has pointed it out, that the White Paper still sticks to the concept of procedural autonomy for the Member States in enforcing Community law. But there are examples in other areas of law where the principle of national procedural autonomy has been set aside. Why not set it aside it here as well, and establish a duty for national competition authorities to act on complaints based on EC competition rules? Third point: Commission guidelines, notices and non-infringement decisions should be binding on the national courts and national competition authorities. If they are not binding on national competition authorities, then what are they worth? On the other hand, we might end up in a situation as in the Brother judgment of the European Court of Justice5: one cannot challenge the Commission's policy guidelines in an EC court because they are not binding, and one cannot challenge the decisions of national authorities following the Commission's guidelines because such guidelines are not binding. Fourth point: forum shopping. We do not really know the extent of the risk of forum shopping as far as national competition authorities are concerned, and I am not even talking about courts. But again, the figures of the Commission study from 1994 tend to indicate that there is indeed an important risk of forum shopping, as the economic impact of the agreement was felt in more than one Member State in two thirds of the cases. Fifth point: there should be a one-stop shop. In the debate following the publication of the White Paper, the Commission suggested that, if more than one national competition authority takes up a case, problems could be avoided by 4 5

Case T-24/90 Automec v Commission (Automec II) [1992] ECR 11-2223. Case 229/86 Brother v Commission [1987] ECR 3757.

2 - Coherence

275

the Commission 'pulling in' the case. But isn't this defeating the very purpose of decentralisation? There should be one-stop shop, and the conditions for it should be created. Thefirstcondition is that there should be a clear set of rules on the allocation of cases: what has been proposed so far boils down to some vague criteria, and conflicts of jurisdiction should be solved on a case-by-case basis in the Advisory Committee. This is perfectly non-transparent and, I think, not acceptable from the point of view of practitioners. The second condition is that the national competition authorities should have the instruments to act under a one-stop shop system. They should also have the possibility of carrying out investigations in other Member States, or should at least be able to require the help of the national competition authorities in other Member States. Such co-operation has already been achieved in other areas. If this is possible under EC customs law, why should it not be possible for competition purposes? The third condition is that full faith and credit should be attributed to the decisions of national competition authorities. They are not enforcing national law, but EC law, which is binding throughout the Community. Consequently, their decisions should have binding effects throughout the Community. • LAURENCE IDOT—Out of the many issues related to the White Paper, I would like to make some remarks concerning three points. First, we have to think about the reform in a context of dual control. I cannot agree with the opinion according to which the implementation of Articles 81 and 82 raises some type of problems as in otherfields,because Articles 81 and 82 are the only provisions of the Treaty in relation to which there is parallel application of EC and national law. It is not too difficult to apply the kind of dual control resulting from this parallel application of EC and national law when you prohibit certain types of conduct, but it is much more difficult to reach an agreement about exemptions. This is the reason why personally I am in favour of what can be called a one-stop shop system of rules. But I know that this is a dream. If the current legal system is preserved, there are two points we have to think about. The first relates to the interpretation of the scope of application of Article 81, that is to say the interpretation of the requirement that trade between Member States must be affected. Here I fully agree with Mr Paulis: there is great uncertainty about this jurisdictional requirement. This is because the European Courts and the Commission have often applied this test either in order to determine the scope of application of EC competition rules or as a substantive criterion indicating the illegality of certain agreements (that is, agreements affecting trade between the Member States). When it comes to Articles 81 and 82 EC Treaty, national competition authorities frequently still have a restrictive perception of the jurisdictional scope for the application of EC law, and I could give you some examples in this sense. I am not going to deal here with the problem of allocation of cases between national competition authorities, or with that of the legal effects of both the

276

Panel Discussion

Commission's and the national competition authorities' decisions, as a lot of work has been done on these subjects since the publication of the White Paper. I cannot agree with the Commission's view that there is no need to harmonise national procedural rules. When this issue is raised, we do not know in fact if the Commission refers to national competition authorities or to national courts. I agree that, for now, there is no need to harmonise procedural rules in so far as national courts are concerned; but I strongly disagree that there is no need to harmonise procedural rules for national authorities. This is because there are many differences between the competition authorities of the Member States and, in particular, there are many differences in so far as the margins of their powers of intervention are concerned. Do national competition authorities have the power to select their cases, or not? Do they have to address all the complaints received, or not? Can they issue conditional decisions—or decisions with commitments—or not? And there are other examples in this sense. As far as coherence between the decisions of national courts is concerned, I would like to insist on two points, which were forgotten in the Commission's reasoning. First, I have the feeling that we over-estimate the potential of the Brussels Convention to solve the problems. We should not forget that for the practitioners one of the main provisions of the Brussels Convention is Article 5 (1) which introduces an option for contractual disputes. This provision is difficult to apply in the way that the ECJ has indicated, and there are in fact many divergences in the jurisprudence of national courts on the application of this provision. All specialists in private international law agree that Article 5(1) of the Brussels Convention should be abandoned. Second, most of the cases brought before national courts in Europe deal mainly with contractual issues (Is the contract valid or not? Is it possible to enforce this particular clause or not?). In most cases, competition law is only a defence against a contractual liability charge. This fact changes the perspective on coherence. Disputes over agreements, for example, generally do not go before national courts. Most agreements contain an arbitration clause, and I fully agree with what Judge Edward said this morning, when he pointed out that there was a gap in the Commission's proposal with respect to arbitration. We get the feeling that there are two separate worlds: the 'competition' world on one side, and specialists in arbitration on the other side. I am lucky enough to belong to both worlds, and I can say that we have a lot of problems with the application of EC competition rules in arbitration proceedings, in spite of the Court's decision in the Eco Swiss case.6 I would like to add a few words on the idea of restricting the application of Article 81 (3) to specialised national courts. I am not so sure that this is a good idea because, I repeat, in many cases EC competition law is only a defence. If it is only a defence, restraining its application to specialised courts will add some delays in the court proceedings concerning the main cases. Finally, I wanted to add something in relation to the discussion this morning. We have to be aware that there will be a change in the interpretation of Article 6

Case C-126/97 Eco Swiss [1999] ECR 1-3055.

2 - Coherence

277

81 (1), and there will be a new balance between the principle of prohibition on one side and the rule of exemption on the other side. I think the French example demonstrates it very clearly. • ULRICH IMMENGA—Although this affirmation might surprise you coming from myself, I do not think that the reform will really imply a sacrifice of coherence, at least not if some modifications are brought to the original reform proposal and if some of the ideas already discussed here are taken into account. I will be very brief, but first I would like to refer to the controversy about specialised courts in light of the experience we have had with such courts in Germany. German competition law empowers the Lander to establish their own courts specialised in thisfield.I am referring here only to private law suits, and not to the judicial review of decisions of the competition authority (this has to be distinguished very clearly, of course). We do have this kind of specialised court. The system is well established, and it should be introduced in all Member States. I think coherence requires the effective application of EC competition rules by national competition authorities and national courts, clear criteria for the allocation of cases and competencies in the area of EC competition law, and a clear distribution of competencies between the Commission and national bodies. In my view, the direct applicability and justiciability of Article 81 (3) is a problem of coherence. Insofar I refer to previous discussions. What is strange for me is that, in relation to case allocation and the uniform application of Community competition law, the White Paper generally refers to the relationship between the Commission and national competition authorities, and respectively the national courts, but there are no substantial remarks on the relationship between the national bodies themselves. This relationship is crucial with respect to both the allocation of cases and the uniform application of the EC's competition rules. To briefly mention a point often neglected in the discussion: the necessity to transfer cases might arise even if criteria for the allocation of cases are established (perhaps according to a 'centre-of-gravity' model). One must consider the consequences of such eventual transfers of cases. The complainants do not always know where they have to take their case to, or they might choose the wrong forum, and then cases will have to be transferred. In such cases, there must be a special system for the exchange of information and, in particular, solutions must be found for the problem of confidentiality. I think this is a very important point in the context of the debate on allocation of cases. As to coherence, I will be very brief because I can refer to what Mr Bourgeois has just said. I am a firm supporter of the view that the decisions of national competition authorities should have binding, Community-wide effects. I think this is a pre-condition for coherence in enforcement. Alternative solutions have been presented. One is a kind of mutual recognition system, yet according to

278

Panel Discussion

what was referred to here about the Brussels Convention, it would really be a problem to extend this model to the EC's competition rules. Once again, problems will arise with respect to private court proceedings. You might have the defence of the national ordre public. This is really a very delicate question. Another proposal was to establish Community courts in all Member States. This would be similar to the American system where there are Federal and State courts. Yet another proposal was to allow national bodies to make requests for preliminary rulings on the interpretation of Community competition rules, and to involve the Court of First Instance in the issuing of such preliminary rulings. I think that the most interesting proposal was made by the Anti-Trust Section of the American Bar Association, and I would call it the solution of 'conditional Community-wide effect'. This system would consist of publishing national decisions throughout the Community, with the opportunity to appeal against such decisions within a fixed period of time (let's say, within two months). If appeals were not made within this period, the decisions would automatically acquire Community-wide effect. In the case that an appeal is made, it might be considered either by the Commission, or—and I would prefer this option, as Mr Bourgeois does—by the Court of First Instance. I think this is a very intelligent and sophisticated system for introducing the Community-wide effect of national decisions. Very briefly, two other points. I think there is a very valid reason for the Commission to reserve jurisdiction over certain cases. The White Paper refers to the most serious restrictions of competition on the single market, where the Commission would reserve the right to prohibit the agreement approved through positive decisions of national bodies, subject to the principle of res judicata between the parties. Furthermore, there is a general reference in the White Paper in the sense that, in the case of agreements prohibited by the national bodies, there will only be references to the ECJ based on Article 234.1 think that Commission guidelines explaining situations in which it is likely that the Commission will claim jurisdiction are necessary. I think this is an issue of transparency and, as Mr Monti briefly reminded us this morning, there is also a question of subsidiarity. A final sentence on the necessity of procedural harmonisation: I think procedural harmonisation is very important in order to prevent forum shopping. An efficient application of EC competition rules necessarily requires at least a certain degree of procedural harmonisation at a national level with respect to enforcement, and perhaps also with respect to remedies. • FREDERIC JENNY—I must say I am a bit frustrated with the discussion so far. I understand what everybody would like. What I do not understand is the organising principle of the discussion. How are we going tofindlight at the end of the tunnel? Possibly this is because I was trained as an economist rather than as a lawyer. It seems to me, and I am going to re-phrase some of what I have heard in economic principles, that what the Commission is trying to achieve is

2 - Coherence

279

a more efficient and relevant enforcement of EC competition law. The means that the Commission talks about switching to a legal exemption system in order to have decisions that are more rooted in economic analysis and theory, and in order to increase in the investigative powers of the Commission. One must recognise that this will have various effects. One of them is a shift of the costs of enforcement of EU law from the centre to the national bodies, a shift of the cost towards private enforcement, and also the possibility of allowing the shift of resources within the Commission towards the most important cases of anti-competitive behaviour. The rule of reason and decentralisation are considered to be necessary means for a more efficient and more relevant enforcement system; though, as Barry Hawk said, there is tension between economic and legal certainty. I would also say there is tension between decentralisation and legal certainty, and there is also the issue of coherence in enforcement. If there are tensions, we also have trade-offs. As an economist, I have been trained to say 'if there is a trade-off, the question you must ask is: what is the social cost of the measure compared to the social benefit of that measure?' In other words, what is the social cost of decentralisation of enforcement and a more economic effects-based approach in enforcement? Is this cost commensurate to the social benefits of the reform? I do not believe that we have asked ourselves these questions. We have talked about possibilities, but not about actual costs. On the benefit side, Alexander Schaub has mentioned in other forums that the Commission was spending most of its time on worthless procedures and that this was due to the notification procedure. We all know that in Europe there are very few decisions on hard-core cartels, whereas there are many more in other jurisdictions of the same size; so potentially there is under-enforcement in Europe. There is overenforcement in cases that might not be so important, and under-enforcement with respect to practices that impose a social cost. This is on the side of the social benefits of the reform. The problem is that we have no idea of what its social cost will be, whether there will be a lot of incoherent decisions by national courts or national competition authorities, or whether the cost that will be imposed by those incoherent decisions is sufficiently high to worry about. I would also add that, from my point of view, inconsistency might not even be the real problem; if we will have a problem with the consistently wrong enforcement of EU law, we won't be very far from the current situation. When we look at the possible instruments for achieving the objectives of the reform, it seems to me that it would be necessary to rank the possible instruments from the least costly to the most costly ones, and to solve the problem with the least costly. If we look at the different instruments that were proposed, we heard of specialised courts, we heard about creating a network of competition authorities and co-operation, we heard about the possibility that the Commission would intervene in national court proceedings, about the possibility of appeals, etc. I would suggest that there are two instruments that are practically costless. One of them is specialisation, or the concentration of cases in the hands of a few judges. There are some countries where this is already done (but of course

280

Panel Discussion

they do not have a good name, so you cannot do this). What you could do is to specialise a few 'competition'judges in a court. We happen to have some experience with this in France, because this is the way things have developed both at the level of the Competition Council and also at the level of the Court of Appeal. I think that developing a competition culture and developing the ability of judges to ask questions to the extent that they could thus influence the outcome of the process of gathering information in competition cases is very important. And it is not going to cost much. My second suggestion is the organisation of co-operation between national competition authorities. Clearly, there are going to be transnational competition issues. Clearly, there is practically no co-operation among national courts and national competition authorities across the Member States, or between them and the Commission. It is slightly ironic that the EU is preaching to the rest of the world that it should have a system of international co-operation for antitrust enforcement in the context of the WTO, when we all know that there is no co-operation going on within the EU itself. Clearly, it would not be very expensive to try to develop methods of co-operation. This would also be a useful example in the context of the debate about international co-operation in general. When we move to other instruments, what they have in common is the fact that they imply that the Commission is going to take an active stance, either as amicus curiae or on appeal, and so on and so forth. Those are more costly instruments. There is obviously a need to have some instruments whereby the Commission can intervene, but if we adopted all the instruments that we have talked about, maybe the benefits of the reform would be equal to zero, because it would all be so costly to the Commission which would have to deal with national cases all the time. I would suggest, and Barry Hawk also suggested this earlier on, that the EU should push for specialised courts and co-operation, and should try to set up the logistics of co-operation between competition authorities, and intervene in national proceedings. Which instrument should be used for such intervention? Should it be through appeals against national bodies' decisions? (By the way, I think that if the Commission could appeal the decisions of national bodies, this should be done at the Court of First Instance.) That would solve some of the problems mentioned by Mr Burrichter earlier on. Or, the Commission could intervene as amicus curiae in national proceedings. In any case, I think such intervention should occur at the appellate stage, because that is much less costly than trying to intervene in the proceedings at a lower level. As for the rest, let us wait forfiveyears and see whether the costs of incoherent decisions are such that something more is needed. After all, this is the approach we took in the field of merger control. • KAREL VAN MIERT—I should like to invite now Mr Kist, Head of the Dutch competition authority, to make his presentation. Let me remind you, Ladies and Gentlemen, that the Dutch Competition Authority was only recently ere-

2 - Coherence

281

ated, but it is so successful that the Dutch business community is looking to behave in a way they never did before. Within a really short period of time, they have learned what a national competition authority is, how it operates, and what its meaning is. As occurred in some other Member States as well, I think we have here really excellent proof that, on the basis of clear national rules in combination with European competition rules, a credible competition authority can be very successful. Spreading this around could build a true network of competition authorities, as we are trying to do.

• ANNE WILLEM KIST—After your introduction, what can I say other than 'Your Honour, I rest my case'. I have always been a sceptical supporter of this reform project. From the outset, I have taken the position that the reform and its objectives are too good to be put in jeopardy by insufficient preparation. What we do in meetings like this, or what we are doing with Emil Paulis and others in Working Groups, is to learn exactly which tunes are played and which tunes will be played if we are going to face the music. I would like to make three points about how to face the music in time, and not after we have already embarked on the reform project. The first point is that we need as much clarity as possible, and I cannot stress enough the importance of clarity before launching the reform. Let me give you a few examples. We are now discussing rules on the allocation of cases, which should of course be very, very clear. Also mentioned, and again of crucial importance, is the need to spell out criteria for individuating cases with a Community dimension, and the role of the Commission in relation to such cases. Clarity on the burden of proof is very important as well. We are heading towards a new, more economic effects-based approach in enforcement, and let us be frank: nobody knows exactly which pro-competitive elements are to be handled under paragraphs 1 or 3. Yet, from the point of view of the burden of proof, this is a very important issue. Clarity is also needed with respect to the way in which national competition authorities—and I hesitate to say national courts, but I think they will also face this problem—are going to deal with the EU policy objectives newly incorporated into the Treaty. I think what we need from the Commission is guidance and clarity. One can think of many other aspects in need of clarification, but we are already in the process of clarifying them. The purpose is to create clarity beforehand, and we face great danger if we do not take enough time to realise what it is all about. Second, and this is in fact an elaboration of my first point, I think we should also consider the procedural and institutional ramifications of the reform project. As a lawyer, I know there is substance only through procedure, and in many cases it is the procedure that decides the substantive outcome of a case. I would therefore call for a little bit more ambition on behalf of the Commission in setting some minimal harmonisation requirements as to national procedural rules. In addition, we should also face the problems posed by divergent institutional arrangements in the various Member States. There are Member States where,

282

Panel Discussion

as in the Netherlands, all enforcement powers are gathered in one institution. But there are also Member States where they are divided between two institutions, and even cases where one of these institutions is a judicial body. I think it is really dangerous not to consider the effects of such different institutional arrangements. Finally, if we want to make this work, our greatest efforts should be oriented towards preventive measures, that is, making the network of authorities work. I think that the way to go there is to go all the way. We should decide to give national competition authorities the power to issue prohibition decisions and even non-infringement decisions with EC-wide effects, but under conditions guaranteeing that there is a sufficient level of coherence and consistency in enforcement. • KAREL VAN MIERT—I would like to introduce a special guest. He is the only participant coming from a country that is a candidate for EU membership. And we all know that, while we are discussing the reform of EC competition policy, there is another big issue outstanding: the enlargement of the European Community. As this discussion has shown, we all care for coherence and how to make the reform work in light of the very different legal systems and traditions inside the Community. With enlargement coming up, and certainly materialising in the coming years, there is even a bigger challenge from this perspective. • GHEORGHE OPRESCU—I have noted fourteen issues to comment on: half of them are a brief presentation of the current competition law enforcement situation in Romania, which might be applicable to the other candidate countries. The other are comments about the potential impact of the proposed reform in the candidate countries. Romania and the other candidate countries currently implement 'cousins' or 'brothers' of Article 81(1) and (3). I say 'cousins'—and my paper contains a brief discussion on this—because, in the Romanian legal system, anti-competitive agreements may be exempted if they fulfil four cumulative conditions, plus one out of another five. Therefore the approach is even tougher than under the EC Treaty. The current Romanian legal system relies on notification and prior authorisation, including for agreements falling under block exemptions. This means that companies should not defy the law even in the case of block exemptions. The national competition authority by means of formal decision gives exemptions, and these can be appealed to the Court of Appeals. I would therefore say that, under the current system, the national competition authority—or authorities, if I may speak in the name of more than one candidate country—play an important role. The system reserves a moderate role for the courts, even though—we should not forget—the courts might be the ones taking the final decision. I read in your written contributions to this Workshop that the European Court of Justice has imposed some limits on its

2 — Coherence

283

judicial review when dealing with decisions based on Article 81 (3). National courts in Central and Eastern Europe did not have such limitations in mind, yet they would nevertheless take a more mechanistic approach to the review process. In Romania, the prior authorisation system did not create an overwhelming number of notifications, though there are three kinds of notifications that must compulsorily be made to the national competition authority: concentrative agreements, applications for individual exemptions, and agreements falling under block exemptions. Most of the notifications received refer to block exemptions, where the analysis is more or less formal, or more or less close to being 'a stamp of approval'. At the same time, the number of notifications for individual exemptions, is close to zero. This means that the actual system of compulsory notifications, treating all agreements in the same way and without making many distinctions between them—though we are aware that there is a clear distinction between concentrative agreements and those falling under block exemptions—has a sort of perverse effect. If we look at statistics we may see, for example, that over recent years the number of cases on anti-competitive practices has constantly decreased, while the number of cases on economic concentrations or other notifications simultaneously increased. The newlyestablished competition authorities in Central and Eastern Europe might take the path towards a bureaucratic approach in enforcement, being satisfied that they have a lot of work to do and not being aware that they are still missing the real competition issues. What we have to do is in the European Agreements. The Agreements require the associated countries to adopt competition rules similar to those of the Community, which means that the countries must (somehow) comply with stricter rules than the Member States themselves. As for enforcement capability and coherence, we should be aware that, if all the notifications that needed to be made to our competition authorities had indeed arrived, these authorities would have been overwhelmed by their workloads. In such a case, it would clearly have been hard to discern the 'good' cases from the 'bad' ones, and thus resources would have been diverted from more important matters. We should also not forget that, from the point of view of the applicant countries, the enforcement procedures are already decentralised. In a short-term perspective, however, it is unreasonable to expect that newly established competition authorities of Central and Eastern Europe will apply the law uniformly and with an eye on Community competition law. In all likelihood, they will look at competition issues from the national point of view first, rather than from that of Community law. If this is a serious problem in the current Member States, it will be even more serious in the candidate countries. On the other hand, I would say that the proposed reform represents a shift from a regime in which 'what is not authorised is forbidden' to one in which 'what is not forbidden is authorised'. The former system is consistent rather with the type of regulation prevalent under the communist regime, and which still retains a strong hold in the post-communist mindset. The old system of applying 81 (3) simply

284

Panel Discussion

reinforces this mindset, which is not what is needed for creating a market economy. To conclude, the implementation of the reform might be problematic in the Central and Eastern European countries in the short-term perspective, but it would not be a problem in the medium-term when it comes to national competition authorities. And it would perhaps be even less of a problem when speaking about national courts. • EMIL PAULIS—First of all, I would like to thank all the participants at this Workshop, because the Commission is extremely interested to see which angle you approach the proposed reform from, and this debate will help us in the drafting of a proposal for the new Regulation. I would like to outline the general framework conditions for setting up a coherent enforcement system. And I would like to do that basically following the philosophy outlined by Mr Jenny. That is, how can we reduce, at the lowest possible cost, the risk of conflicting decisions, so that the reform becomes acceptable? The first condition is to strive for the widest possible application of EC law. It is obvious that the more you apply the same law, the better the basis for agreeing on its application, and vice versa, because when different laws are applied, the risk of incongruent application increases. On this, I particularly wanted to point out that, in our view, we should not narrow the scope of application of EC law by narrowing the interpretation of the condition that trade between the Member States must be affected for EC rules to apply. This would lead us in the wrong direction, or towards the re-nationalisation of competition law. What we should do, and here I fully agree with Ian Forrester, is to reach a more reasonable interpretation of the scope of application of 81 (1), on the basis of an economic approach. However, that is a different issue. In the direction of the widest possible application of EC law, we should also tighten the rule of primacy of EC law over national law. Another general framework condition, and one that has also been stressed this morning, is that the Commission should make the rules of substance as clear and predictable as possible. Here I would like to make two points. The first is that the Commission (and, in the future, the Commission together with the national cartel authorities) should focus on what is prohibited, and not so much on what is exempted. By focusing on what is prohibited, we place ourselves in line with the new approach, which is to block-exempt everything except what is explicitly prohibited. We should also focus on the essentials in legislation, and in the approach to rule making. The second point that I wanted to make, and here I want to support Barry Hawk's statement this morning, is that there is a trade-off between clear rules (that promote coherence), and the price we have to pay for them in terms of flexibility and the correctness of decisions from an economic point of view. I think we should go as far as possible in making rules predictable, but not so far that we produce incorrect outcomes.

2 - Coherence

285

The third in the series of general conditions is very important: in the field of competition law, the Commission should always preserve its autonomous power of enforcement. It is very important that the Commission can adopt its own decisions, and set policy, and try to encourage the coherent application of the rules. That is a very important feature improving coherence in a system of parallel competencies. The fourth condition, which is again essential and perhaps not sufficiently understood as yet, is that we propose an enforcement system whereby constitutive exemption decisions are altogether eliminated. There will be no more exemption decisions, but only exceptionally certain positive decisions. Yet these will not be taken in the interest of legal certainty for companies, but in the interest of setting policy and furthering the coherence of the system. This might, of course, overlap with the interests of companies in terms of legal certainty, but the elimination of constitutive exemption decisions is essential in order to reduce the potential for conflicting decisions. These are the decisions that would create the highest potential for conflict in a system of parallel competencies. The fifth general condition is fundamental if we aim for a more economic approach in the application of Article 81 (1). We cannot impose on companies a system of parallel competencies if we keep, in substance, what I call 'an unreasonable scope of application of Article 81 (1)', because then the companies are likely to systematically fall under the prohibition rule and they would systematically have to invoke Article 81 (3) in defence. Allow me to state that, if I say we should come to a reasonable approach in the application of 81 (1), it is not equivalent to promoting a relaxation of the rules. 'A more reasonable approach' means 'a more economic-effects oriented approach': a less strict application of the rules when there is no market power involved, but stricter application when market power is involved. The last general condition, which unfortunately I will not have enough time to cover, is the most important: the Commission's monopoly is going to be replaced by co-operation. The whole Community is a matter of co-operation. Article 10 of the Treaty provides for an obligation to co-operate, and not for a faculty to co-operate. The co-operation obligation is reciprocal, meaning that the Commission also has an obligation to assist national courts and national cartel authorities. And here I return to what Mr Jenny has said: it is true that we should look for the most efficient and least costly instruments to guarantee coherence, however we (the Commission) should also be responsible; we cannot go for a system of parallel competencies without at least assisting the national bodies. To reply to Mr Kist: he said we should go all the way. Yes, let's go all the way, but not in one day. Europe was not constructed in one day. We have to do it step by step. • JAMES RILL—In the US we have been struggling with modernisation since around 1787. Prior to that time, uniformity came to the colonies through some other force. The fact is, however, as it relates to antitrust enforcement, that the

286

Panel Discussion

federal system is a relatively new phenomenon. I would say that it dates back not more than about twenty years, when State Attorneys General asked the Assistant Attorney General for Antitrust what they should do in this field. He replied 'close up', which of course led exactly to the opposite course of conduct. Furthermore, it has become clear only during the last ten years, following the Supreme Court's decision in California v American Stores Co case,7 that the States have plenary authority (for example, in the divestiture of assets obtained in an illegal merger). Yet, just by way of example, there were only four merger cases conducted by the States during all the years between the adoption of the 1914 Clayton Act and 1980. In the 1980s, there were twenty-nine cases, and in 1988 the States brought two more merger cases than the two federal enforcement agencies could buy. I do not view this as a salutary result. The situation has improved since then, and it has improved in ways that may be instructive to this forum. But I would like first to make one structural, or jurisdictional, remark: the enforcement of US law is not hierarchical but horizontal. Following a variety of Supreme Court decisions, the States have authority to bring antitrust cases under federal law—either on their own, or on behalf of their citizens under the notion of parens patriae—that is equal to that of the federal enforcement agencies. I would also say there is another difference between the EC and the US systems, in that there is legal and political leeway for disagreement in the US, though that might not be so much of a difference. For about ten to fifteen years there have been other State political concerns that have intruded on antitrust enforcement beyond consumer welfare. The State Attorneys General are, for the most part, elected officials. It is not by accident that the National Association of State Attorneys General is known as the 'National Association of Aspiring Governors'. As a result, we have had a number of conflicting decisions over time. I could name a couple. In the 1990s, the National Association of Attorneys General issued Merger Guidelines that were deliberately, intentionally and directly in opposition to much of the 1992 Merger Guidelines issued by the Department of Justice and the Federal Trade Commission. It got a little worse in 1994 when the federal government issued Vertical Restraint Guidelines (which, I must confess, many of us felt were not wise anyway). The States did not merely issue conflicting guidelines, they issued outright 'nuclear war' guidelines, deliberately accusing the federal government of 'selling out' in the area of vertical restraints. This is not a situation that I would recommend Europe should follow. In some cases, the States also issued remedies that were not exactly what would be called 'competition remedies'. For example, in the West Point merger case, the consent agreement entered into with the State of North Carolina required a company to keep an inefficient plant open for North Carolina's benefit. In the case of the Whitman Chocolates/Russell Stover Candies merger,8 7 8

California v American Stores Co., 495 U.S. 271, 295-98 (1990). Commonwealth of Pennsylvania v Russel Stover Candies, Inc., 1993-1 Trade Cas.

(CCH) (E.D. Pa. 1993).

2 - Coherence

287

the State of Pennsylvania required that Whitman buy a certain amount of raw chocolate from Pennsylvanian retailers for the processing of candy. There is nothing wrong with Pennsylvania chocolate, I just wonder about the consumer welfare implications. What is the situation in the US today? First of all, the States themselves are in need of some co-ordination, and we have a National Association of Attorneys General fulfilling that function. There are a number of working arrangements that have been entered into between the States and the federal government, which I think is quite clearly a move in the direction of greater convergence and coherence. The federal government does not have the authority to withdraw cases from the States. There is no likelihood of any Treaty between the federal government and the States for an appropriate allocation of cases, although that might make some sense. Elimination of uncertainty is illusory, but I think work can be done in that direction in the process of moving towards greater coherence. I think coherence is desirable, and I tend to disagree with Barry Hawk's comment that the business community is less concerned about uniformity than they let on. Let me just give you some examples of steps towards convergence and coherence that have been taken in the US; they might be instructive for what is going on now in Europe. I would certainly echo the comment of Emil Paulis that it does take time, and it has taken a lot of time in the US. We think that relatively formal working groups are a good idea. In 1989 we established a joint working group on antitrust between the Department of Justice, the Federal Trade Commission and the Committee of State Attorneys General. The result of that working group was to produce very similar Merger Guidelines. As a result of those relatively formal working groups, by the early 1990s there was a minimal amount—and almost elimination—of the stridency that had existed prior to the 1990s. This brings me to a second point: a mechanism should be established for the adoption of common standards, guidelines and the like. Speeches are very helpful. Most guidelines are good. I don't want to get into the Joint-Venture Guidelines recently issued by the Department of Justice and the Federal Trade Commission. Sometimes judges look at guidelines with scepticism. I remember how Judge Thomas Jackson spoke against the 1992 Horizontal Merger Guidelines. He viewed them as a statement against 'private interest' made by the government from a public platform. I would also very much encourage the use of so-called 'non-infringement decisions' and more, but on a selective basis. An attempt to do this across the entire panel of enforcement work will cost more than what it will produce. Decisions that are not particularly useful should be eliminated where a landmark decision has already been taken. Decisions with an articulated reasoning would also be very helpful for the business community, for the Bar, and for coherence purposes. A couple of examples: when the 'big eight' accounting firms merged into the 'big six', we described the unilateral effects theory on which we acted. The theory was, if not novel, at least not ordinarily expressed before that time. When

288

Panel Discussion

we did not bring the case of the acquisition of Uniroyal we got into large buyer response to possible price increases in public speeches, and anticipated the 1992 guidelines. I know that Commissioner Van Miert had also issued statements on the accounting cases, which I consider a positive step. I think that Business Reviews or Business Letters, or what would be considered to be advisory opinions, are quite frankly of limited use. In the US, they take too long to prepare, and they are usually not used because there is a question of whether they are binding or not. Furthermore, they tend to deal only with the most commonplace of situations. At the same time, I would encourage work sharing, as the Final Report of the International Competition Policy Advisory Committee also encourages. This implies co-ordination of staff to the extent that resources and confidentiality rules allow. It worked very well in the PrimeStar and Nintendo cases,9 as well as in other cases, and has brought the States and the federal government closer. It also appears to be working, notwithstanding press reports—and one does not always believe the press—in the Microsoft case.l0 We hear many objections on breaches of confidentiality in cases of enforcement co-operation. I think this is an area that cries out for uniformity, not only within the jurisdictions of the European Community but also in the area of transatlantic co-operation, about how we could ensure intraCommunity co-operation and cohesion, as well as in our increasingly global market-place. In the US, the federal agencies cannot withdraw cases from courts, but they can use the amicus curiae briefs which have proved very effective in the insurance anti-trust litigation. It is also possible to use procedures such as the consolidation of related cases within a single jurisdiction, so that two cases on the same matter are not going into the same district court before different judges. Under Section 1407 of the United States Judicial Code we have the opportunity to consolidate cases across districts, where at least for procedural purposes (and often more than that), it is useful to bring about uniform results where there are multiple cases and multiple jurisdictions. I do not know if this could work in the EU, but I think it has been very helpful to achieve uniformity in the US. I think also clarity is important, and that clarity and uniformity can be achieved without a loss offlexibility.Sometimes flexibility is difficult to achieve even within a single jurisdiction. The ABA, the ICC and the UNICE have all cried out for uniformity, clarity and transparency in conjunction with the modernisation proposal; and at the International Competition Policy Advisory Committee we certainly heard a great deal from the business community asking for clarity and coherence.

9

United States v Primestar Partners, L.P.,\ 994-1 Trade Cas. (CCH) (S.D.N. Y. 1994) and New York v Primestar Partners, L.P. 1993-2 Trade Cas. (S.D.N. Y. 1993). 10 United States v Microsoft Corp., No-98-1231 (D.D.C.filedMay 18, 1998), see also 84 F. Supp. 2d 9 (D.D.C. 1999) for findings of fact, and 87 F. Supp. 2d 3 (D.D.C. 2000) for conclusions of law.

2 - Coherence

289

• MARIO SIRAGUSA—Let me try to be provocative: I would say that the real danger of the reform proposal is not lack of legal certainty or lack of coherence in enforcement, but the marginalisation of the Commission as the engine of enforcement and EC competition policy-maker. I will explain my thoughts with three questions, which I will then try to answer. First, what is the role of the Commission's monopoly to apply 81 (3) today? Second, will the elimination of this monopoly induce national competition authorities or judges to apply EC law? And third, what will be the legal status of the Commission's decisions in the new enforcement system? On the first question—related to the role of the Commission's monopoly to apply 81 (3)—in Europe today, the large majority of cases are determined under national law and by national competition authorities. The third paragraph of Article 81 is the tool by which the Commission can reserve to itself certain areas of implementation of EC competition policy, such as block exemption regulations, which are very effective means to regulate matters at the European level. National competition authorities are not impeded from acting in areas covered by block exemption regulations (such as the area of vertical restraints). Block exemptions are also a tool of decentralisation, because once they have been issued, national judges and national competition authorities can of course apply them. The second use of the monopoly relates to the issuing of individual exemptions in those few cases of agreements of EC-wide relevance, or, as I said in my paper, in the cases of investments that are very important at the European level. In such (very few) cases, the Commission's protection is granted through individual exemptions. That is why the current number of notifications in Brussels is only a couple of hundred, which in a market of more than three hundred million people does not seem to me to be a big number. The possibility of issuing individual exemptions is a powerful tool because such decisions are binding on the national competition authorities and on national judges. As to the second question—whether the elimination of this monopoly will induce national bodies to apply EC law—I doubt it. Why should they? First of all, the Member States adopted legislation inspired from Articles 81 and 82. I would really be surprised if the British, who have now finally adopted a modern competition law, would suddenly abandon the application of their national law and start applying Articles 81 and 82 following the reform, particularly when (as some people observed this morning) the trend towards limiting the scope of 81 (1) is already strongly manifest. The Italian Banks case" is an example of the re-thinking of the notion of effect on trade between the Member States even before further limiting the application of EC law, and of leaving more space to national law. I doubt that national competition authorities will be induced to apply EC law; if they chose to apply Article 81, they would possibly have to tolerate interference from the Commission and need to 1

' Joined Cases C-215/96 and C-216/96 Carlo Bagnasco and Others v Banco Popolare diNovarrasx., Cassa di Risparmio di Genova e Imperia SpA (Carige) [1999] ECR 1-135.

290

Panel Discussion

co-ordinate with other national competition authorities. I therefore think they will be inclined to continue to apply mainly national laws, and will apply EC law only in rare cases. As to the third question—on the status of the Commission's decisions in the legal exemption system—the strength of exemption decisions is that they have binding effects on national competition authorities and courts. The legal status of the new decisions is doubtful. Fortunately, the Commission is now trying to give some strength to these decisions, and it talks about 'binding effect', though a binding effect that is based on the general principles of the Treaty rather than as a clear consequence of the existing exemption monopoly. So, the binding effect of exemption decisions will be questionable. The role of the Commission as competition policy-maker might therefore be largely reduced because the effects of its decisions will be doubtful, whereas national competition authorities will continue to apply mainly national laws and I do not see an incentive for this to change. What are then the solutions? One solution could be to adopt a different enforcement system, but that is not the topic of today; today we are supposed to try to improve the Commission's proposal. On this approach, I think that we should first try to bring back the binding effects of the Commission's decisions, and there are many people who have suggested different ways for doing that. Many people here, including Mr Paulis, said that we must have a strong rule about the primacy of EC law. Yet the strongest rule of primacy is the Commission's monopoly on the application of the third paragraph. That is the real and strongest rule of primacy, and if we eliminate it the primacy of EC law becomes much more questionable than it was in the old regime. I understand the Commission's positive decisions will continue to be issued. This is surely a positive development. Yet what is the system through which these positive decisions can be obtained? That is all open for debate. It could be following notification, but there is a strong reaction against re-instituting a notification process. There are even discussions aboutfindingways to give binding effects to decisions by the national competition authorities. Therefore, there is preoccupation not only about the effects of the Commission's decisions (where, through the primacy of the Treaty provisions, one can find some binding effect), but also about the effects of decisions by national competition authorities (which seem to be much weaker unless their binding effects are provided for under the implementing Regulation). Another solution, and this has also been said today by many people, would be to create clear rules about the allocation of cases between the Commission and the national Authorities. To conclude, I believe the two main areas in which the reform proposal could be improved are (1) the binding effects of the Commission's decisions and (2) clear rules of co-ordination between the Commission and the national competition authorities. • JAMES VENIT—I wonder if I could say much, after Mario Siragusa has already said basically what I wanted to say. Let me trace back, like the captain

2 - Coherence

291

in the trench in World War I, as the shells are flying over his head and he is trying to figure out how it was that he got where he was, and see how we got to where we are now. Since the EUI workshop four years ago, a consensus has emerged: 81 (1) was applied too broadly; the split between 81 (1) and (3) was a bad thing; the consequences of the automatic application of Article 81 (2) in litigation were not acceptable. A lot of the drawbacks of the existent enforcement system were traced to the split between 81 (1) and (3) and the notification system. We diagnosed the problem and accepted that there was a need for reform. To my mind, there were two implications arising from the acknowledgement of the need for reform. One was that, by applying economic analysis more, you could get rid of the artificial and formalistic split in Article 81. Legal certainty would not necessarily be improved, but there would be at least an acceptable kind of legal uncertainty that had to do with the subject matter: economic law. The second implication that I envisaged was that there would probably be an increase of private enforcement. That did not disturb me particularly, because one assumed it would be private enforcement of Article 81 as a whole, and not just of Article 81 (1). I felt comfortable that the judges could handle that, and there would be a system of judicial review that would work. Therefore, I was frankly taken aback by the boldness of the Commission's reform proposal when it finally hit the light of day, because what I describe as a 'system of cohabitation' struck me as going further than needed and giving rise to some potential major problems. I heard Jim Rill talking about the very uneasy relationship between the federal government and the States in the area of antitrust enforcement in the US, where political consensus is much firmer in terms of cohesiveness of the nation. I am not sure whether we can accept in Europe the degree of friction that can be absorbed in the US system. This is one area where I would say 'don't follow the US model: it might not work'. In analysing the reform proposal, I think one has to distinguish between the role of national courts (which do not have policy agendas and decide cases based on the facts and on an individual basis) and the role of national competition authorities (which do have policy agendas). That is where I think friction comes in. What I am suggesting—and what I seem to hear from a lot of the speakers today—is that there is a possibility that the Commission might have gone too far. The issue now is whether can we pull some of that back and get things in order. A number of suggestions have been made today which, on reflection, I regard as positive steps in that direction. First, there is recognition that there are certain types of cases that you might not want to have dealt with at the national level. I was involved in the airline alliance cases, and I can tell you that I do not see any positive value in having a 'cohabitation' kind of enforcement system in such cases. No authority has primacy under such a system; there are only 'equal authorities'. In the airline cases, the Commission was able to establish the duty of loyal co-operation applying to the Member States' authorities in order to, essentially, seize back jurisdiction. But for those of us who went through those

292

Panel Discussion

cases, it is clear that a cohabitation system creates a layer of potential procedural and substantive diversion that is not positive, and that I would see this as a tremendous inefficiency of the regulatory system. One wants to avoid that, but how? A number of suggestions were made in this sense. First—and Emil Paulis touched on this before—the primacy of Community law must be established. Second, the scope of Article 81 (1) should be narrowed. Adopting positive decisions is another policy suggestion that I consider very important. And Judge Bellamy's suggestion that jurisdictional rules must be very clear is also something that needs to be reflected upon. But what concerns me is that, when you look at some recent developments like the Italian Banks decision of the European Court of Justice, you start wondering about the scope for the application of Community law by national authorities. If purely national agreements fall outside the scope of EC law, then its scope of application is reduced to agreements that are multi-jurisdictional, and you might start to have a very serious problem when moving away from a centralised model. I therefore think that, at this stage, the Commission has taken a very bold approach; perhaps too bold. It is time to go back and think very carefully about the potential negative consequences of an overly bold approach in terms of decentralising EC competition law enforcement. I do not think that our system is robust enough to tolerate high levels of dissonance in cases that involve multiple jurisdictions, or parties from multiple jurisdictions, and where different national policy agenda items can start to dictate. The Commission has to maintain the power to set the competition policy agenda for the Community. In fact, if you go back to the original purpose behind the reform, it was to free the Commission from what was regarded as unnecessary and unimportant work deriving from notifications (whether that justification was warranted or not is another issue). To get to the point of imposing your agenda as an enforcing authority by giving up control over that agenda is a little bit questionable, and we have to really ask ourselves whether the reform is moving in a completely coherent way or not. • KAREL VAN MIERT—It is striking that a lot of people, particularly the practitioners, care for the status of the Commission and accuse it of boldness. This is something different to what I have been used to, and I still have to learn to live with it; but it seems to be a good sign. Having said this, there are indeed some important issues to be clarified within the reform proposal. No one will contest that, but I think that some of the examples given here are not entirely appropriate. Talking about the airline business, the origin of the problem lies with the fact that the Commission has not been given jurisdiction over agreements between companies from inside and outside the European Union. Let me remind you that, in this case, the airlines concerned were not even prepared to 'notify' of their agreement. We obliged them to do so. Indeed, we had to use the old Article 89 [now 85] just to get involved, but the companies could not get away with it.

2 - Coherence

293

No one can get away with it, because at the end of the day that would imply facing a situation of illegality within the European Union. Therefore, if the Commission has an idea of what the problem is. There were and there are real problems, and the Commission can act in such circumstances. This seems to contradict the fears just expressed by Mario Siragusa. It is all about a determined Commission, one that properly analyses what the real competition issues are and attacks them, even if it has not been given the right instruments to do so, as it was the case in the airline example. There we had to deal with US Authorities, which made life difficult for the Commission, because in this case they had a completely different approach from ours (and not exactly a competition approach, for that matter). Their approach favoured American interests, and that was all. In spite of everything, the Commission could block agreements that were considered detrimental to competition, whether right or wrong. This is my main point: a determined Commission can do its job, even without the necessary instruments, even if the companies concerned are not prepared to 'notify' of their case in the first instance, even if national authorities take a different view. It might take longer. This case did take a long time, and our intervention was not very efficient because of our lack of jurisdiction, but in the end the Commission got its way. Second, under the existing system, legal practitioners can drag a case on for years and years and get away with it. They 'notify' of an agreement, and when the Commission is about to say 'no', they 'notify' of a slightly different agreement, and the process starts all over again. This tactic can mean it takes years and years before a final decision is reached. This is to show you that some of the remarks that have been made here should be a little bit more nuanced. It all comes down to having an independent central competition authority, and developing strong co-operation with the national competition authorities. Let me give you the example of the CocaCola Company, which tried to by-pass the Commission and play one national competition authority against the other. All we needed to do is sit down together and speak about it, and you know the result. This is another example of how the Commission can handle this kind of situation once it knows what to do and if it is determined to do it. • CHRISTOPHER BELLAMY—I agree with you one hundred percent. What we are saying is that in these kinds of situations it should be the Commission, and only the Commission, to have jurisdiction. Once you involve the national competition authorities in some kind of messy cohabitation, either it all becomes too complicated, or procedural disputes slow it all down. Therefore, keep the jurisdiction over cases with a Community dimension exclusively at the Commission. • KAREL VAN MIERT—I fully agree with you: it would be much better if it were up to the Commission to handle cases such as the ones I mentioned before. It

294

Panel Discussion

could be done in a much speedier way, and without the complications that, at the end of the day, are detrimental to the interest of the companies concerned. • JAMES RILL—I think we are all in agreement on that. The only point I am trying to make is: having lived through that situation, would one want to consciously recreate it? I think the answer is 'no'. • PETROS MAVROIDIS—I have basically one comment on consistency and cooperation in enforcement. Let me start with something that Frederic Jenny said before, which I think is very true, namely that consistency in itself is not a value unless we are consistently right. You see, this is classical lawyer talk. We talk about something without knowing what we are talking about, because how do you define consistency in an ever increasing rule of reason world? We should be consistent about what? What are the consistency criteria? For constitutional reasons, we might not have the same criteria throughout Europe. Legally speaking, consistency would hardly be a measurable matter in the field of competition law. But let us assume that we somehow overcome this difficulty and consistency, as Frederic Jenny said, would simply signify correct decisions—which means one would like informed decisions by the national competition authorities, which also means that one would like good people to work there. I can see the intellectual merit of this, but I tend to have a slightly different view. To my mind, this is not the crucial issue, for two reasons. First, you can easily address this kind of 'incompetence externality'. You can certainly find five to ten good people in any Member State. Second, uninformed decisions do not have a disintegrating effect. Incompetence is national interestblind: you can be incompetent and go against your national interest. What is the disintegrating effect of this? That is what I argued in our paper, and I saw nothing in the White Paper to address this issue from the legal point of view. To the contrary, the White Paper seems to offer national competition authorities precisely the possibility of basing their decisions on national interests. Why do I insist on this point? Recent practice, like the Boeing/McDonnell Douglas merger case,12 shows clearly that when stakes are high, co-operative outcomes are forgotten, and everybody goes his or her own way. On the one hand, game theory would apply. On the other hand, we do not have the legal instruments to ensure that national competition authorities move towards cooperation. To my mind, this should be the priority. Unless we 'immunise' people who work in national competition authorities, we will never really create the network of competition authorities envisaged by the White Paper. • DAN GOYDER—What I am about to say will echo what has just been said. I was just thinking about the meaning of coherence. In our context, coherence 12

Commission Decision M 877 Boeing/McDonnell Douglas, OJ L 336/16 of 8.12.1997.

2 - Coherence

295

seems to have two meanings. The first is that the rules are logically consistent. We talked a lot about that. But equally, or even more importantly, is the second meaning of coherence; namely, that there is confidence among those affected by the rules that their application will actually produce both a fair and a logical result. We have to be concerned, of course, not only with the feelings of Member States and their competition authorities, but also with those of business and their advisors, with the people who are actually the subjects of individual cases. At various moments this afternoon I have asked myself whether there is such coherence. Has such a point been considered? Very briefly, one or two points that have not been mentioned yet. The first, and one that puzzles me, relates to the fact that for ten years now we had a very effective system of notification, which nobody complained about, in the field of merger control. In fact, it worked extremely well until very recently, and it is only because the inputs are now starting to exceed the outputs of that particular system that people are starting to get worried, and articles published in the Financial Times show that we have arrived at a point at which a resource-adjustment has to occur. This shows that in the case of transactions where you basically want a 'yes' or a 'no', or a 'yes' plus conditions, the notification system can be very good. I just wonder why we have not talked about it at all, and why the White Paper mentions it only indirectly. Why don't we look at the lessons to be drawn from the application of the Merger Regulation? If legal certainty is a problem—and it is a problem for major investments—why don't we use what we have learned from the Merger Regulation? We have learned what works about it and what does not. For instance, one of the benefits from the modernisation should be that a wider range of joint venture cases should be put through. After all, joint venture cases are a 'grey area' between Article 81 and the Merger Regulation in most cases. It seems to me that, without great adjustment, all structural joint ventures above a certain threshold could be considered as falling under the Merger Regulation, which then would relieve the rest of the Commission and allow it to deal with other priorities. My second point is the following: as has come across very clearly from the last speaker, giving extra responsibility to national competition authorities does require that those authorities operate within a very clearly defined structure. It is very important that the Regulation—or Regulations—that are going to embody the modernisation lay down the minimum requirements for a national competition authority to be given these new responsibilities. It might be a bad example, but if you think about EMU—and perhaps the British are not the right people to think about it—no country could enter the monetary union unless its economy met certain requirements. There were no exceptions. The tests were applied, I assume objectively, by the experts of the Commission. Why don't we say that a national competition authority has to meet certain minimum institutional standards that I think are obvious? That includes the relationship between national competition authorities and national courts. However, as we will talk tomorrow about how to help courts and judges, I

296

Panel Discussion

would not like to discuss now this point in detail. There is a very important integration effort required here. We cannot treat national courts and national competition authorities, which are given very different functions, as totally separate entities. If the system is to work, they have to be able to help each other. Finally, it does seem to me that if the Commission is to take the lead in setting standards and continue to be the prime mover of competition policy within the Community, it should set standards for itself that are at least as high as those set for everyone else. I am not making any personal point here, except that everybody who has acted in a competition cases before the Commission knows that there are complaints—some justified by the facts, and others not—that sometimes the standards of administration are not as high as might be required. In particular, being both prosecutor and judge effectively imposes impossible problems upon the Commission officials. Therefore, in setting the rules on the conduct of the hearings, for example, the Commission should be very self-critical and set very high standards for itself. Of course, through its decision, the Court of First Instance, to some degree acts as a 'gate-keeper', but perhaps such decisions should prompt the Commission to adopt even stricter internal rules. • DAVID EDWARD—I will start with two comments on what was said this morning. First, Professor Immenga quoted Professor Mestmacker saying that national courts should not consider issues other than competition ones. Here I would have to agree with Professor Tesauro. A court faced with the interpretation of an article in the Treaty cannot divorce that article from the rest of the Treaty. The Treaty contains a number of other considerations, and courts cannot ignore them simply because they are dealing with a competition case. Second, Mr Burrichter asked why the Community courts do not take a wider view of their competence as administrative courts of review. The short answer is that, in the majority of Member States, the judicial review of administrative decisions by an administrative court is much less wide than in Germany. The Community judicial review system is broadly based on the model of the French Conseil d'Etat. If you want to change that, it can be changed, but one can't just say: 'change!' In relation to that, I would like to support very much what Mr Kist has said this afternoon about the need for preparation, clarity and definition. Laurence Idot raised the issue of the Brussels Convention. The Brussels Convention is an inadequate tool for the allocation of competencies in this field between the courts of different Member States. Without adequately preparing rules on the allocation of competencies, I can foresee a lot of time being spent on preliminary references about the Brussels Convention before the national courts even get down to discussing the competition issues. This would multiply the problems. It is clear, I think, that the Brussels Convention is not sufficient on its own.

2 - Coherence

297

One comment on the burden of proof. The burden of proof is of importance in civil law proceedings from two perspectives: it is the basis on which civil law judges operate (as they have to look at cases from the point of view of 'who has to prove what'), and it also important from the point of view of the parties (because they have to know what they are expected to produce in terms of proof). The issue of the burden of proof is of particular importance in jurisdictions where civil litigation is largely conducted in writing: the parties must know what proofs they must produce in writing, and then the judge must know what to make of it. A classic example of this is found in the jurisprudence of the European Court of Justice, in the French electricity and gas monopoly case.131 was Rapporteur in that case, and the eventual solution was to define the scope of the Commission's initial case,14 and then to decide how far France and other Member States had to go in replying to that case.15 Defining the burden of proof is also very important if you go into the question of 'the rule and the exception', as opposed to an overall economic assessment. In other words, you have to decide in advance whether 'there is a prohibition rule, and the defender of the contested agreement has to demonstrate that the agreement does not fall within its scope', or whether 'the judge will make a general overall assessment of the merits of this agreement'. You must decide beforehand which of the two it is going to be; and please, decide beforehand, do not ask us the European courtsfiveyears later—and once there are already a number of conflicting national decisions—to tell you what the system is. In that context, it is true, as Professor Tesauro said, that the courts are used to balancing Article 28 and Article 30 and the exigences imperatives. But, as Mario Siragusa has said, they do so when reviewing State measures with regard to the reasons that the States have stated or are required to justify. This is different from reviewing the Treaty provisions against private justifications for potentially anti-competitive agreements. That is a different operation, although there is an analogy with the former. • BARRY HAWK—I would just like to revisit a point made by Mario Siragusa earlier. I would raise that argument again, and see if everyone agrees with it. If I understood well, Mario Siragusa's concern was that, in giving up the 81 (3) monopoly, the Commission relinquished its most 'powerful tool' to the Member States. It seems to me that the area of competition policy in which the Commission was perhaps most successful over the last ten years is that of liberalisation—telecoms, transport, etc.—where it had to fight the Member States (though it seems that in the area of liberalisation, the 'tool 'used by the Commission was not 81 (3) but Article 86—the old Article 90—and the Merger Regulation). The Merger Regulation was used because it required that transactions had to be 'notified' and approved ex ante. That gave the Commission the 13 14 15

Case C-159/94 Commission v France [1997] ECR 1-5815. Points 24 and 25. Points 90 et seq.

298

Panel Discussion

possibility of saying to the Member States' governments: 'well, if you want this transaction to be approved, you will have to liberalise first, or we are not going to approve it'. Therefore, I do not see the Article 81 (3) monopoly as the most powerful Commission tool. I rather see this tool in the old Article 90 and the Merger Regulation. • ULRICH IMMENGA—I feel I should raise three objections to the general conditions for ensuring coherence under the new enforcement system formulated by Mr Paulis. One of the conditions he mentioned is securing the primacy of Community law. I think that in the new enforcement system the Community rules will autonomously achieve the highest degree of primacy because, when applying Article 81 as a whole, you can no longer distinguish between paragraphs 1 and 3. Under the existing enforcement system, paragraph 3 was considered to embody a sort of primacy vis-a-vis the application of national law. Under the new enforcement system, the distinction between paragraphs 1 and 3 will no longer be possible. According to my understanding, this means that all the decisions taken by the Commission on the basis of Article 81 will in fact have primacy vis-a-vis national law. As to the second condition—namely, the autonomous power of enforcement of the Commission—I think this does not really comply with the principles of decentralisation and subsidiarity. To the contrary, I think that this will promote, at the national level, a tendency to apply national law. Whether this is •desirable or not might be subject to discussion, but if one is interested in extending the reach of EC competition law, I do not think it is useful to set incentives to apply the national law. Even if the primacy of EC law is extended, there will still remain areas where EC law and national law can be applied in parallel. My last comment regards the condition of a more economic approach in the application of Article 81 (1). Here the 'catchword' was market power. I see some risks in such an approach. If one looks at the US antitrust law, there are at least some borderline per se rules whose application does not require that decision-makers look into the economic effects of agreements. The same should apply to Article 81 (1). Otherwise, the danger is that you look at efficiency considerations only when you look at economic effects, and then you return to the Chicago School approach, which I think we have already overcome in Europe. • DAMIEN NEVEN—I would like to make two brief points. One is to reinforce what Professor Goyder said with respect to minimum institutional standards for national competition authorities. I just wanted to add that this is actually done in other areas, for instance in the area of telecommunications where the Commission imposes to the Member States independence standards for the national regulatory agencies. Actually, in this particular case the independence standards imposed by the Commission might even seem ironic but, without

2 - Coherence

299

going into this discussion, there is no reason why this could not be the same with respect to antitrust. My second comment is with respect to the issue of being 'economic-friendly'. I am getting a little tired of this discussion, that we have been listening to for a number of years, and I do not see much in the way of results. Here I would also like to echo one of the points that was made earlier by Simon Bishop. Some might say, and I am sure that the Commission believes, that the new treatment of vertical restraints is more economic-friendly. I do not think it is so; even though the Commission has understood that market power matters, it has used the market power criterion in a completely wrong way. What economists have been saying for some time is that if there is no market power there is no issue with respect to vertical restraints. So, if there is no market power, 81 (1) does not apply. Instead, the current approach is to say: 'well, if you fall below the market share threshold, you may get the benefit of 81 (3)'. Nonsense. If you do not reach the market share threshold, you do not fall under 81 (1), and there is no competition issue to start with. All of this is to say that I am in favour of a more economic-friendly approach in enforcement. With respect to this particular issue, I also think that there might be a bit of confusion in the idea that being more economic-friendly will introduce more uncertainty. In the end, I think the concept of legal uncertainty is much too narrow. What companies do care about is business uncertainty, and I could very well think that the current environment induces a lot of business uncertainty simply because the particular agreements that firms want to enter into, or the particular joint ventures that they want to create, have to be tailored to fit into the straight jacket of Commission decisions or of block exemptions. When companies face a particular deal, they do not know whether fitting it into the straight jacket of the Commission will actually deliver the business benefits they are after. At the end of day, what companies are worried about is uncertainty about business benefits, and not uncertainty about particular legal forms. Therefore, I believe that the issue of the trade-off between economic analysis and uncertainty has been somewhat exaggerated. • ELEANOR FOX—Two different points: first, and picking up on the notion of liberalisation introduced by Barry Hawk, I believe that the removal of the notification system is very much a part of the ongoing liberalisation process. The notification and prior administrative approval system frustrates the effects of liberalisation. The fact that the notification system was adopted in the area of merger control really stands as an exception. In the case of mergers, we probably do need notifications, because the parties are about to integrate and the unscrambling of their assets would be very difficult. In other cases, we probably do not. I would point out, however, that even in the case of mergers the situation has became so complicated lately, particularly with respect to international merger notifications, that in order to alleviate the heavy burden of multiple merger notifications around the world we probably need some

300

Panel Discussion

international co-ordination. Diane Wood would actually have a very good proposal in this respect. My second point refers to the burden of proof. I agree that it is important to specify the burden of proof. The structure of Article 81 causes the usual economic analysis to fall through a crack. This is because Article 81 (1) is so wide that it catches restrictions on competition that are much less than anti-competitive. Once an agreement is caught by 81 (1), the burden to prove that the benefits or pro-competitive aspects of the agreement outweigh its restrictive effects falls on the defendants, whereas the Commission (as plaintiff) does not have the burden to prove that the restriction is significant enough to cause the transaction to be anti-competitive. That is something that should necessarily be proved. This is a point that Damien Neven made in his intervention, and Simon Bishop made it also. I think it will be very useful if the Commission revisits those burdens of proof, and puts a burden on the Commission to prove first that the restraint is of an anti-competitive dimension, rather than just a restriction or distortion of competition. • GIUSEPPE TESAURO—The experience of the Brussels Convention can help the Commission to make better rules on the allocation of cases. It can perhaps even provide for an arbitration clause in the hands of the Commission; if there are problems about the interpretation of the allocation criteria. Why not provide an arbitration clause for the Commission? Second point: abandoning the Commission's monopoly over the application of 81 (3) does not mean leaving the application of 81 (3) to the courts only. The Commission has to guard very strongly its power to maintain some or all of the cases it wants, such as when a case is difficult or new. Therefore we can have a solution for differences between the national competition authorities and the temptation not to use European law. Third point: Bagnasco, or the Italian Banks case. I do not know if one judgment only can make case law. It is one judgment, only one. I have been reading the judgments of the European Court of Justice for many years now and I always tried to wait until there are three judgments on the same matter, so as to make a 'bouquet offlowers',otherwise there can be surprises. Take for instance the Emmott case,16 it was just one judgment. I tried to say then 'Wait, my friends, wait for a month, two months, a year'. The result was that Emmott was just a judgment, but the case law was quite different. So, Bagnasco is one case, and there is probably time to re-think about it. I do not think the Court consciously made with it to affect daily life, and I am sure that national competition authorities would not like to apply EU law that is not well founded.

16

Case C-208/90 Emmott v Minister for Social Welfare [1991] ECR 1-4269.

2 - Coherence

301

• IAN FORRESTER—I would like to echo and sympathise with the remarks of Mr Oprescu from Romania. There are, I believe, twenty-five or thirty countries under an obligation to implement EC competition rules, ranging from Morocco in the West to Russia in the East. Most of these countries have reacted to their duty by simply introducing versions of Articles 81 and 82 into their national legislation. The danger is that the competition agencies in those countries, which still lack confidence and enforcement experience, are likely to apply in this century the law from the last century, which we all agree we should try to get away from. I think, therefore, that one of the most important tasks of the Commission over the next years is to give us guidance, be it via decisions or via guidelines, on where Article 81 (1) falls and what the criteria are for the application of Article 81 (3). Otherwise we risk these countries discharging themselves of their competition law obligations in a bureaucratic, black letter, ticking boxes approach: 'I'm doing fine because I have a good Merger Regulation'. We all know that is not enough. Second observation: there is only one competition authority in the European Union to have a multi-country enforcement duty, and that is the European Commission. The role of national competition authorities is only national. This not an insoluble problem, but if we look at Belgium, for example, a Report from the competition authority has not been published there since 1995. Why is that? It because they have no photocopiers, no officials, and since recently no Conseil de la Concurrence either, because its members have all resigned in disgust at the lack of office facilities to work with. The Belgian phenomenon is to be cured by Belgium. It is not the Commission's fault, but it will be if the Commission says: 'competition enforcement in Belgium is up to the Belgian authorities'. The Commission must take reality, or the weaknesses of the enforcement process at national level, into account. Finally, echoing the elegant words of Mr Kist, I would say that a reform whose concept and goals are splendid and to be commended, is too important to be damaged by imperfect preparation. I think that what a number of practitioners and commentators around this table are actually saying is 'piano, piano.' Or, using different words in their respective languages, they are saying: 'caution, you are going in the right direction, but the passengers on the train, which encouraged the train to get moving in the first place, are now knocking at the doors, trying to climb out and slow the train down'. • JOCHEN BURRICHTER—First of all, I wanted to clarify my remark this morning when I referred to the administrative nature of the exemption procedure, which is in fact the reason why the EC granted the Commission discretion in applying Article 81 (3). I did not ask for an amendment of the EC judicial review system: I am fully aware of the limited scope for judicial control of administrative decisions. What I wanted to say is only that this discretion is not a logical consequence of the notions used by Article 81 (3). The discretion is a

302

Panel Discussion

result of the administrative process. Such discretion would not be recognised by the courts of all Member States. My second remark concerns a question posed to Mario Monti this morning, and which was also repeated during the debate: why should the Member States, or the competition authorities of the Member States, have the incentive to apply EC law after the reform has been introduced? I think they will have such incentive because they will be forced by the business community to apply EC law. Take, for example, a clear-cut case in which an agreement affects trade among the Member States. The business community will argue that the agreement no longer has to be 'notified' to the national competition authorities for approval, even if national legislation requires notification. Undertakings would instead just invoke the exemption provision under Article 81 (3). The national competition authorities would then be forced to deal with Article 81 (3) if they wanted to question the legality of the agreement. If national courts can apply Article 81 (3) and the national competition authorities are allowed to, and requested to submit comments during civil law proceedings, the national authorities will have to deal with Article 81 (3). • DIANE WOOD—A couple of the participants have referred to Mario Siragusa's comment that the Commission will lose power. In terms of coherence it probably will lose power if there is a greater plurality of enforcers. I deliberately use that word—enforcers (as opposed to 'national authorities')— because we are talking about a couple of different cases. We're talking about cases where national competition authorities are perhaps using procedures available to them to enforce EU law, but we are also talking about the way in which the national courts will apply 81 (3) in ordinary cases coming before them, where private parties might simply litigate about a contract or arrangement, and where the national competition authorities might not be involved. It might just be the last case, and these changes will make it difficult for the national courts to apply 81 (3). The national courts will bear a heavy responsibility for developing the coherence and consistency of Community law through whatever mechanisms are available, whether by making greater use of the possibility of requesting preliminary rulings from the European Court of Justice based on Article 234, or by requesting some form of intervention by the Commission. I would make two other observations about the Article 81 (1) and (3) issue. One is that I really quite disagree with what Barry Hawk said this morning, namely 'pull everything into 81 (1) and let 81 (3) be limited to non-competitive criteria'. The truth is that competition law does not need to take care of every single other area of law (such as environmental rules, labour rules, and industrial and development rules). It is necessary to determine which law takes precedence. It seems to me quite clear that Article 81 (3) is a competition-based rule. In fact, I was interested to hear Mr Immenga saying that you have overcome the Chicago School approach in Europe. I would have thought that in fact

2 - Coherence

303

the consensus on the fact that competition is a body of law about efficiency and increasing production actually means that the Chicago School took the field, but perhaps not, and I'll be glad to be corrected if need be. What Eleanor Fox said about the burden of proof might be the most important question left open, because if you thought that the burden of proof shifted when it begins to be an Article 81 (3) issue, that would be a very serious matter indeed. • MARIO SIRAGUSA—I do not see how one could induce the national competition authorities to apply EU law before their national law. There is the traditional principle of the double barrier. The reform will eliminate the prior authorisation process, so that positive Commission decisions will be nothing more than declaratory under Article 81(1) and (3) combined. I think that if a national authority says 'This is a declaration made by the Commission under 81 (3), and I will apply my national law', there is nothing to oblige the national authority to apply Community law. This brings us back to what I think is a very important issue: what is the role of the Commission's decisions? Can we rely on the basic principle of the primacy of EC law under the Treaty, or do we need to give those decisions specific force under the new Regulation? My second point concerns what Barry Hawk said about the success of the Commission under the old Article 90. The very reason for that success was that the Commission had the power to issue directives, which were obligatory vis-avis the Member States, and that the Member States had to comply with. I really wonder what would have happened in the liberalisation process if the Commission had not had that power. Remember that power was often contested by the Member States, which brought the Commission to the European Court of Justice on this. That is a very good example in the sense that, unless it has the power to overcome the opposition of the Member States, the Commission is going to be marginalised. The only modality the Commission has to impose its views under the present system is the third paragraph of Article 81. This is, in a sense, akin to the compulsory role of the directive under Article 90: with the third paragraph of Article 81 you can oblige national competition authorities and Member States to follow your rules. And that is what you are giving up. • JAMES VENIT—I think the issue here—and this is really why I am very anxious to hear what Alex Schaub and Emil Paulis have to say—is that you can try to move away from the problems of a system of excessive control, towards a system that is going to be less centralised; but, at the same time, you do not want to throw the baby out with the bath water. What are the legal mechanisms to do it? Is retention of the Commission's monopoly over Article 81 (3) the only way to do that? I do not believe that is necessarily the case. But the issue is: what other means are at the disposal of the Commission, bearing in mind that that the adoption of new legislation might require the co-operation of the Member States to provide those

304

Panel Discussion

legislative means and to adopt a system that will eliminate the defects that we had before without plunging us into a fairly chaotic type of universe. Again, I want to make it clear that the views that I have expressed should not be interpreted as hostile to the role of national competition authorities. The national authorities have played a very important role in the Member States, and they continue to have a vital and important role to play, also in consultations with the Commission. What is terrifying to those of us in private legal practice, is the thought that our clients will become a prize to be fought over by various regulatory bodies seeking to exercise their jurisdiction. That is not an acceptable risk to impose on the private sector. And that is the spirit in which these issues need to be addressed. • BARRY HAWK—Since I do not think I have ever disagreed with Judge Wood on anything, I am now going to try to clarify what I just said this morning and also respond to Professor Immenga. If you have the full economic analysis brought under 81 (1), that does not mean taking the Chicago School approach or anything else. You can have both per se rules and whatever you want to call non-per se rules. But whatever you are doing, do it under 81 (1). I agree that, ideally, if you want to trump competition for environmental reasons, or for jobs, you should do that under another statute and not under Article 81 (3). But I did not want to be too radical: in the end we do have Article 81 (3), even though everyone seems to agree now that this provision was a giant mistake back in 1957. If you put all of the competition analysis under Article 81 (1), you still have to do something under Article 81 (3). If you decide to include environmental, employment or industrial policy considerations in competition law enforcement, you should do that only under Article 81 (3). And if you are concerned that Member States—maybe not the national competition authorities, but another part of the national governments—will disagree with, say, liberalisation, or if you are concerned that you will give national governments the power to grant exemptions based on national industrial policy considerations (and I frankly think that is what all this is about, though nobody wants to say it openly), then you should put the competition analysis under Article 81 (1) and keep industrial policy under 81 (3). This eliminates the concern about national authorities or government by retaining the Commission's exclusive jurisdiction over Article 81 (3). • JOCHEN BURRICHTER—Getting back to what Mario Siragusa just said, I understood that the direct applicability of Article 81 (3) would be the consequence of a new Regulation, replacing Regulation 17/62. For me it is crystal clear that a positive Commission decision is an act of the Commission having primacy over national law, so that any attempt by a national authority to apply its own rules on notifications or whatever would contradict the principle of the primacy of EC law.

2 - Coherence

305

• DAVID EDWARD—I would simply like to follow up what Giuseppe Tesauro said about the interpretation of the European Court of Justice's judgments. Remember that, in order to determine who is to decide a case, the Court uses the following broad principles: (a) if the case raises a major issue of principle, it will be brought before the plenary; (b) if it is a simple question of technical interpretation, like the interpretation of customs rules or a simple question of applying existing jurisprudence, it goes to a chamber of three: and (c) inbetween there is the chamber of five, whose basic assignment is that it may develop existing jurisprudence, but not create new jurisprudence. So, in addition to the point Giuseppe Tesauro made, which was to look at the decisions on a specific issue over a period of time, also look at how many judges were involved, because the number of judges involved indicates a priori whether the Court considered that it was necessary to review or change the jurisprudence. It might appear that a chamber offivejudges is saying something new, but that is a reason for waiting to see whether this novelty is confirmed by more decisions in chambers of five, or instead (as has happened in many cases) because of this 'something' said by the chamber offivejudges, the next case is sent back or kept before the plenary to decide whether that was the right way to go or not. So, do use some degree of insight in reading the cases. • CLAUS-DIETER EHLERMANN—First, I think that delegation and decentralisation requires us to clarify what is delegated, to whom, and with what content. This really seems to me to be a powerful argument for what Mr Kist forcefully and convincingly required; that is, clarity of the reform in several respects. One of them, is 'who is doing what' in terms of the distribution of responsibilities under the new enforcement system. All the written contributions I have read so far suggest that this distribution should beflexible.Yes, there is a lot to be said forflexibility,but there is also a lot to be said for transparency, which is one of the requirements for democracy. Another aspect of the reform that absolutely needs to be clarified is the burden of proof. The German criticism with respect to the possible inefficient effects of the reform would be justified if the reform were to bring about a change in the burden of proof with respect to Article 81 (3). For the moment, Article 81 (3) has to be proven by the party who invokes it. I think this should be maintained and written into the new Regulation, because otherwise the burden of proof requirement could be interpreted differently in the future. This goes to the heart of the main objective of the reform: efficiency. Unfortunately, this also means that the difficulty we today have to distinguish between paragraph one and paragraph three remains, because I tend to believe that the burden of proof for paragraph one is different from the burden of proof for paragraph three. For paragraph one, the burden of proof is on the plaintiff, whoever that may be, whereas for paragraph three, it is on the party who invokes it. I do not believe that the burden of proof is a question linked only to procedures between private parties. I happen to belong to a court where the

306

Panel Discussion

burden of proof plays a much more important role than it has ever played, and probably will ever play, in Luxembourg. In practically every case we have before the Appellate Body of the WTO in Geneva, the question of the burden of proof arises. These are cases among States—not among private parties—and there are good reasons to explain why the importance of establishing the burden of proof arises, as neither the Panels nor the Appellate Body have investigative powers. In the light of our debate, a final aspect on which clarity is required is the supremacy of Community law. The question about whether Community law will or will not apply, or to what extent it will apply in the future (again, very prominent in the German debate), has to be addressed. These things should be spelled out in the Regulation; therefore, the Commission should make a proposal. If the Member States do not agree, it will be for the Court to clarify such aspects. But these aspects have to be addressed in the Commission's proposal. My second point is that I slightly disagree with Mario Siragusa: one of the extraordinary phenomena in EC competition law is the fact that the Commission is entitled, and will remain entitled, to decide on individual cases. Whether the Commission leads the EC competition policy debate or not will result from its power to apply the first paragraph of Article 81. Only the additional extraordinary and—from today's point of view—anomalous power of the Commission, the monopoly over 81 (3), will be abandoned. The third point is more of a philosophical nature. Referring to the German debate in February, I can very well understand those who say this is a false decentralisation initiative: in effect, its outcome would be more centralisation, and we do not want more centralisation. Why do I say this? First of all, if Emil Paulis is right—and I think he should be right—Article 81 (3), regardless of who applies it, will generate a pre-emptive effect similar to that which was previously generated only by the Commission's positive decisions. Consequently, there will be more supremacy of Community law than today. Moreover, if you decentralise, you would like to see more homogeneity between the players, whether the players are national competition authorities or national courts. This explains the implicit or explicit request for the respect of some minimum standards. And, indeed, in those areas of the internal market where the EU has moved to mutual recognition, there is minimum harmonisation. Competition is an exceptional field in which nothing like a directive imposing minimum harmonisation has ever been imposed. Nevertheless, we observe a phenomenon of voluntary, spontaneous approximation of national law. We all hope that this trend will continue. Otherwise certain national competition authorities will be obstructed by their status, by limitations of power, and by their procedures. However, the process in which we are now engaging will take time, and indeed it should take time. It would be suicidal if the Commission put forward proposals that conditioned the reform upon minimum harmonisation at this stage. This would probably be politically unacceptable. I would prefer that this does not have to happen at all, because a request for obligatory harmonisation would be an additional argument for those who say this is not a decentralisa-

2 - Coherence

307

tion process, but rather a process of re-centralisation. That is not the Community I would like to see, one in which every time you delegate something, centralisation returns through the back door. I have sympathy for the concerns expressed by both sides. However I would not like to go into the very merits of the remedies proposed, even though I fully share the concerns of those who apply institutional economics and say 'where are the incentives for the national authorities to act by taking into account Community-wide interests?' • EMIL PAULIS—I will try to say a few words on a couple of points raised during the day. First of all, there is of course quite a lot of questioning about the functioning of the 'network of competition authorities'. This morning somebody asked what happens if a national competition authority does not act at all. Take, for example, a case having the 'centre of gravity' in Belgium and where there is no authority wanting to deal with it because none has the resources. Or, in the perspective of enlargement, there might be competition authorities in the candidate countries that are not yet ready to act. For precisely these kinds of situations, the Commission proposes a system of parallel competencies that keeps the Commission in the game and allows it to intervene. However, this should not lead to a situation whereby the Commission has an obligation to intervene each time a national cartel authority does not act. The Commission should not become the institution that corrects all inefficiencies at a national level. There should be a certain degree of discretion for the Commission to decide whether or not to act in such cases. Another question was what would happen if more than one national competition authority is willing to act. It would be a nightmare to have a complaint before several national competition authorities, all acting in parallel. The Commission thinks that the rule should be 'one-stop-shop', though without totally excluding parallel action, particularly in cases in which there is a clear and serious infringement of competition law (such as a cartel whose effects are felt in more than one Member State). When the infringement is so clear and obvious there is no real risk of diverging decisions, two or even three national competition authorities can act in parallel. But the rule should be a one-stop shop, and here again the Commission would have a powerful instrument—the equivalent of Article 9(3) in Regulation 17/62—with which to intervene. The Commission will be informed, via an Intranet that we will create for the national competition authorities, of all actions started either by way of complaint or ex officio. Then there is a demand for clear criteria about the allocation of cases between competition authorities and for transparency. On the allocation of cases, some have even pronounced themselves in favour of a system similar to that in the Merger Regulation (where quantitative thresholds determine which cases have a Community dimension—and should therefore be of the competence of the Community—and which should be left to national competition

308

Panel Discussion

authorities). I do not think that such a distinction can easily be made in the case of behavioural agreements. Therefore I doubt that clear and strict quantitative thresholds could be established in this field. Notwithstanding this, the Commission intends to publish and discuss a set of criteria setting out the allocation of cases inside the network of competition authorities, so that companies know where they should launch a complaint in the first place. I believe that there is some merit in keeping a certain degree offlexibilityin the allocation of cases, particularly if we want to ensure an assessment of cases in the best place. But, of course, all of this can be discussed. Then the fear was expressed—which I understand as well—that national competition authorities might not be sufficiently independent, or that they might be influenced by national interests, and this might cut in two ways. Professors Neven and Mavroidis thought this might lead in practice to a situation whereby national competition authorities become too strict in enforcement, because—if I understood well—they would define national markets and then come down with too narrow a view, not taking into account the benefits of agreements from a Community-wide perspective. Others worry that national competition authorities might be too lax, or might not be strict enough with particular violations of Community law, either because they would be tempted to preserve national champions or because they would be put under pressure by their governments. Here again I would say 'yes' and 'no': 'yes', because that risk might exist—we are all human beings and we know that pressure exists—and 'no' because one of the great merits of the new system is that, by getting national competition authorities to apply more and more EC law, these authorities will become more and more bound by EC law. Moreover, national competition authorities will be stronger when applying EC law, because then they can say to national instances: 'Wait a minute, we are not among ourselves only, we have to take account of the fact that this is EC law, and ultimately the legality of our decision will be tested by an independent national court which may refer a question to the European Court of Justice. Therefore, we have to live up to the EC standards.' So, applying EC law might help national competition authorities whereas, at present, under national law only, the authorities might be under more pressure. I also think that, the more and the longer we apply the same EC standards, the more we will develop a common culture among national competition authorities, and that will also work towards a harmonised application of EC law. If, however, the worst comes to the worst—you have always to consider that scenario too—the Commission will have a powerful tool to intervene, which is the possibility of withdrawing cases from national competition authorities. This is a powerful tool both in the sense of preventing a too strict application of the law and in the sense of preventing its too lax application. • ALEXANDER SCHAUB—I agree with everything that Emil Paulis has explained so convincingly. This afternoon we witnessed a fascinating further development of our debate on the reform. My feeling is that a number of topics that have

2 - Coherence

309

preoccupied us for the last months—like efficiency, predictability and legal certainty—are now starting to be exhausted, and we have a broad understanding about where to go. This seems to me to be a logical development. We are getting now into a new phase of the debate, a phase that was triggered this afternoon by Mario Siragusa, Jim Venit, Ian Forrester and a number of other speakers, who asked questions that until now were not really at the centre of attention: doesn't the Commission go too far in its reform euphoria? Has it forgotten that co-operation might not work? Is the Commission aware that it could risk its leadership position and loose control over the enforcement process? What are the guarantees that the Community rules will be applied? I think these are extremely important questions. In any case, we would have needed to discuss them. Now, what are the answers to these questions? First, maintaining the Commission's monopoly over the application of Article 81 (3) is certainly not the only method by which to ensure the necessary dominance of Community principles in the area of competition law. What is the Commission aiming to achieve with this reform? We want to explore the national potential in the enforcement of European competition rules, and we do not want to do this by delegation—I think this is a misleading term—but by sharing enforcement competencies with the national authorities, which is a different thing. Moreover, we want to do this in a way that keeps the risks of such an exercise at a tolerable level. That is, I believe, the major challenge of the reform. The future 'cohabitation'—as it was described today—has to be organised in a satisfactory manner. I am convinced that the key to this problem is a well functioning network of competition authorities, as we are suggesting. The network of competition authorities is a concept to be understood in a broad sense. I would distinguish between two major components. The first is made of the rules dealing with the functioning of the network. Here again, I would distinguish between what Emil Paulis described this morning as 'the highest degree of primacy of Community rules', and the rules on the allocation of cases within the network. These are two different aspects, and it is quite clear that we have to discuss them both in more detail. Our discussion on these aspects is, in a certain sense, 'amputated', because a Commission 'non-paper' on the functioning of the network already exists (and has for some time), though it has not yet been widely circulated. However, this 'non-paper' has already been discussed quite extensively with the representatives of the Member States and it offers a number of answers to the questions raised on the functioning of the network. Coming back to the highest degree of primacy of Community rules and the satisfactory allocation of cases: I do not believe that there are simple solutions. I believe instead that we need a quite sophisticated degree offlexibility.We want to give the Member States a maximum role to play in the enforcement process, but we must also have some kind of safeguard in case things go wrong—for example, the Member States' authorities do not exist, or do not function—so as to avoid a blockage of the enforcement system.

310

Panel Discussion

Moreover, I believe that rules alone will not suffice to ensure a successfully functioning network. The great charm of the reform concept that we are developing is that, in addition to the rules on the functioning of the network in the broad sense, we also count very much on the development of a common competition culture as a result of co-operation, and we count on the Commission's ability to maintain an accepted leadership position within the network that is not legally imposed. The relationship between the Commission and the national competition authorities is very satisfactory nowadays; it has very much improved and it will further improve, but there is no guarantee that tomorrow things could not go wrong (in this unregulated area, things can easily go wrong, since what matters very much is the quality of persons, the chemistry between the sides, the general atmosphere, and so on). We therefore need to establish rules strong enough to allow the survival of the network even if the accepted leadership position of the Commission comes to be doubted. The conclusions to be drawn from all this are, in my view, relatively simple. First, we must focus over the next months on the issues related to the functioning of the network of competition authorities—both as a legal construction and as cultural concept—more than we did before. I suppose we will also need to create more transparency about our 'non-paper' on the functioning of the network, refine it, and ensure that this element—which is crucial to the reform concept—achieves a higher degree of maturity. Second, I agree with those saying 'piano, piano', but I wanted to remind them that, from the very first day, our position at the Commission was that we have to take our time, and that this operation is much too important to be rushed. Somebody said this afternoon: 'We are fed up with this discussion. Why don't you just do it?' That is an understandable sign of impatience, but that would be the worst mistake we could make, because clarifying these things is so important. And if we were to head this once towards a mistaken concept, we risk destroying what has been so painfully achieved over recent decades. So we will take our time, and most likely in September 2000 we will make a formal reform proposal. But, as you know, our formal proposal—which will be presented after more than three years of intensive work on this issue—is only the beginning of the formal reform procedure. We have always said that the whole reform exercise will certainly not be operational before 2003. So we do not feel in a hurry; we will take all our time, and we are extremely grateful to all those who have ensured that this afternoon's discussion was so exciting and fascinating. It brought us one important step further; we will think about it, we will listen further, and I am sure that then we will be able to come up with a refined network concept, offering all the reassuring guarantees that have so rightly been asked for this afternoon. • KAREL VAN MIERT—Thank you all very much for your contributions. It was a very lively and instructive day^ and I gather from what Alex Schaub said that the Commission will consider all your remarks very carefully. I felt privileged to be here and chair this meeting, and I did my best to be even-handed, and not

2-Coherence

.

311

get dragged into controversial discussions (at least not too much). Thank you all very much. • CLAUS-DIETER EHLERMANN—I just wanted to thank Karel van Miert on behalf of all of us for having spent this day with us. As Giuliano Amato has done in the past—Giuliano could not be with us because he is in Berlin with his fellow Prime-Ministers, sharing ideas on modern forms of governance—Karel made the impossible and disagreeable very pleasant and extremely interesting, particularly when sharing his experiences and convictions with us. Once again, many thanks on behalf of all of us.

PANEL TWO COHERENCE

2 WORKING PAPERS

Sir Christopher Bellamy Q. C. The Modernisation of EC Anti-trust Policy Some Reflections: Don't Throw the Baby Out with the Bath Water

As one who has had the privilege of grappling with the existing system of EU competition law as an advocate, a Community judge, and now a national judge, I warmly welcome the broad concepts set out in the Commission's White Paper. Apart from any other considerations, it is in my view conceptually absurd to maintain the Commission's monopoly over the grant of exemption under Article 81 (3). Similarly, the formal system of ex ante notification is unnecessarily bureaucratic in its present form. Even if it means that the new British competition law, recently introduced under the Competition Act 1998, has to be substantially modified, I am in general, supportive of the Commission's White Paper. Nonetheless 'the devil is in the detail'. I would therefore like to make a few comments about (1) the role of national authorities and (2) the role of national courts, in order to draw attention to various problems and pitfalls which could occur unless detailed thought is given to the implementation of the Commission's proposals. It would indeed be a tragedy if, in an effort to liberalise and improve the Community system, one finished up by seriously damaging what has been one of the real success stories of the Community to date.

1. The role of national authorities The United Kingdom authorities do not at present have power to apply Articles 81 and 82 directly, except in very limited spheres, one of which is air transport to third countries. The single United Kingdom experience to date of seeking to apply the Treaty in this area was not a happy one. It concerned the proposed alliance between British Airways and American Airlines. The national authorities took the view that the national carrier should not have to surrender to foreign airlines an excessive number of slots at the national airport, Heathrow. On the other hand, the Commission considered that a great many more slots ought to be surrendered as a condition of the deal going through. Eventually an impasse was reached and the deal was abandoned. I mention this case simply to illustrate that it is very possible for a difference of view to develop between a national authority and the Community authorities where a national authority is applying Community law especially in a case where its own national interests are also engaged. It is not a question of bad faith, far from it, or even of bias, conscious or unconscious. It is very largely a question of perception. If, when dealing with a case under Community law that

316

Working Paper I

involves both its own nationals and other Community nationals, a national authority decides in favour of its 'own' side, there may well be a perception of bias, even if the decision itself was totally correct. In other words, it is no accident that the European Commission has hitherto been seen as the best arbiter, and on the whole an impartial arbiter, in cases where the decision involves interests from different Member States: the Commission's involvement safeguards the perceived integrity of the system. In my view, when implementing the Commission's proposals, it is essential that everything be done to preserve the perception of impartiality and fairness in the system. That is even more important if one considers that in some Member States the competition authorities are not, or at least are not seen to be, entirely independent of government, or do not possess the experience or resources to deal with cases where significant cross-border issues are involved. It follows in my view that there is a limit to the extent to which the Commission may properly and appropriately abandon its natural role as arbitrator in cross-border disputes in the competition field. I would therefore argue that any case in which the nationals of more than one Member State are opposed to each other, there should be a rule, or at least a strong presumption, that such a case should go to the Commission for decision. What I have just said seems to have been the view expressed by the British Office of Fair Trading in its evidence to the House of Lords Select Committee.1 Secondly, rules are needed to determine who does what. Take for example a Hispanic agreement between a Spanish and a Portuguese company. They fall out. The Portuguese company complains to the Portuguese authorities, the Spanish company to the Spanish authorities. Who is to handle the case? In my view that cannot be left to informal arrangements between the authorities potentially concerned, in this case the Spanish, Portuguese and European authorities. That is because the question of who has jurisdiction to determine a complaint, or to impose a fine, is a question of law, which determines the procedure to be followed, the language in which the procedure is conducted and the rights of appeal of the parties concerned, to name but three basic elements. In all legal systems where there is a potential conflict of jurisdiction, rules exist for determining which jurisdiction is competent.2 It would, in my view, be contrary to the principle of legal certainty to have a system in which the question of who takes jurisdiction is decided not merely by informal, and, presumably, unaccountable and non-transparent, contacts between the various authorities. I note in passing that under the United Kingdom regime there are no less than six different competition authorities, that is to say the Office of Fair Trading and the regulators in the spheres of telecommunications, electricity, gas, water and railways. There exist detailed rules, called 'the concurrency rules', for determining which competition authority is to have responsibility for the case in the event that more than one author1 2

See 'Reforming Competition Law Procedures', HL Paper 33, 2000 p.145. See for example Article 9(3) of Regulation 17/62.

2 - Coherence

317

ity has jurisdiction. In the event of a dispute between authorities, the Minister is to decide. In my view, thought should be given to developing transparent, objective and binding criteria "for the allocation of cases between national authorities under the Commission's modernisation proposals. Thirdly, if the proposal is to abolish in its entirety 'one-stop shopping', in my view it is counter-intuitive. It is also swimming against the tide. In cognate areas such as mergers and trade marks it is widely recognised that a system of onestop shopping is beneficial, and indeed in the merger area the trend is towards reducing interventions at national level by lowering the threshold at which mergers are considered to have a Community dimension. I understand also that, in the parallel case of patents, there are serious moves afoot to establish a self-contained Community system, with a Community court dealing on appeal both with questions of grant and revocation, and also with infringement actions. Against that background, and without getting distracted by questions of terminology relating to 'positive' or 'negative' or 'partially positive' decisions, in my view there must also continue to be the possibility for the parties to obtain guidance from the Commission of the lawfulness of their agreements. In addition, the Commission should be able to rule in a legally binding way on certain agreements where the parties reasonably request such a ruling. Such agreements should include those involving major investments or new technology, including the new dot.com economy, or where agreements have an obvious Community dimension—for example pan-European distribution systems, or cases involving networks of European significance: for example infieldssuch as telecoms, energy, transport, financial markets and sport. Such rulings by the Commission should also in principle be susceptible to legal challenge, so that actions can be introduced before the Court of First Instance either by the parties themselves or, where there is appropriate standing, third parties. The question of the right to bring a challenge before the Court of First Instance, and ultimately the Court of Justice itself, is in my view of fundamental importance. Community law has been shaped so far largely by case law, through the process of appeal to the Community jurisdictions. In my opinion 'guidelines' and/or 'informal decisions' or 'indicative guidance' by the authorities are no real substitute for properly argued legally binding decisions having the status of law. In my view, neither the right of appeal against a legally binding decision of the Commission, nor the principle of the contradictoire before the Court, should be abandoned for the sake of administrative convenience. Beyond this line of reasoning there is a question which troubles me a little. By adopting these proposals is the Commission in effect, excluding or weakening a control juridictionneH To put the matter more concretely, under the new provisions what would happen in a case such as Perfumes! In that case there was a European-wide distribution system, which was the subject of a Commission decision, which was challenged by third parties. The result was a detailed judgement by the Court of First Instance. Is one abandoning that process in favour

318

Working Paper I

of pure 'administrative guidance'? Similarly, what would happen under the new system to a case such as European Night Services! Again, that case concerned a European-wide transport project, and was the subject of a decision by the European Commission granting an exemption subject to conditions. The need for an exemption, and the conditions subject to which the exemption was granted, were challenged by the parties before the Court of First Instance. There was subsequently a judgement of that Court which clarified important aspects of the scope of both Articles 81 (1) and Articles 81 (3). Is that kind of case no longer going to be subject to any legally contestable 'acte' by the Community authorities? If it be said that the route of legal challenge is henceforth to be no longer from the Commission to the Court of First Instance but by way of requests for preliminary rulings posed by national courts to the Court of Justice under Article 234 of the Treaty, my own view would be that that is a wholly unsatisfactory solution. The Article 234 procedure is extremely slow; it often involves parties having to reach the highest national court before a reference is made; and it involves posing questions in the abstract, often at one remove from the facts of the case. In any event the procedure of the Court of Justice is not adapted to deal with intricate questions of fact in the same way as that of the CFI. It will also be remembered that one of the main purposes of the CFI was to relieve the Court of Justice of competition cases and to improve the jurisdictional protection for those affected by decisions in competition matters. The CFI has, over the past ten years, carried out painstaking factual evaluation of the cases that come before it, and has acquired very considerable expertise in both procedural and substantive matters involving the application of competition law. It would in my view be regrettable if an unforeseen side effect of the new proposals was to bypass the CFI and to bring back to the Court of Justice cases which it had successfully divested itself of more than ten years ago. Finally, may I just put up my hand for a small but apparently fast vanishing right, the right to address a public authority in one's own language, and to have any decisions that are taken available in one's own language. What provisions will be adopted to ensure that parties have the right to address national authorities in their own language? What arrangements are to be made to ensure that decisions taken by national authorities are available in other Community languages or at least in one or more of the main working languages of the Community? My overall view is that there may well be many cases with a Community element that can conveniently be dealt with by national competition authorities, either under their own national law, or acting under Community law. However, for the reasons I have given, there is a limit to the extent to which the Commission can in fact abdicate its own responsibilities under the Treaty to develop and enforce Community competition law. That is perhaps particularly true outside the area of cartels, which it is fashionable to consider, at the moment, as the priority of the authorities.

2-Coherence

319

2. National Courts Having been judicially responsible for applying Article 81 (3) at Community level, and now being judicially responsible for doing so under the equivalent provisions of national law, I can hardly say that judges are not capable of applying the economic concepts inherent in that Article. Although Article 81 (3) is expressed in apparently vague language, I am sure that with a little more case law it would not be difficult to fashion more clear rules for the application of Article 81 (3). However, on the assumption that national courts as well as national authorities are to have jurisdiction to apply Article 81 (3), I would like to point out that there are limitations to what one can reasonably expect of a non-specialised national court in this respect. Most of these limitations result not from the lack of expertise in the court itself, but from the nature of the litigation process. First, in general a national court has only a limited vision of the parameters of the case. In almost all systems the national court is limited to what the parties put in front of it, and does not have the same investigative powers as national authorities. In common law systems the national court is almost entirely limited to material produced by the parties. It is true that in civil law countries and in some other systems the national courts have various powers of 'instruction', but these in my experience are not in practice anything like equivalent to the powers of a competition authority. The national court will therefore sometimes lack the material it needs to decide such questions as what is the relevant market, what are the main competitive forces at work, what are the advantages and disadvantages, from a competition point of view, of granting or withholding an exemption, and what conditions such an exemption should be subject to. Although it is open to the courts in some legal systems to demand 'expertise', or to receive expert reports, I would not myself be in favour of a system which risked effectively devolving the decisions on the development of Community competition law to experts appointed by courts at national level. Secondly, before the national court there is typically an immense disparity in strength as between the parties before the court. Typically, litigation will take place between a strong—possibly dominant—company, perhaps part of a multinational group, which is pursuing, or being pursued by, some smaller company. In general the larger company will have much greater financial resources to sustain litigation, and through that fact alone may force the smaller company to desist or settle on unfavourable terms. In the event of the dispute continuing, the larger company will generally have the resources to place before the court all the necessary research materials, experts opinions and skilled argument to convince the court that, after all, there is no breach of Article 81 or Article 82. The smaller company, by contrast, will typically have more difficulty in assembling the materials that are needed, or even in being able to afford to fight the case at all.

320

Working Paper I

If the present proposals are intended to herald a shift of emphasis whereby parties are to be left more and more to their remedies in the national courts, with less intervention from the Community or national authorities, such a development by its nature favours the big pocket of the large firms at the expense of the small and medium-sized firms. It has always been an honourable and in my view excellent feature of the Community system that even a small company from anywhere in the Community can go, at moderate cost, to the Commission and seek effective protection. If I may put the point in a rhetorical way, under the new system which jurisdiction would handle a case such as ECSIAKZff! I doubt whether is it appropriate or desirable that such leading cases should no longer be primarily for the Commission to pursue. Thirdly, it is worth reminding ourselves that in the parallel area of patents, previously mentioned, some of the judges are in despair at the problem of forum shopping, as emerges from the evidence given by the English judges to the House of Lords Select Committee.3 Many techniques are employed. If one fears being sued for infringement, one can start anticipatory proceedings in Italy, secure in the knowledge that it will be many years before the case comes to trial, and that in the meantime other national courts will have to decline jurisdiction under the provisions of the Brussels Convention. Alternatively one may proceed in a jurisdiction such as the Netherlands, and obtain very quickly injunctive relief that might, on the face of it, be effective throughout the European Community, and would, again, preclude another national court from taking a different view. These problems are part of the reason why, in the case of patents, the movement is towards a Community system with a Community jurisdiction, and away from a fragmentation of responsibility at national level. There is, therefore, some risk of forum shopping if the new proposals are intended to transfer to national courts the role that has hitherto been performed by the Commission. It is, in my view, naive to suppose that the provisions of the Brussels Convention constitute an adequate answer to the problems that arise. There is also the problem that a very large number of disputes in any case could go to arbitration, which falls entirely outside the national and Community legal systems. Fourthly, what procedure is going to be adopted for dealing with cases before national courts that have important Community consequences? I can illustrate this question by raising another question, which I call 'the Cardiff question'. The Cardiff Rugby Football Club was in dispute with the Welsh Rugby Union, and claimed that the various rules to which it was subject were contrary to the competition provisions of the Treaty. The case, as I understand it, was pending before a national court in Wales. The International Rugby Union intervened in the case and obtained a stay of the proceedings on the grounds that the International Rules of Rugby had been notified to Brussels and that an exemption decision was expected. The situation in this case was that the national court 3

Op. cit. pp. 63-70.

2 - Coherence

321

was seized with a problem that had repercussions far beyond the borders of its own jurisdiction, and in effect involved the whole international organisation of the sport of rugby. The national court, by definition, could not see the whole picture. The case therefore cried out for a mechanism whereby it would be decided by an authority with the means to take into account all the relevant considerations from the point of view of the entire Community interest. In other words, under the modernisation proposals, what mechanism is there going to be for ensuring that such decisions are ultimately taken not by national courts but by institutions of the Community? For the reasons I have already given, I very much doubt whether the Article 234 route represents a feasible solution. In any event, as can be seen from the recent Masterfoods litigation, it has taken some eight years for the ice cream litigation to finally reach the Court of Justice from the Irish courts under the Article 234 procedure. Is that an acceptable system in what is supposed to be a communaute de droiti Similarly the suggestion that the Commission could intervene as amicus curiae in some cases, is not in my view a total solution. In order to play usefully the role of an amicus, one must know the file intimately, and also be familiar with the language and national procedure of the case. For the Commission to provide such a service to the national courts of all the Member States would seem to me not to reduce the burden on the Commission, but to have substantially the opposite effect. In addition to those points, there is the simple fact that many national courts are already hopelessly blocked, sometimes for a decade or more, and decisions by such national courts are likely to be inaccessible for language reasons. Moreover, although one may, with appropriate training, achieve greater judicial familiarity with competition law issues, it must not be forgotten that judges are trained to seek to uphold contracts where they can. Many a judge is, by natural disposition, rather unsympathetic to a litigant who seeks to escape a contractual obligation in reliance on a point of illegality. Similarly many judges, trained to apply strict legal rules, are likely to feel uncomfortable where what they are being asked to do is, in effect, to develop matters of competition policy, e.g. in the application of Article 81 (3) and in some cases under Article 82.4 The overall result of these reflections is that one should not expect too much if one gives jurisdiction to the courts to apply Article 81 (3). There will be at least some surprising and erroneous decisions emerging, because of the inherent difficulties of the litigation process described. Again, my own solution would be to give further thought to the need to construct mechanisms for channelling at least the more important cases towards the Commission or, perhaps, towards experienced jurisdictions at national level. These certainly already exist, e.g. the Cour d'Appel in Paris, the Tribunal de Commerce in Brussels, the Oberlandsgericht in Frankfurt, the Market Court in Sweden, to name but a few.

4

See in this respect the comments of the English judges to the House of Lords, cited above.

322

Working Paper I

Perhaps, and this is my final suggestion, one might also consider an exchange of judges at national level, so that cases with a Community dimension which are tried before the courts of one Member State could in fact be tried by a court which consists of at least one judge from another Member State. Perhaps that, the ultimate sign of integration within the Community, might come one day.

II Jacques H. J. Bourgeois Decentralised Enforcement of EC Competition Rules by National Competition Authorities: Some Remarks on Consistency, Coherence and Forum Shopping

I. Introduction 1. Topics addressed The present paper addresses two aspects related to the decentralisation of the enforcement of the EC competition rules applying to undertakings that are advocated in the European Commission White Paper on Modernisation of the Rules Implementing Articles 85 and 86 (now 81 and 82) of the EC Treaty:1 the 'consistency and coherence' mechanism, and the options for minimising the risk of forum shopping and divergent decisions suggested by the White paper.2 Although the two subjects are to some extent the two sides of the same coin, they are treated separately for the sake of clarity. This paper focuses on enforcement by national competition authorities ('NCAs'). Decentralisation to national courts is discussed in other papers. The difficulties to which the proposals of the White Paper, if implemented, would give rise for the enforcement of the EC competition rules in newly acceding Member States are also left aside. Before addressing the two subjects, it might be useful to examine the assessment of the risk of inconsistency and incoherence that is implicit rather than explicit in the White Paper. 2. Risk assessment When it issued its 'Notice on Co-operation between National Courts and the Commission in Applying Articles 85 and 86 of the EEC Treaty',3 the European Commission probably already had at its disposal the study on the application of the EC competition rules by national courts that it had asked the Union Internationale des avocats to carry out. 4 1

OJC 132, 12.5.1999, p. 1. I thank Hartmut Johannes, of counsel to my firm, for his comments. Opinions are mine only. 3 OJ 1993 C 39/6. 4 Braakman (1997); see also Behrens et al. (1994), and the literature cited by Temple Lang (1998) at p. 265, fn. 1. 2

324

Working Paper II

So far, the exchange of information on enforcement of EC competition rules appears to have been mainly a one-way affair. Through the Advisory Committee, the EC Member States are kept well informed about European Commission actions. Pursuant to Regulation 17/62 and its parallel sectoral regulations, as well as pursuant to the Merger Control Regulation, the European Commission is under a duty to do so. No EC regulation imposes a similar duty on Member States to inform the European Commission of their activities. It seems that so far only German law provides that the European Commission is to be informed of cases of application of EC competition law so as to enable it to state its views.5 As is apparent from the annual reports on competition policy, which for some time now have listed the enforcement decisions of Member States, the European Commission seems to have persuaded Member States to communicate information on actions that they have undertaken to enforce the EC's competition rules. However, the question is whether such information is systematically submitted. Moreover, while Member States are consulted within the Advisory Committee on the major enforcement steps of each case handled by the European Commission, the converse does not seem to be true. At least to outsiders, the evidence on enforcement of EC competition rules by NCAs is patchy. At least, it was patchy at the time the reports presented at the XVIIIth Fide Congress on National Application of Community Competition Law.6 As appears from these reports, the enforcement is quantitatively limited, if not non-existent, depending on the Member State in question.7 In other words, the risk assessment carried out by the European Commission seems to be based on what is likely to happen in the event of decentralisation rather than on past experience with decentralised enforcement which, it should be stressed, is limited to enforcing Article 81. 5

Gesetz gegen Wettbewerbsbeschrankungen, para. 50 (3). Temple Lang (1998). 7 E.g. Germany: 'increasingly' since the amendment of the GWB enabling the BKA to enforce EC competition rules (German Report at p. 22), but in fact mainly in sectors where German competition law provides for an exception in particular the energy sector (German Report at p. 21); Austria: since 1994 between 60 and 100 proceedings have been initiated (Austrian Report at p. 29); Belgium: non-existent (Belgian Report at p. 53); U.K.: a regulation enabling the NCAs to enforce EC competition rules came into force in August 1996; one pending case reported (U.K. Report at p. 67); Denmark: apparently no case (Danish Report at p. 79); Spain: 98 cases (including some double-counting) (Spanish Report, Appendix at p. 133 et seq.); Finland: a few cases (Finnish Report at p. 148); France: 2 to 3% of the decisions of the Conseil de la concurrence (French Report at p. 159); Italy: no important effect of decentralisation, the Authority having enforced national competition rules (Italian Report at pp. 194-195); Luxembourg: no law enables the national competition authority to enforce EC competition rules (Luxembourg Report at p. 203); Netherlands: at the time of writing the report the new NCA was not yet in operation (Netherlands Report at p. 215); Portugal: only a few cases identified (Portuguese Report at p. 239); Sweden: no cases where EC competition rules have played a substantive role before the Competition Authority (Swedish Report at p. 249). 6

2 - Coherence

325

The modalities of the decentralisation and theflankingmeasures are thus all the more important, even though the likelihood of the risk of incoherence, inconsistency and forum shopping cannot effectively be assessed.

II. Coherence and consistency 1. Coherence and consistency of what? The White Paper's attention is focused essentially on the need to ensure coherence and consistency of actions that would be undertaken by NCAs. Little is said about competing claims of jurisdiction between NCAs, or between NCAs and the Commission. How conflicts of jurisdiction—both positive and negative ones—are to be solved is left aside here. This matter is more appropriately dealt with under the heading 'allocation of cases'. It will be addressed briefly in the section on 'forum shopping'. The only point about conflicts of jurisdiction to be mentioned in this section relates to negative conflicts of jurisdiction, that is, where no NCA is willing or able to act and where the Commission is equally unwilling to act. 2. Non-action 2.1. The issue When dealing with complaints, the White Paper refers only to complaints lodged with the Commission. These are expected to take an even greater importance than at present. Yet, nothing is said about complaints about the infringements of EC competition rules lodged with NCAs that will equally acquire a greater importance. There might be many reasons why an NCA does not want to act on a complaint. There are, however, situations where rejection of a complaint is particularly relevant for the decentralisation process. One of these is when an NCA does not act against national undertakings that 'export' anti-competitive effects, particularly when the conduct has no competitive effects on the national market. In such a situation, redress could probably not be obtained in the courts of that Member State either. Even if accepting, as the White Paper states, that NCAs 'do not set out to decide disputes between parties but rather to guarantee the maintenance of a system ensuring that competition is not distorted',8 non-action by an NCA should nevertheless be addressed. Non-action 8

Para. 91.

326

Working Paper II

by an NCA is as much a problem of consistency and coherence as actions of NCAs and of the Commission. Moreover, complainants risk being smashed between the Charybdis of the 'no Community interest' of the Commission and the Scylla of the NCAs' indifference. 2.2. Possible remedies Should no NCA be able to act (because either no allocation rules exist, or no EC rules confer jurisdiction on the NCA, or the NCA is unable for jurisdictional reasons to collect the necessary evidence), the Commission should be under a duty to act on a complaint. Arguably such duty already exists.9 Should an NCA be unwilling to act in aprimafacie case of infringement, the Commission should accept a complaint if it appears that no effective judicial remedy is available. Even if a duty to act would be provided for NCAs by a new Regulation 17/62, the remedy of infringement proceedings under Article 226 EC would hardly be practicable. Alternatively, the soft option in the form of the mechanism for the exchange of information and co-operation between NCAs and the Commission-—as currently contemplated by the White Paper to apply only in cases in which EC competition rules are applied10-—should be extended to cases where no NCA is willing to act. What if the 'consultation and co-operation' mechanism does not result in action? Is it subject to judicial review? And by which court? 3. Action 3.1. The issues It would seem that the White Paper is underestimating the risk of inconsistent decisions by NCAs. It might well be that the conditions for the prohibition of Article 81(1) have already been sufficiently clarified through case law and decision-making practice to be known to undertakings.'' (Nonetheless private practitioners would do well to review their professional liability insurance policies!) Whether all NCAs are familiar with the conditions of the prohibition of Article 81 (1) and of the exemption under Article 81 (3) is doubtful. As indicated earlier, many have limited practical experience in enforcing Article 81 (1), and some do not have any practical experience at all. Moreover, and as the German Cartel Office pointed out, there remain divergences of views on competition policy among Member States that may result in differing competition assessments of comparable fact situations. Decentralising 9

The CFI has found such duty in certain circumstances in cases where the Commission rejected complaints against infringements of competition rules. See e.g. Case T-24/90 Automec //[1992] ECR 11-2223. 10 Para. 105. 11 Para. 78.

2 - Coherence

ZTl

exemptions will exacerbate this problem to the detriment of the protection of competition.12 One should not discount the risk that policy considerations other than those listed in Article 81 (3) are taken on board by NCAs, particularly where national law mandates NCAs to take other policy considerations into account when enforcing national competition rules.13 The potential conflicts that, according to the White Paper, raise problems of coherence and consistency are of two kinds: (a) An NCA (or a court) takes a favourable approach to a restrictive practice prohibited by the Commission; or (b) An NCA prohibits a restrictive practice (or a court refuses to enforce it), despite a positive approach taken to it by the Commission.14 It is interesting to note that the White Paper does not address either the potential positive conflict of jurisdiction between NCAs or the potential conflict between actions taken by NCAs. This will be examined briefly in the section on forum shopping. 3.2. The remedies Apart from block exemptions with a wide scope of application, the White Paper contemplates another more general measure: guidelines and individual measures to clarify the scope of application of Articles 81 (1) and 81 (3) outside the block exemptions,15 and also more notices.16 The White Paper notes that such notices and guidelines might not be binding on NCAs, except where individual Commission decisions would confirm the approach set out in these notices and guidelines and these decisions would be upheld by the EC courts.17 The reluctance of the Commission to push for, and the opposition of several Member States against, Commission guidelines that would order national authorities what to do, are well known. However, there are arguably precedents in EC regulation in other areas that do exactly that. Another device contemplated by the White Paper is 'individual decisions that are not prohibition decisions', to be adopted by the Commission 'on grounds of general interest' where a transaction raises a new question.18 Such decisions would be of a declaratory nature and would have the same effect as a negative clearance.19 12 BKA: Praxis und Perspektiven der dezentralen Anwendung des ECWettbewerbsrecht 24-25 (23 September 1998). 13 E.g. see the amendment proposed to Article 10-2° of the French Ordinance mentioned by Laurence Idot in her paper. 14 Para. 101. 15 Para. 78. 16 Para. 86. 17 Ibid. 18 Para. 88. 19 Para. 89.

328

Working Paper II

It has been pointed out that these sort of decisions raise a series of questions.20 In addition, if they are meant to be remedies to prevent inconsistencies and incoherence, should they be qualified as 'negative clearances'? On the other hand, their relationship with pending or adopted decisions of NCAs relating to the same transaction needs to be clarified. 3.2.1. Preventive measures The White Paper contemplates that the Commission would have to be informed of the initiation of proceedings by NCAs. It would also have to be informed before the termination of procedures by NCAs, as well as of a planned NCA withdrawal of the benefit of a block exemption. This is expressly not meant to be 'a machinery to impose solutions to conflicts in the application' of the EC competition rules.21 The Commission would, however, still have the possibility of pulling in a case and thus taking it out of the jurisdiction of an NCA.22 This calls for two remarks. First, one wonders whether this will work effectively. One of the practical reasons for the decentralisation exercise is the manifest absence of sufficient Commission resources. Will the ending of the notification and authorisation system free enough resources to deal with collecting information about national proceedings (the latter being likely to increase significantly in number) effectively and in a timely fashion? There would, at any rate, be a need to subject the Commission intervention to strict time limits.23 Second, there are possible alternative preventive measures to the Commission pulling in a case, such as granting the Commission the possibility to veto a planned NCA decision or to issue binding instructions to NCAs. From the perspective of undertakings, the choice between the preventive measures would depend inter alia on whether there would be a one-stop shop system. If not, the Commission pulling in a case would be preferable. 3.2.2. Corrective measures The White Paper contemplates three possible cases:24 Case 1: An NCA adopts a [positive?] decision that is still open to appeal; the Commission may at any time adopt a contrary decision.25

20

Mestmacker (1999), at 523-527. Para. 105. Apparently, only conflicts between N C A s and the Commission are contemplated. 22 Ibid. 23 There may also be some technical difficulties for certain N C A s to comply with such information requirements. In addition, there are concerns about confidential business information. 24 Para. 102. 25 The time limits will differ between Member States. In some cases the Commission will have to make u p its mind in a short period (e.g. 4 weeks to appeal a decision of the Kartellgericht in Austria, and also 4 weeks to appeal a decision of the Competition Council in Denmark). 21

2 - Coherence

329

Case 2: An NCA adopts a positive decision that is no longer open to appeal or that has been confirmed on appeal; the Commission may always intervene to prohibit the agreement (subject to res judicata). Case 3: An NCA adopts a negative decision; the Commission normally should not intervene (except on a reference for preliminary ruling by a national court). Understandably, Case 2 does not find much favour on the part of undertakings. If the information and co-operation system between NCA and the Commission operates well, why should the Commission be given the right to prohibit an agreement that has been approved by an NCA after the time referred to in Case 1 has run out? If the information and co-operation system does not operate well, why should the undertakings' legal certainty be sacrificed on the altar of competition correctness? If the concern not to allow incorrect decisions goes beyond a dislike of letting the parties 'get away with it', and is perceived to be a major problem that cries for a solution, then another remedy should be sought. One such remedy could be a kind of 'recours dans l'interet de la loi' by the Commission in the Court of First Instance of the EC. As to Case 3, it is unclear whether the contemplated policy of nonintervention applies to negative NCA decisions in general, or only to negative decisions that are no longer open to appeal. Case 3 raises a question in relation to the Commission's intention to adopt 'individual decisions which are not prohibition decisions'.26 Will such decisions also be adopted with respect to transactions against which an NCA has taken a negative decision? If so, it would seem that the 'individual decision which is not a prohibition decision' and is assimilated to a negative clearance would let the negative NCA decision stand.

III. Forum shopping The concerns voiced by Member States, industry and lawyers about forum shopping relate mainly to judicial enforcement.27 They are not dealt with in this paper which focuses on NCAs. Stricto sensu, 'forum shopping' refers to complainants searching for the NCA that is most likely to act on their complaint and offer them the most attractive remedy. It could alternatively mean that parties to a restrictive agreement will try to have their agreement somehow blessed in such a forum. In comments on the White Paper, 'forum shopping' has also been used (inappropriately) to describe the fact that more than one NCA might have jurisdiction and may exercise it over a given transaction. 26

Para. 88. See supra at p. 7. The White Paper does refer t o the Brussels Convention on Jurisdiction and t h e Enforcement of Judgments in Civil and Commercial Matters (para. 100) but does not address the question whether this Convention should be amended. 27

330

Working Paper II

Whether the feared risks of true 'forum shopping' or the exercise of concurrent jurisdiction by NCAs are genuine and widespread possibilities is difficult to say without an analysis of the rules denning the jurisdiction of the NCAs and of the practices they follow. At any rate, it has been noted that, even though NCAs may generally only act within their national territories, it would seem that many of them apply an 'effect' doctrine of some sorts. Cases involving more than one NCA, and which produced asymmetric results, have been reported. This is likely to increase significantly if enforcement is decentralised. At least two systems are conceivable. One consists of introducing a mutually exclusive scope of application for EC and national competition rules on the basis of a reinterpreted intra-EC trade effect criterion. Another is to define the respective Commission and Member State jurisdictions to enforce EC competition rules.

1. Mutually exclusive scope of application of EC and national competition rules Several writers—some of whom draw inspiration from the Merger Control Regulation—advocate a narrower scope for EC competition rules and a system of exclusive jurisdiction (as opposed to the current system in which EC law and national law are applied in parallel).28 From the perspective of competition policy, the main argument in favour of a system of mutually exclusive EC and national competition rules seems to be that the supremacy of (unenforced) EC competition rules effectively precludes enforcement of both EC and national rules.29 It is true that substantive national competition rules are now much more like EC competition law, but harmonisation is still imperfect and incomplete.30 Moreover, in a system where two sets of competition rules apply concurrently, if one competition authority is not doing its job, is preferring to lead a quiet life, or is being too activist, another competition authority with concurrent jurisdiction can contest that market by providing its own 'product' of government services.31 If they apply different sets of competition rules, forum shopping is even more likely. In addition, one wonders what sort of brightline test to separate the respective scopes of the EC and national competition rules would appropriately take account of the real world, while at the same time ensuring an effective compe28

For a discussion of this problem, see Bos (1997), at p. 410. Wesseling (1997). Also advocating a limitation of the input of cases in the EC competition law system in Rodger (1995). 30 Temple Lang (1998) at p. 286. 31 Marsden (1997), at 238-239. 29

2 - Coherence

331

tition policy through the enforcement of either EC or national competition rules. According to an analysis submitted by the Commission staff to the Advisory Committee in 1994, one third of the cases handled by the Commission had their competition and economic impact in a single Member State. Assuming that these cases had no impact whatsoever in other Member States, this means that two thirds of the cases handled by the Commission had a competition and economic impact in more than one Member State.32 The option of the White Paper is not a system of mutually exclusive spheres of application for EC and national competition rules, but a sharing of authority in the enforcement of EC competition rules between the Commission and NCAs with regard to co-ordination, flanking and preventive and corrective measures. For the reasons set out above, it is submitted that this is the right option in the current circumstances.

2. Defining the respective Commission and Member State jurisdiction to enforce EC competition rules 2.1. The White Paper The White Paper does not set out precisely how responsibility would be allocated between the Commission and the NCAs. The Commission would act where a transaction raises a new question33 and would concentrate 'its activities on the most serious restrictions of competition',34 'on the basis of sufficient Community interest'.35 When a case 'whose effects are essentially concentrated in a single Member State' comes before the Commission, it would pass it on to the NCA 'best placed to deal with it'.36 Where 'the effects of a disputed practice are felt primarily in one Member State' the Commission would send the whole of the file to the competent NCA, and the same procedure would apply conversely.37 Nothing is said about the allocation of cases among NCAs, except for a reference to 'the authority best placed to deal with it'.38 This is hardly satisfactory for users. If left to the vague criteria put forward in the White Paper, cases would be allocated between the Commission and NCAs on a case-by-case basis in a less than transparent way (probably through consultations within the Advisory Committee, although it is unclear whether there would also be such consultations about the allocation of cases between NCAs). This is obviously not conducive to the creation of legal certainty. 32 The question is whether cases handled by the Commission reflect what is happening in the real world. 33 Para. 88. 34 Para. 118. 35 Para. 119, it being understood that a notice would clarify this concept. 36 Para. 96. 37 Para. 97. 38 Para. 96.

332

Working Paper II

In addition, there is the legitimate wish of industry that a transaction would not be subject to the concurrent jurisdiction of several NCAs, each one enforcing the EC competition rules in its own territory. Such concurrent jurisdiction could lead to conflicts of jurisdiction and to conflicting decisions. The figures mentioned earlier showing that two thirds of the cases handled by the Commission in the relevant period had an economic impact in more than one Member State tend to indicate that the potential that more than one NCA would handle a given transaction is significant.39 The White Paper suggestion that the transaction be handled by the best placed NCA is a half-measure when a transaction has a competition and economic impact in more than one Member State. As a reaction to comments on the White Paper, the Commission is apparently considering the possibility of taking charge of transactions that would have an impact in more than one Member State and when there is a risk of conflicting decisions. Would that not be defeating the purpose of the decentralisation exercise? And, more importantly, from a practical point of view would the Commission have the necessary resources to do this, particularly if the decentralisation exercise leads to increased enforcement of the EC's competition rules? 2.2.

Why not a 'one-stop shop'?

Under a 'one-stop shop' system—for which there is strong support on the side of industry—a decision of a national authority applying EC competition rules would be binding and enforceable in all Member States. In other words, such a decision should be given 'full faith and credit' throughout the EC. It should be borne in mind that NCAs are enforcing rules that apply in all Member States when they deal with transactions under EC competition rules. In other areas of EC law, departures from the principle of the territorial limitation of Member States' enforcement have already been accepted.40 Why would this, as a matter of principle, be unacceptable for the enforcement of EC competition rules? Such a system would need to embody some brightline test to ensure predictability for industry and to make it administratively practicable.41 If would also apply to the allocation of cases between the Commission and NCAs. 39 As indicated earlier, a proper assessment would require an analysis of the rules defining the jurisdiction of NCAs and of the practice they follow. 40 E.g. under the Second Banking Directive (Directive 89/646/EEC, OJ 1989 L386/1), the decision of the 'home' Member State of a financial institution finding that it fulfils the requirements of the Directive is binding on other 'host' Member States on whose territory offers financial services or establishes a branch or a subsidiary; see also for insurance companies (Second Insurance Directive 88/357/CEE, OJ 1988 L172/1); for decisions of Member State customs authorities see Community Customs Code (Regulation 2913/92, Art. 250, OJ L 302 of 1992). 41 See the several possibilities examined by the BKA (1999) n. 12, at pp. 23-24.

2 - Coherence

333

NCAs are currently subject to territorial limitations when investigating cases. A new Regulation 17/62 should set out the necessary duties of NCAs to cooperate with the NCA investigating the case. Concerns have been expressed about the fact that some NCAs are less efficient than others, and that this would create a playingfieldthat will be anything but level. This risk would be heightened where the application of the allocation rules would result in a case being referred to a NCA that would not be up to the task or would not enjoy the necessary confidence of the parties. In order to take this concern on board, a sort of opposition procedure could be introduced resulting in the Commission pulling in the case. Other solutions are conceivable. However, allowing a Member State to opt out of the effect of a decision on its territory should be ruled out.

Ill Laurence Idot A French Point of View on the Radical Decentralisation of the Implementation of Article 81(1) and (3)

I. Introduction For a French legal expert, the proposals made by the Commission in its White Paper are, generally speaking, not very surprising and lead to a rather positive reaction. This might be explained by the fact that we are familiar with the socalled 'legal exception system', which the Commission would like to use to replace the existing 'authorisation system' that has been in force since 1962. France defended this solution back in Messina in 1956, and the French competition law system has always adopted it. Established from the start in the decree of 1953, which prohibited anti-competitive agreements for the first time, it was not questioned during either the main reform in 1986, which lead to the adoption of Ordinance 86-1243 (1 December 1986), or during later modifications of this text.1 In this debate on the White Paper, it might be useful to first refer to the French experience. Title III of the 1 December 1986 ordinance dealing with anti-competitive practices is, like the EC competition rules, based on a principle of prohibiting restrictive agreements and abuses of dominant position under Articles 7 and 8 (1). Further, an exemption mechanism has been inserted under Article 10. Article 10 (2), in particular, provides that practices which fulfil the four conditions of the text, similar to Article 81 (3) EC Treaty, 'do not fall under articles 7 and 8'. Besides the fact that exemption might be beneficial to abuse of dominant position—something that has never happened in practice— the main difference with respect to Community law is one of a procedural nature. Indeed, there are no provisions for a notification mechanism in order to obtain an exemption. Moreover, all bodies enabled to implement Title III of the ordinance—that is, the specialised authority which is the Competition Council and the ordinary courts—may apply Article 10 (2) to the practices they have to handle and that, in their view, fulfil the conditions of the prohibition. Accordingly, it is not possible to agree with the criticism that the Commission challenges the very principle of prohibition enshrined in the Treaty.2 The French example shows, to the contrary, that the principle of prohibition no way excludes application of exemption by different bodies. 1 Thefirstmajor reform in 1996 did not affect the Title III of the Ordinance no. 861243 on anticompetitive practices. A new major reform is presently being discussed, but is no significant debate on this issue. there 2 Mestmacker (1999), at p. 523.

336

Working Paper III

However, some paragraphs of the White Paper, especially points 88 and 89, are ambiguous in their current form. First, it is true that the Commission would no longer be able to adopt constitutive exemption decisions under Article 81 (3). Nevertheless, a decision in which the conditions provided for by Article 81 (3) are fulfilled is a decision of exemption, whatever its name (positive decision or other) even if its effects will no longer be the same. Second, it is confusing to say that the Commission would have to take a global approach, whereas the structure of Article 81 requires reasoning in two stages. The French example shows precisely that there might be decisions in which the benefit of Article 10 (2) is granted after an infringement of Article 7 has been noticed. However, decisions referring to article 10 (2) are rare.3 At the French Competition Council, the number of debates about the application of Article 10 (2) is indeed limited.4 Most of the decisions are either decisions about prohibition, inadmissibility or lack of evidence. This might be explained by practical as well as more theoretical reasons. The parties will avoid losing time with proceedings at the Council if their agreement obviously fulfils the conditions of exemption. There is a prior evaluation. This applies particularly if the French Ministry of Economy introduces the action. Moreover, the Council largely applies the 'rule of reason' approach under Article 7, enabling it to interrupt its reasoning if the global outcome of the agreement is positive.5 It is much more difficult to evaluate the application of Article 10 (2) by ordinary courts as there is no systematic publication of the decisions. The analysis of published judgments does not enable us to find any applications of this text. However this acknowledgement is insignificant, as most of the cases that are referred to ordinary courts also fall under Community law. There are frequently debates on the scope of block exemption regulations, especially in the field of car distribution. Moreover, in the absence of 'BER', (which is the case, for instance, in the perfumery sector), undertakings often oppose the fact that the Commission has been notified of their agreement. Accordingly, it is very tempting to conclude that—in view of the French example—the so-called legal exemption system works perfectly well. We should however be careful when drawing such a conclusion. The French system works in the French context of a very centralised and unitarian state. Moreover, as the Commission is systematically notified about relevant co-operation agreements to, the French authorities have a smaller amount of work. More important difficulties arise when a legal exception system is set up in an almost federal context, in which Community law co-exists with national laws and different types of bodies with different types of training have to apply the same rules. One of the major challenges of this reform is precisely to maintain a coherent application of EC competition law. 3 4

The benefit of article 10 (2) seems to have been granted for the last time in 1993. For some explanations of the non-application of article 10 (2), see Blaise (1997), at

p. 9. 5

Blaise (1997), at p. 85.

2 - Coherence

337

A preliminary remark can be made. The issue of coherence is typical for any decentralised application of the law. This is far from being a new debate because the prohibitions provided by Articles 81 (1) and (2) have always been considered to be applicable by national authorities and courts. Moreover, it is widely known that the Commission has, especially since the beginning of the nineties, always been in favour of such a decentralisation. However coherence was not the most important criterion during the debate on subsidiarity; the debate focused instead on the distribution of competence and on the famous notion of 'community interest'.6 This might be explained by the fact that only the repressive aspect was at stake at that time. It is undoubtedly easier to reach an agreement on harmful conduct that should be prohibited, rather than on exemptions. This is demonstrated by way of international texts. It is significant that the only text dealing with substantive competition issues that has been adopted within the OECD framework is the 1998 recommendation on 'hard-core cartels'.7 On the contrary, as soon as there is a possibility of exemption, facts change. Even if everybody refers to the same text—Article 81 (3)—its content is vague enough to allow room for divergent interpretations. The answers given by the Commission in order to guarantee a coherent and uniform application of EC law within this new context are mainly of a procedural nature. This is a first step, but the first solutions that were suggested are insufficient, whatever their advantages and drawbacks (Section II below). There is one basic fact to be remembered: contrary to merger control, Articles 81 and 82 are not applied independently, but in conjunction with national laws. The White Paper automatically leads to thoughts on the parallel application of EC law and national laws (Section III below).

II. Coherence and isolated application of EC competition law If considering nothing but the application of EC competition law (without taking into account the parallel application of national laws), the White Paper reveals a first positive factor in favour of coherence. When dealing with complaints, the Commission has always tended to put national specialised authorities on the same level with ordinary courts; in our opinion, this is groundless and dangerous. However, in the White Paper, the Commission considers the difference that exists between both instances. Indeed, the problems encountered, and consequently the safeguards that may be offered, are not exactly the same; they depend on the authorities by which Article 81 is being applied, be they national authorities (1) or ordinary courts (2).

6 7

There are many studies on this issue. We have expressed our opinion in two articles. OECD (1998).

338

Working Paper III

1. Coherence and application by national competition authorities In order to guarantee that national competition authorities apply Community competition law uniformly, the Commission mainly proposes to create a network of authorities that would act in close co-operation and along common principles.8 In more technical passages of the White Paper, the Commission suggests various devices in order to resolve difficulties: • Removing obstacles to information exchange between authorities.9 • Requiring competition authorities to 'inform' the Commission.10 • Enabling the Commission to withdraw a case from a national authority.1' The starting point of such reasoning is undoubtedly good, and the idea to create a network is excellent. There are a few more technical suggestions made in the White Paper and they are more than welcome. The third device is the equivalent of Article 9 (3) of the present Regulation 17/62. It seems to be necessary to maintain coherence. On the second point, it will be necessary to indicate precisely the cases that have to be notified to the network. Clear rules on the allocation of cases between different authorities are also necessary. From that perspective it is clear that the obstacles resulting from the European Court of Justice's jurisprudence on the confidentiality of information12 have to be eliminated. It is completely unjustified to impose stricter limits on information exchange between the Commission and national authorities than exist in relations with the EFTA Surveillance Authority or even with US authorities. At this stage, we also have to draw consequences from a single market and a decentralised application of competition rules. Furthermore, it is essential that the new Regulation deal not only with the vertical relationship between the Commission and the national competition authorities, but also with the horizontal relationship. The Commission is working with the national authorities on these different issues and it is too early to make an appraisal of the results. From a general perspective we would offer a reminder that the institutional context as well as (1.1.) the procedural modes of enforcing community law should (1.2.) be taken into account as well.

8

White Paper, no. 91. White Paper, no. 96 to 98. 10 White Paper, no. 105. 11 White Paper, no. 105. 12 CJ, 16 July 1992, Case C-67/91 Direccion General de Defensa de la Competencia xl Association Espanola de Banca Privada (AEB) [1992] ECR 1-4785. 9

2 - Coherence

339

1.1. Reasoning on the institutional context Two points appear to be of importance in this context: (1.1.1.) the multiplicity of specialised authorities in the Member States, and (1.1.2.) their ability to apply EC law. 1.1.1. The multiplicity of specialised authorities in the Member States The idea of reinforcing the existing network between national authorities and the Commission does not give way to many remarks. However the Commission neglects an important administrative aspect. It is rare to be in front of one authority, as in the case of the Commission at Community level or the Bundeskartellamt in Germany. In most of the Member States, control is split between two specialised authorities.13 This is the case in France, for instance, where duties are divided among the Ministry of Economy (dealing mainly with inquiries) and the Competition Council (in charge of examination and decision-making). The same system exists in Belgium and Spain. In some of the Member States, the relations between the different authorities are sometimes rather strained.14 Furthermore, the preparatory work on the White Paper between the Commission and the national authorities already shows that, in some Member States, the most powerful authority tries to appropriate the reform in order to obtain new powers. 15 Let us take an example: which authority—the Ministry of Economy or the Independent Administrative Authority—will be entitled to adopt conditional decisions, when such a possibility does not already exist in the national system? Accordingly, it is essential for the Commission to be sure that all authorities in each Member State will effectively participate in the network and will exchange information between themselves. 1.1.2. The ability to apply EC law Entrusting specialised national authorities with the task of enforcing Article 81 means, first of all, that they must have the capacity to do so. This depends: (i) on their judicial abilities, and (ii) on their technical and practical abilities. i) The Commission is the first to notice that, for the time being, national authorities in seven Member States are not as yet in a position to apply Community competition law because of the judicial systems in which they operate. This had already been underlined in the Report on the application of Articles 85 and 86 prior to the adoption of the 1997 Commission Notice on co-operation between national competition authorities and the

13 14 15

This is a minimum. Some regulators may also apply competition law. The Belgian example is the most obvious, but it is not isolated. It is still very often the Ministry of Economy in spite of the generalisation of socalled 'Independent Administrative Authorities'.

340

Working Paper III

Commission in handling cases falling within the scope of Articles 85 or 86 of the EC Treaty.16 Facts have not changed a lot since then.17 The White Paper mentions that those seven Member States will have to enable their competition authorities to apply Community law, but it does not provide the judicial means to do so.18 It would be interesting to have more precise information on this, all the more because the principle of Community loyalty provided for by Article 10 has to be reconciled with the principle of institutional and procedural autonomy, established by the European Court of Justice. ii) Assuming we overcome this significant judicial obstacle, there still remain practical obstacles. The specialised national authorities do, a priori, have enough expertise to decide on whether to grant exemptions or not and they are the best placed to proceed therewith. They have the technical expertise to apply their national law as well as Community law. Whatever their technical aptitudes, there still remains a question mark regarding the practical means the national authorities have at their disposal. Exceptions to this are the German Bundeskartellamt and, to a lesser extent, the British OFT. Most of the remaining national competition authorities, and mainly those in the so-called Latin Member States (Italy excepted), complain about a lack of means to apply their national laws. It is not clear that the Member States will be ready to provide them with the means they will need in order to accomplish their new functions. Although it is understandable that such questions cannot be dealt with in a White Paper, it does not mean that they have to be left out altogether. On the other hand, it is hard to understand why the White Paper overlooks some relevant facts regarding the procedural tools that national authorities can use to apply EC law. 1.2. Reasoning on the procedural modes of enforcement of EC competition law The creation of the network of authorities is only afirststep. In our view, the true issues raised by the White Paper are the following: (1.2.1.) shouldn't procedural rules, at least to some minimal extent, be harmonised? (1.2.2.) shouldn't the territorial range of some of the national authorities' decisions be extended? 1.2.1. Minimal harmonisation of national procedural rules? The reform assumes that national authorities have equal powers as far as the application of Community law is concerned. The Commission abandoning its 16

OJC 313 of 15.10.1997. In the United Kingdom, the new Competition Act (1998) has not dealt with this issue, though one of the objectives of the reform was clearly to make British law close to the EC solution. 18 White Paper, no. 97. 17

2-Coherence

341

exemption monopoly reflects this view. Any national authority will be entitled to evaluate whether an agreement fulfils the conditions for exemption, and respectively to proceed at the withdrawal of the exemption. The White Paper only deals with parts of these questions.19 Indeed, it does not refer at all to the procedural framework within which these powers will be applied. The principle according to which national authorities apply Community law along with their national procedural rules is not being questioned. The principle of procedural autonomy still prevails. It is, however, essential to prolong the reasoning and to analyse whether such a decentralised and overall application of Community law will, without giving way to a harmonisation of all of the procedural rules, at least refer to identical principles when essential questions are at stake. Experience shows that differences that might be of importance do not concern procedural development stricto sensu. Indeed, all procedures have to comply with the fundamental requirements of Article 6 of the Convention on Human Rights because competition law (at least, law on anti-competitive practices) belongs to the category of quasi-repressive laws. This means that, as far as the rights of defence are concerned, more or less equal standards have to be ensured. Difficulties arise from the different margins of manoeuvre that national competition authorities have when starting procedures, or in the procedures' outcomes. i) It is important to know whether the national authority may decide itself which cases it is going to handle. Some authorities, such as the French Competition Council, may not choose and have to handle all actions that are not declared to be inadmissible. Of course the Council may address the case right away, but it is very often overwhelmed with work. This means that the French Competition Council is not always in a position to influence competition policy, which the Ministry of Economy itself wants to direct. Another problem arises in the details of the case submission, even though difficulties vary according to the practices aimed at, be they obviously prohibited practices (hard-core cartels or abuse of dominant position) or agreements likely to be exempted. In the first case (obviously prohibited practices), the most difficult thing is probably to disclose the infringement. In order to reach consistent results, it is essential to solve the problem of information exchange as exposed by the Commission. 20 However this is not sufficient: the problem has to be analysed along with leniency programmes. Imagine, for instance, a British undertaking providing the OFT with key information on a market-sharing agreement with a French company in order to benefit from a leniency programme and to avoid a fine. To what extent could the OFT subsequently 19 20

White Paper, nos. 94 and 95. White Paper, no. 97.

342

Working Paper III

pass such information on to the French Competition Council if the French law ignored the leniency device?21 The second case (agreements likely to be exempted) raises the possibility of maintaining notification systems when Community law and other national laws are based on a so-called legal exception system. Experience shows that a notification system at the Community level can work alongside a legal exception system at a national level; this was the case in France. It is not clear that the opposite is true as well. ii) In so far as the outcome of the procedure is concerned, the question regarding room to manoeuvre and freedom of appreciation is just as sharp. There is indeed a fundamental opposition between authorities that may negotiate and accept commitments, and authorities without any power of negotiation that may only make prohibitions once the infringement has been noticed. In France, the Competition Council falls into the second category. If national authorities are empowered to grant exemptions under Article 81 (3), it is important that all of them have the power to negotiate details of the contract with the parties, as the Commission can do at present. In our view, this is much more economically efficient. It would also be appropriate that authorities have similar or identical powers regarding prohibited practices. The importance of some behavioural commitments in cases of abuse of dominant position is obvious. Even if the only possible remedy is an injunction along with a fine, as in the case of hard-core cartels, there should be a minimum of common rules. It is completely abnormal for sanctions to vary significantly between different Member States, all the more because this influences the efficiency of competition policy. Such common rules could, for instance, affect the conditions under which a fine is imposed, mainly depending on the question of prescription or the rules on ceilings (or percentage of turnover made on the national territory). A differentiation should be made between the content of rules and their application. However the Commission forgets about the second factor. Even if the reform cannot be achieved in one single stage, it is necessary to think about the procedural context in which Community law will apply in the future. 1.2.2. Extension of the geographical impact of national decisions? Contrary to the Commission, the national authorities may only intervene within their national territory. This is the application in competition law of a traditional principle of international law: the territoriality of the power of constraint or imperium.22 These facts are not new but are not always taken into

21 In the present draft law under discussion, rules close to the EC ones have been introduced. 22

Idot (2000), forthcoming.

2 - Coherence

343

account. Later on, a distinction should be made according to the state of the proceedings, be it the progress of the proceedings or their outcome. i) Geographical obstacles during proceedings are not invalidating at all. They are indeed non-existent, or there are technical means for eliminating them. If an agreement might benefit from an exemption, the parties concerned will probably give the authorities all the useful information available, particularly as it is in their own interests to do so. It should indeed be remembered that undertakings do not a priori request the benefit of an exemption in a so-called 'legal exemption system'. Only when starting proceedings before the national authorities will they argue that such a practice is not prohibited. The burden of proof will influence their conduct. In that respect, it will be important to introduce rules on the burden of proof into the Regulation. For the application of the prohibition of Article 81 (1), the competition authorities bear the burden of proof. The situation is reversed in the application of Article 81 (3), where undertakings have to prove that their agreement fulfils the conditions. In reality, the principle of territoriality might only be a problem for the controlling authority in cases involving prohibited practices with undertakings that refuse to co-operate. Technical means do exist, yet they have been developed between competition authorities in international co-operation agreements. Those are more precisely agreements on information exchange but could be developed in the future network as the Commission suggests. Transfer of files and clear allocation of cases is necessary. One could even go further by introducing, for instance, a principle of positive comity between national authorities in the application of Community law. In any event, the Commission may handle the case itself, as it has done in the past. The suggestions of the White Paper point in the right direction.23 ii) There is an even bigger problem regarding the geographical range of a national decision. This question has already been raised and we will have to trace the true difficulties. When only national authorities were able to make prohibitions, the system worked quite properly. It mainly functioned well because national authorities generally intervened in practices located on their territory and arising from national undertakings.24 Facts change totally with the 'right to grant and withdraw' exemption being transferred to the jurisdiction of the national authorities. This difficulty has been mentioned in the recent draft Regulation on vertical restraints, where the national authority may withdraw the benefit of exemption but the withdrawal is only valid on the national territory.25 23 24 25

See supra n no. 9, 10. Ehlermann (2000), no. 103, p. 38. Commission Regulation (EC) no. 2790/1999 on the application of Article 81 (3) of the Treaty to categories of vertical agreements and concerted practices, 22.12.1999, OJ L 336 of 29.12.1999.

344

Working Paper III

Should the same solution apply to a national decision, noticing that the conditions of Article 81 (3) EC Treaty are fulfilled? What if an exemption is granted only upon the fulfilment of some conditions, or if the proceeding has ended pursuant to negotiations that led to commitments? Given the situation, there is no mechanism that would confer a broader scope than the national territorial one on these types of decisions, all the more that they are not even sure to impose upon their own national courts. The exequatur procedure for the recognition and enforcement of foreign judgments does not have any equivalent in the relations between administrative authorities. The question is whether the national decision should have an EU-wide effect or not. The debate on this point is intense.26 There seems to be a contradiction between the geographically limited effect of the national authorities' decisions and the fact that agreements cover the entire Community market. For the moment, however, the Commission seems to deny any EU-wide effect of national decisions. There is no such difficulty when considering the problem of coherence in the application of Community law by ordinary courts.

2. Coherence and application by national courts Under the current system, national courts may have to draw consequence from the infringement of the Article 81 (1) and 82 prohibitions. Accordingly, the reform does not have any effect on pleas for damages, which are moreover not very numerous; nor should it have any consequences on actions brought before an administrative judge, when available, as in France. On the other hand, changing over to a legal exemption system might noticeably modify contractual litigation, because judges will have to integrate Article 81 (3) EC Treaty in their reasoning. As far as the judicial prerequisite is concerned, we can wonder whether Article 81 (3) may be considered to be of direct effect?27 Accordingly only technical procedures that confer a minimum of coherence will be taken into account. The judge of the contract is not only the state judge but also an arbitrator. Any search for coherence has to take this fundamental fact into account; it is a fact that the Commission completely ignores. 2.1.

Coherent application of EC competition rules by national courts

First of all we must note that, even if facts may vary considerably from one country to another, the passage to the legal exception system re-launches the 26 27

Ehlermann (2000), no. 104, p. 39. See also, Ehlermann (2000), no. 46 and sub., p. 20.

2 - Coherence

345

debate on the part played by ordinary courts in the application of competition rules. Here, the Commission gives an established affirmative answer because it apparently does not wish to enter a debate on the qualifications of judges. However the French experience in this field leads to the following two observations: First, we do not have any hindsight about the way in which ordinary courts apply the economic balance under Article 10 (2) of the ordinance (in most cases, at first instance by commercial courts). Second, we start having some doubts regarding judges' abilities to apply technical matters such as competition law. In the draft law that is currently being discussed, appointed courts would be in charge of litigation in competition related matters. However it is possible to pick out the following elements that are in favour of the reform. Not only is application of Articles 81(1) and 82 rather difficult in itself, it is not certain that the application of Article 81 (3) is more difficult regarding the important decision-making practice. Moreover, a lot of contracts, especially in thefieldof distribution, will be covered by block exemption regulations that should considerably ease the task for judges. In the White Paper, the Commission only concentrates on preventive means, that is, those that will help the judge to solve competition issues. The Commission does not analyse the questions of international private law. 2.1.1. Support in solving competition issues The Commission proposes three types of mechanisms to help the national courts solve complex competition cases and thus guarantee the coherence of national decisions. Those mechanisms concern either the European Court of Justice, the Commission itself, or national competition authorities. The first mechanism is well known: it is the application of preliminary rulings under Article 234. Even though we do not want to challenge a method provided for in the Treaty itself, nor should we overestimate its results. A major reform of the European Court of Justice is under way. There is a debate about whether to limit the number of national courts that will be entitled to request preliminary rulings. The mechanism would be less efficient if courts of first instance could no longer question the European Court of Justice. Time is of essence in thefieldof competition, and the Commission is the first to acknowledge this.28 This is why it suggests other methods in order to guarantee a minimum of uniformity.29 The logic is the same, that is, enable the national judge to benefit from the technical expertise of a specialised authority, be it the expertise of the Commission or of one of the national authorities. i) The Commission makes the following three suggestions with respect to its role: to impose an obligation upon national judges to inform the 28 29

White Paper, no. 104. White Paper, no. 107.

346

Working Paper III

Commission as soon as a case deals with an issue of Community competition law; to add the co-operation mechanism provided for by the Notice of 13 February 199330 to the future Regulation; and to intervene as amicus curiae in pending matters. These proposals have different levels of impact. The obligation to inform does not raise major objections. However, it is obvious that it does not have any impact in itself. The obligation to inform only makes sense if the Commission subsequently uses the information effectively. The Commission already complains of being overwhelmed with notifications; in the future, the Commission might be overwhelmed with . information requests from national courts. The co-operation procedure of February 1993 and the newly suggested procedure of amicus curiae raise likely problems regarding their impact and nature. Experience has already shown that the procedure of February 1993 has not been used very often. The parties dread it because of its lack of transparency, its lack of supervision regarding rights of defence, because of the principle of contradiction, etc. If we want the new procedure to be more successful, the mentioned deficiencies need to be compensated in the future regulation, and equivalent guarantees need to be anticipated for the amicus curiae. However this new mechanism of amicus curiae raises an even bigger problem. The main difference between both procedures seems to lie in the fact that, in the first case the judge requests the Commission's advice, whereas in the second it is the Commission that decides to intervene, ii) Finally, the Commission proposes to use the same co-operation mechanisms that exist on the national level31 under French law:32 that is, the request for advice from the Competition Council. From a theoretical point of view, this indeed corresponds to the logic of a decentralised application system of Community law. From a practical point of view, it might not be that efficient having regard to the long deadlines and the increasing workload national authorities have to face. The reluctance of national courts to use this device must also be taken into account.33 2.1.2. Issues of international private law Contrary to some observers, I do not think that it is useful to claim for an approximation of rules of judicial procedure. It is too much controversial. The only exception concerns issues of international private law. These seem to be ignored by the Commission, which always refers to the Brussels Convention of September 1968 on international competence, recognition and enforcement of judgments to solve the difficulties. In my opinion, this is not sufficient. 30

European Commission: Notice on cooperation between national courts and the Commission in applying Articles 85 and 86 of the E E C Treaty, OJ C 39 of 13.2.1993. 31 White Paper, no. 107, in fine. 32 Ordinance 86-1243, 1 December 1986, art. 26. 33 In France, only penal courts have used this device. No commercial courts ask the Council to give its advice.

2 - Coherence

347

i) As far as the competence of courts is concerned, there is a serious risk of forum shopping. It is true that the rules have been harmonised within the Community and in relation with the EFTA countries by the Brussels and Lugano Conventions. Nevertheless, all specialists in international private law will confirm that Article 5 (1) of the Brussels Convention, which enacts an option on jurisdiction in contractual matters, raises a lot of debates. It is so complicated that courts of first instance interpreted the text in many various ways, despite case law of the European Court of Justice that did not clarify its meaning and introduced a conflict of law approach. A three-step approach is necessary. First, is it a contractual matter? Second, what is the obligation that has to be taken into account? Third, where is this obligation localised according to the law applicable to the contract? The Regulation proposal that will take precedence over the provisions of the Brussels Convention is not written in the simplest way. The simplest solution would be to suppress this option and to return to the general rule of Article 2 {'actor requiturforum ref). ii) The issue of the EU-wide effect of judgments of national courts might be rather easily solved by the application of the rules of the Brussels Convention. A definitive judgment in a competition case will probably benefit rather quickly from exequatur in other Member States of the European Union. The mechanism of enforcement is effective and works well. Reasoning only about relations between decisions of national courts is not sufficient. No analysis has been made of the relations between judgments of the courts and administrative decisions taken by the national competition authorities. This question existed prior to the reform, but it will have to be dealt with more precisely in the future. 2.2.

Coherent application of EC competition rules by arbitrators

In this matter, the White Paper overlooks some serious points. The Commission ignores an important reality in business litigation. In contracts, and mainly in cases involving significant contracts that are rarely covered by block exemption regulations, most actions are submitted to arbitrators. This is possible because the fact that litigation can be arbitrated in competition matters is currently admitted in all Member States.34 As judge of the contract, the arbitrator has the same powers as a national judge. If a judge has the power to apply Article 81 (3), the arbitrator shall have the same power. However there are no measures meant to guarantee the coherence of the application of EC competition law by arbitrators. In compliance with the Nordsee decision,35 arbitrators cannot ask the European Court of Justice for 34 35

Idot(1996)p. 563. Case 102/85 Nordsee [1982] ECR-1095.

348

Working Paper III

preliminary rulings, nor do they have the ability to request technical expertise from competition authorities. The obligation imposed on arbitrators is not accompanied by any procedural measure, even though the Court of Justice judged in the EcoSwiss case36 that rulings have to comply with EC competition rules (that refer, of course, to public order). Without challenging the arbitrability of competition matters, it might be useful to think this question over.

II. Coherence and parallel application of EC and national laws There is another main fact that has to be considered. The national authority or court will rarely apply anything but EC competition law, but its application will go along with national law. This is not new, and in its White Paper the Commission reminds us that parallel application of EC and national competition law has existed ever since 1962. Then it notes that that this has raised few problems.37 This conclusion is somewhat quick (1) and it is about time to re-evaluate relations between Community and national laws (2).

1. A need for more consistent reasoning The Commission has made a quite optimistic analysis of the need for more consistent reasoning: on one hand it over-estimates the coherence of Community law as applied so far; on the other hand it underestimates the risk of 'renationalisation' of Community law.

1.1. Overestimating the past application of EC competition rules At first glance, the application of EC competition rules at the national level might be assessed as satisfactory. However the results need to be analysed in different ways, having regard to whether competition authorities or national courts make that application. We will take the French situation as an example. i) The parallel application of EC and French law by the Competition Council did not pose any major problems. This was mainly because EC rules were applied in most of the cases to prohibited practices.38 36

Case 126/97 Eco Swiss [1999] ECR 1-3055, see also Rev. arbitrage, 1999/3, note by L. Idot. 37 White Paper, no. 102. 38 Every year, in its annual report, the French Competition Council makes some comments o n the application o f EC law. The decisions can also be found in this annual report.

2 - Coherence

349

Prosecuted undertakings also referred to the fact that the Commission had been notified of the agreement in order to defeat the application of French law. In such cases, the notification was evidently intended to cause delay,39 and the Council justly prohibited the agreement right away: when in doubt, the Council waited for the Commission's position before applying national law.40 In reality, the problems lie upstream. A thorough analysis would indeed reveal that the French authority interprets the condition of 'affecting trade between Member States' in a somewhat restrictive manner.41 Moreover, from a procedural point of view, competition authorities do not have any obligation to apply Community law, even though the rapporteurs in charge of the case would, in practice, proceed with verification, ii) There are only few examples of parallel application by national courts, because of the kind of litigation submitted to them. Most actions before civil or commercial courts—either contractual or delictual—deal with the application of block exemption regulations when there is an allegation of unfair competition. There is no discussion of national law since the principle of the supremacy of EC law is well accepted. The same remark can be made for actions before administrative courts because it is easier to invoke EC law than French law to deny the validity of a decree or administrative contract. Furthermore, there are fewer risks because of the procedural rules and the duty to apply EC competition law.42 The difficulties in interpreting the condition of 'affecting trade between Member States' are however most probably the same. A notice to summarise the principles that have to be followed to check this jurisdictional prerequisite would be useful. 1.2. Underestimating future risks of 're-nationalisation of EC law All solutions given by the Commission in the White Paper assume that national authorities or courts apply their own national law, and the Commission will intervene at the Community level due to its exclusive jurisdiction to grant an individual exemption under Article 81 (3). However, the facts will be quite different under a legal exemption system where national authorities will have to evaluate agreements in light of their national laws, as well as in the light of EC 39

See for instance the Facam case in 1992 upheld by Paris, 3 March 1993, B O C C R F no. 6, 26 March 1993. 40 See for instance the Quaniel case; decision 91 D. 53 and decision 93 D.I 1. 41 See for instance the Aeroport de Paris case, which leads to a decision of the Tribunal des Conflits, 15 October 1999. E C law obviously was applicable because foreign airtransport companies complained about a restrictive agreement between Air France and ADP in relation with the new organisation of Orly Ouest. Only French law was applied in this affair. 42 Joint Cases C 430/93 and C 431/93 Van Schjindel [1995] ECR 1-4705.

350

Working Paper III

law and especially in light of Article 81 (3), in case a block exemption regulation is not applicable. The difficulties are minor in the present system because there is a mainly convergent approach towards prohibited conduct within the EU. The issues are mainly upstream and concern the applicability of competition law, such as the debate on the submission of public authorities to competition law in France. In the future, undertakings might try to enlarge the application of exemptions. The application of Article 81 (3) is likely to be more diverse than the application of prohibitions in Articles 81 (1) and 82. First of all, the national authority or court might be tempted to generalise a 'rule of reason' approach to the application of Article 81 (1), and to save on further reasoning if the agreement is globally good. The French example shows that there is indeed such a risk.43 Second, it might also be tempting to add additional grounds for exemption to those provided for in EC law, and to distort the traditional application of Article 81 (3) by, for instance, taking into account social questions,44 the problem of crisis agreements, etc. Would the future network and the Commission's supervision be effective enough to eliminate the risk of dilution? This is far from certain. There is probably a lesser risk as far as national courts are concerned. They will be less tempted than specialised authorities to 're-nationalise' EC law. In both cases, the Commission should be cautious, and precise guidelines on the application of Article 81 (3) would be useful. This does not exclude the possibility that other solutions could also be sought.

2. A need to re-think the relations between EC and national law As far as merger control is concerned, the 'one-stop shop' system is a satisfactory solution. It might be interesting to generalise this system to antitrust law. Presently Italy is the only Member State to alternatively apply Articles 81 (1) and 82 or Italian law, whereas all other Member States resort to cumulative application when interstate trade is affected. The latter approach does not make a lot of sense: in case of conflict, the principle of primacy of Community law applies and the Community solution prevails. Cumulative application is even less justified when Article 81 is applied in its entirety and in a decentralised way at the national level. The Italian system leads to far more coherent solutions and avoids any risk of re-nationalising EC law. The solution is not so difficult to reach since it does not require any modification of the Treaty. The only thing that might have to be modified is the 43

Blaise (1997). In the French draft law under discussion, social purposes have been added in the text of Article 10 (2), which is the copy of Article 81 § 3. 44

2 - Coherence

351

interpretation of the condition that inter-state trade be affected. So far, the same rules apply to both this jurisdictional prerequisite and to the substantive condition that is the restraint of trade. It is a source of confusion in many minds. In such a system, there must be a clear-cut division, such as the 'Community dimension' criterion in merger control. In our opinion, we will have to continue in this difficult direction if we want to maintain the coherence of the community system. In conclusion, it is clear that a reform was necessary and the White Paper is a first step in the right direction. A lot of things remain to be done. The choice is now clear. Do we make this first step without waiting for any other improvements (harmonisation of minimum procedural standards, new relations between EC and national law, etc.) because it is necessary to begin the reform as soon as possible, or do we have to work for years to improve the present system and thus risk abortion?

IV Ulrich Immenga Coherence: A Sacrifice of Decentralisation?

The White Paper proposes the decentralised enforcement by national competition authorities and courts of Article 81(3), allowing exemptions from the general rule forbidding cartels under Article 81(1). The merits of such decentralisation are obvious. Decentralisation corresponds with the principle of subsidiarity (Article 5 (2)), brings enforcement competencies closer to the markets relevant in competition cases, and diminishes the workload of the Commission, which at present has to apply a system of authorisation. Most important, however, is the fact that national bodies will be empowered to apply the Article 81 (1) prohibition along with the exemption rule. Such joint application has been requested by national competition authorities for a long time. The practical effect of the new distribution of competencies is that undertakings will no longer be able to invoke an exemption in Brussels when they are attacked nationally under the prohibition rule. There are, however, serious concerns about detrimental consequences of decentralisation, particularly with regard to coherence. This paper focuses on the principal points of the debate on coherence, and is structured as follows: —Requirements for effective application by competition authorities and courts (Section I) —Allocation of cases (Section II) —Uniform application of Community competition law (Section III) —Decentralisation and the Commission's competencies (Section IV) —Conclusions (Section V).

I. Requirements for effective application There is a general concern that national competition authorities and courts will not really be suited to apply Community competition law, especially the Article 81 (3) exemptions from the general prohibition on horizontal restrictive agreements in Article 81 (1). National competition authorities must, first, have the express power to apply Community competition law. Yet the competition authorities of several Member States still lack this power. Furthermore, they must have the power to imposefines,they must dispose of effective procedures, and they must also have sufficient independence, professionalism, and objectivity. There are already some doubts about whether these requirements can be achieved within a reasonable period of time, doubts that increase in light of the

354

Working Paper IV

upcoming eastward enlargement of the European Union. In addition, these requirements can be fulfilled only with a minimum degree of harmonisation of procedural rules, a harmonisation that will be particularly necessary in order to achieve uniform application of Community competition law. In the context of the necessary independence of national competition authorities, an issue that arises is the general understanding of the role of competition policy in each of the Member States. The role of competition policy is still quite differently understood in the various Member States, though a certain degree of approximation in this respect can nevertheless be observed. The White Paper sees several benefits in the enforcement of Community competition law by national courts: these are close to the European citizens (no. 99), and undertakings will no longer be able to invoke Article 81 (3) in a directly applicable exemption system (no. 100). As a result, the availability of Article 81 (3) will reduce delays in court proceedings. However, these 'advantages' can be questioned in the light of the 'countervailing' arguments. The most important task of courts lies in their involvement in commercial disputes between private parties. The validity of contracts with restrictive elements is at stake. Numerous courts of first instance can be involved in such matters, but are these courts really suited to decide whether or not contractual clauses fall under Article 81 (1) or (3)? A lack of experience in applying Community competition law should be assumed. Basically the same considerations apply to the function of judicial review of the decisions of national competition authorities or lower courts. There is an argument that national courts are incapable of applying Article 81 (3) adequately since, as expressed by the European Court of Justice, the application of this provision requires complex economic analysis. The Commission has had a significant margin of discretion in applying this provision in the past. It does not seem appropriate to leave the same margin of discretion to national courts. Specialised courts might be the answer to such concerns. To date Germany has had favourable experiences with such courts. The Kammergericht in Berlin had a special chamber hearing appeals from the Bundeskartellamt. The same system continued to function after the Bundeskartellamt moved to Bonn, though competence was transferred to the Oberlandesgericht in Diisseldorf. Furthermore, and probably more importantly, in Germany there is the formal legal ability to concentrate the competition cases of several districts in one court. Thus court specialisation is already is a reality in Germany. Another of the Commission's proposals is the institution of a Community Law Adviser for each Member State. The opinions of such Adviser would not be binding on the national courts. The Adviser would stand in national court proceedings as amicus curiae, or a kind of national Advocate-General on EC law issues, and thus the knowledge and acceptance of Community competition law would thereby be increased. This should be considered as a point for further discussion. Another point is substantive and crucial: Article 81 (3) is not exclusively competition-oriented. In its 23rd Annual Report on Competition Policy (no.

2 - Coherence

355

190), the Commission declared that the application of Article 81 (3) requires the assessment of complex economic elements and involves the Commission's discretion if different objectives of the EC Treaty have to be considered. For instance, the exemption granted by the Commission to the co-operation agreement between Ford and Volkswagen regarding the production of vans in Portugal was based inter alia on EU regional policy considerations (OJ L 20, 28.1.1993). Furthermore, the application of Article 81 (3) requires a weighing of the benefits brought about by a restrictive agreement in terms of 'technical and economic progress', which is an industrial policy rather than a competition policy consideration. The specific policy function of Article 81 (3) is the very reason why it would be wise to preserve exclusive competence for its application at the Commission. National competition authorities cannot handle EC policy considerations, and this is furthermore true for national courts. Therefore, decentralisation requires a new understanding of the exemption rule. This is not only an issue of efficiency, but of uniform application as well.

II. The allocation issue, or hazards of forum shopping The White Paper contains no express statement as to the territorial reach of national decisions. Of course, court judgments only can have inter panes effects, and will therefore not bind third parties. But what about decisions of national competition authorities based on Article 81 (3)? It might be concluded from the Commission's references to the allocation issue (White Paper no. 95 seq.) that the jurisdiction of each national competition authority would be confined to one Member State. This suggestion is very important and must be discussed in the context of the potential problems it could give rise to. The allocation issue arises every time a case concerns companies of different Member States or has trans-border competitive impact. According to the White Paper (nos. 87, 88; see discussion below IV), the Commission still claims the right to assume jurisdiction in certain individual cases and under specific circumstances. However, the rule is decentralisation. Therefore, in order to avoid forum shopping to the largest extent possible, it is necessary to determine the competent competition authority or court. The White Paper is silent on this point. It contains relevant remarks only about the allocation of cases between the Commission and national competition authorities (no. 96, 97). National competition authorities may call upon the Commission if a complaint requires investigation in several Member States, or where the Commission could pass a case to the authority best placed to deal with a case since the effects of the agreement are essentially felt in a single state. But what are the criteria? The allocation issue is most intense for court proceedings. To the extent that there is an inconsistent application of the Community competition law, this will foster forum shopping. There is also the risk of multiple litigation, that is,

356

Working Paper IV

litigation concerning the same agreement or network of agreements in several Member States. The risks are particularly high with regard to pan-European distribution systems. In cases of complaints to national competition authorities there is a risk of forum shopping also when the facts and the effects of the contested agreement involve several Member States. Furthermore, complainants do not necessarily know whether companies in more than one Member State are involved, or where the relevant evidence is situated. If a complaint is made to one national competition authority, that authority might discover that another authority from a different Member State is already investigating the same agreement or practice, or that some important evidence is in another jurisdiction. These situations make the transfer cases between authorities inevitable in a decentralised system. Particularly in cases involving an implicit exchange of information that might include confidential elements, such transfers are not desirable. There is no doubt about the necessity to develop a clear system of case allocation that ensures legal certainty and allows aflexibleco-operation between the members of the network but that avoids forum shopping and the transfer of cases. The White Paper does not present proposals for such a system. Non-legally binding guidelines might provide appropriate solutions to reassure potential complainants and to avoid multiple proceedings. 'Centre of gravity' criteria might be an adequate option but would need more specific descriptions of the parties and markets concerned. In addition, any guideline must be legally enforceable in order to make the exchange of information between national competition authorities possible.

III. Uniform application of Community competition law Coherence in the application of Community competition law is essential in a single market. This is particularly true with regard to exemptions from a general rule against restrictive agreements or cartels. Competition policy requires distinct lines to distinguish forbidden practices from permitted ones. Thus, in addition to group exemptions, the application of Article 81 (3) plays a decisive role. Coherence is ensured under the present system with its reference to the exclusive competence of the Commission. By its nature, decentralisation requires instruments to achieve as much coherence as possible. The high risks of divergent interpretations or even conflicting decisions in a single case have to be overcome. The White Paper is aware of potential inconsistencies. However, it refers to well-known principles for the resolution of conflicts arising between the Commission and national competition authorities or courts (no. 102, 103). These are principles enforcing the supremacy of Community law. It has been accepted, however, that it would nevertheless be advisable to establish mechanisms to avoid conflicts between different national bodies applying Community

2 - Coherence

357

competition law (no. 103). The costly and cumbersome instrument of Article 234 (previously Article 177) requires courts of last instance to refer questions of interpretation of Community law to the Court of Justice for preliminary rulings. In addition, the need for flexible and rapid mechanisms for the exchange of information and co-operation between competition authorities, courts and the Commission has been acknowledged (White Paper no. 104). The White Paper proposes an obligation to inform the Commission of the initiation of Article 81 and 82 proceedings in national competition authorities. However, the possibility that the Commission could take a case out of the national competition authority should be preserved (no. 105). Furthermore, the role of the Advisory Committee on Restrictive Practices and Dominant Positions should be reinforced, and made into a full-scale discussion forum (no. 106). To achieve consistency, national courts should be obliged (as competition authorities) to 'inform' about proceedings conducted before them in which Arts. 81 and 82 are invoked (no. 107). Interestingly enough, both obligations already exist under German law. Paragraph 50 (3) of the Restriction of Competition Act requires the competition authority to provide such information; the requirements for courts are more limited, having to inform the Bundeskartellamt only of cases in which Community law is applied (Paragraph 96 with reference to Paragraph 90). Finally, in this context, the White Paper suggests allowing the Commission to intervene as amicus curiae, subject to the leave of the court (no. 107). Again, these considerations concern only the relationship between the Commission and national competition authorities, and they neglect the need for co-operation and the exchange of information between the national authorities themselves. To involve the Commission might be important. It is as important, however, to co-operate with other national authorities which might be involved. The exchange of information should cover the main features of the case, the evidence collected and, later in the proceedings, intentions as to how the case will be decided. Beyond these considerations there are also serious objections and concerns at different levels. A substantive objection (as earlier mentioned) is the specific character of Article 81 (3), whose application requires the assessments of different EC policy objectives; this is a task that cannot be undertaken by national bodies. Furthermore, a re-nationalisation of Community competition law must be feared because national authorities will take national interests into account. In this context, it might be argued that national authorities might be less independent than the Commission and more sensitive to political pressure. From the point of view of some German commentators, obstacles to coherence seem inevitable if two characteristics of the procedural rules in private law proceedings are considered. First, fact-finding is inherently limited because the judgment has to be based only on that evidence that was provided by the parties. Not only is this element an obstacle to coherence, but also to the application of Article 81 (3). Second, the burden of proof (that the facts satisfy this

358

Working Paper IV

provision) is placed on the plaintiff. This burden will, in practice, seldom lead to a positive judgement. Several proposals have been made about how to preserve coherence. One suggestion is to give national decisions Community-wide effect. At least two concerns arise in respect to this suggestion. The risk of multiple and conflictive decisions will not be avoided, and variances between the reputations of national authorities might not engender respect in particular decisions. Others suggest a mechanism to ensure that an exemption granted by one national competition authority would be mutually recognised by all other authorities throughout the Community. Yet another proposal is the establishment of a decentralised system of Community courts in each Member State to apply Community competition law. And there is also the idea of facilitating the procedure of referring questions of interpretation of Community law to the Court of Justice (which is particularly experienced in competition matters, for preliminary ruling). In my view, besides the establishment of a system of co-operation and exchange of information, the most relevant suggestion is the proposal of conditional Community-wide effect. National decisions based on Article 81 (3) that would become applicable can be opposed by the Commission and national competition authorities within a reasonable period of time. This applies to conflicting decisions as well. If no objections were raised, the decision would automatically acquire Community-wide effect. Appeals from national competition authorities should be decided by the Commission. It has to be seen, however, that even such a procedure does not provide for coherence in court proceedings. Consequently and with regard to other concerns mentioned here, it hardly seems possible to have both uniform application and the direct applicability of Article 81 (3).

IV. Decentralisation and the Commission's enforcement competencies According to the White Paper, the Commission reserves several competencies that do not appear to be very clear. This is an issue of the relationship between the Community and national competencies. This issue is not necessarily one of coherence but it is nevertheless important with regard to subsidiarity and transparency. Only a few points should be taken into consideration in this context. Where its competencies are not exclusive, subsidiarity requires that the Community takes action only if the respective objectives cannot be achieved appropriately on the national level. This approach requires discussion if the Commission has the unspecified ability to take a case out of the jurisdiction of a national competition authority by means of a mechanism equivalent to that in Article 9 (3) of the present Regulation 17/62 (no. 105). The same concern applies to the White Paper's remark that the Commission would be concentrat-

2 - Coherence

359

ing its attention on those restrictive agreements with the most important effects (no. 87). It also applies to the Commission's claimed right to intervene and prohibit an agreement between undertakings that was approved via a positive decision by a national authority, even if the decision is no longer be open to appeal and the principle of res judicata between the parties of the dispute is reserved (no. 102(2)). These proposals grant wide discretion to the Commission to intervene (or not) in a given case, and even the possibility of correcting decisions of national authorities. Such competencies hardly satisfy the criteria of subsidiarity and, moreover, introduce a high degree of uncertainty into a decentralised system.

V. Conclusions Decentralisation requires an absolutely clear notion of the scope of Article 81 (3). This exemption provision should be interpreted in the light of competition objectives and without regard to other objectives of the Community. This is the requirement for the direct applicability of this provision, for its effective applicability and, particularly, for its coherent application throughout the Community. Clear guidelines setting out the relationship between national bodies (competition authorities and courts) are indispensable for an appropriate allocation of cases and for coherence. The White Paper does not contain substantial proposals in this sense, and refers only to the relation between the Commission and national bodies. With regard to national courts, coherence in enforcement can be achieved only if a restricted number of specialised courts in each Member State handle competition matters. The Commission may act as amicus curiae in such cases. Furthermore, in order to avoid forum shopping and to secure an effective decentralised application of EC competition law, a certain harmonisation of national procedural rules, enforcement, and remedies has to be initiated. The Community-wide effect of decisions by national authorities will serve as an important factor to preserve coherence. This is a logical consequence of the application of Community law and there are several means by which to achieve this result. According to the White Paper, the Commission claims wide discretion to assume jurisdiction over individual cases. This is incompatible with decentralisation. The principles of subsidiarity and transparency require guidelines delimiting the circumstances in which it is likely that the Commission will claim jurisdiction.

V Frederic Jenny On the Modernisation of EC Antitrust Policy

It should be noted at the outset that the decentralisation of the enforcement of Article 81 must be looked at in conjunction with other measures, such as: • the elimination of the notification system; • the recognition of the fact that vertical agreements implemented by manufacturers having a small market share do not usually raise a competition problem, and can therefore benefit from an a priori exemption (if they do not include any of the black clauses); and • the strengthening of the investigatory powers of the Commission. The question of coherence is raised in the context of the decentralisation of the enforcement of Article 81 because of a fear that different national competition authorities or courts may make contradictory assessments as to whether a particular agreement constitutes a violation of Article 81. A frequently expressed point of view is that, in a single market, the same rules should be applied and they should be enforced consistently.

I. On the benefits and the costs of decentralising the enforcement of Article 81 Before discussing whether the decentralisation of the enforcement of Article 81 will create a serious risk of legal uncertainty, it is worth mentioning that the previous system had various well-known drawbacks. First, as several commentators have argued, the previous system of notification did not create legal certainty for firms because the Commission could not examine all notified agreements due to a lack of resources and it therefore issued comfort letters which did not have legal weight. It also appears that firms often used the notification system not so as to discover whether their agreements were legal (since they knew that the Commission was too busy to make decisions on notified agreements), but rather so as to discourage their co-contractors from trying to challenge an agreement by arguing that it violated EC law.'

1

Forrester I. (2000): 'Modernization of EU competition law', 23 Fordham Intl. L.J. 4, pp. 1028-1088.

362

Working Paper V

In addition, the block exemption for exclusive vertical arrangements treated all agreements similarly, irrespective of the structure of the industry in which they were implemented. This is not in line with economic reasoning. Finally, because of the importance of the resources devoted to examining and answering notifications of vertical agreements, DG COMP could not concentrate on fighting serious competition violations (which were never 'notified'): that is, horizontal hard core cartels. Thus it is worth remembering that the previous system imposed a substantial cost on society. In light of this, the relevant question is not whether the proposed decentralisation of the enforcement of Article 81 will lead to a risk of inconsistent enforcement of EC law. Rather it is whether it is likely that the benefits of the reform will outweigh the social costs imposed by decentralising the enforcement of Article 81. In other words, if decentralising the enforcement of Article 81 is necessary in order to re-direct the investigatory means of the Commission towards a fight against transnational horizontal cartels or abuses of dominant positions, it might generate social benefits that far outweigh the cost of occasionally inconsistent decisions by national courts or jurisdictions in the enforcement of Article 81. This reasoning would not hold, of course, if the interpretation of Article 81 by national courts or jurisdictions were to be so frequently inconsistent that its enforcement became unpredictable. However occasional divergences by national courts or competition authorities do not mean that it would no longer be possible to predict whether a particular practice would be likely (with a fairly high degree of probability) to be found to violate Article 81. Thus the mere logical possibility of inconsistent enforcement is insufficient to settle the issue of whether or not the proposals of the Commission are acceptable or not. When one considers the instruments which could reduce the legal uncertainty feared by firms in the event of a decentralisation of the enforcement of Article 81, one also has to consider both the (marginal) costs and the benefits of these instruments. Yet such cost-benefit analysis has hardly ever been done. The discussion of such instruments is usually limited to the technical question of whether they would allow the Commission to exercise better control over national courts. Many instruments have been proposed. Some have suggested that an option to notify the Commission of agreements (with the obligation for the Commission to react within a short period of time) must be retained. It has also been suggested that: • production joint ventures should remain under an ex ante control; • the Commission, when acting in the context of ex officio proceedings, should be able to adopt formal non-infringement decisions in cases where it is in the Community interest to do so; • the Commission should be informed of all cases in which Community competition rules are invoked;

2 - Coherence

363

• the Commission should have the right to intervene as amicus curiae or to appeal a decision in the interest of the Community, etc. Most of these instruments would undoubtedly decrease the risk of inconsistent enforcement of EC law. But, taken together, these instruments would also deprive the Commission of a significant part of the expected benefit of the reform: that is, the ability to increase its effort to eliminate horizontal hard core violations of EC law. Indeed, the use of these instruments would be quite cost intensive.

II. Is inconsistent enforcement of Article 81a major risk? First, it should be noted that, under the current interpretation given by the ECJ to Article 81, applying 81 (3) to an agreement involves an assessment of the possible redeeming values of the agreement, both in terms of its procompetitive impact and in terms of its contribution to technical or economic progress (as long as the benefits are to a certain extent passed on to the consumers). Second, it should be noted that hard core horizontal agreements (e.g. price fixing or market sharing among competitors) very rarely have any redeeming value and rarely contribute to economic progress. There is a great deal of consensus on this among competition authorities and courts and, as far as it is known, there are very few indications that national courts or competition authorities would easily exempt hard core horizontal anti-competitive practices. It is certainly more difficult to assess the pro- and anti-competitive impacts of a full-function joint venture, but this category of practice will continue to be notifiable under the Merger Regulation. The Commission has also suggested that production joint ventures should remain under ex ante control because of the extensive nature of the commitment they imply. Some authors have argued that other (horizontal) agreements involving major commitments should also be notifiable. Before venturing an opinion on this issue, it would be useful to know precisely what kind(s) of agreements (that are neither full function joint ventures nor production joint ventures) these authors have in mind. However, it is clear that the larger the scope of notifiable agreements, the smaller the benefits of the reform will be (because of the social cost of handling these notifications). Third, there is very little risk that a national court or national jurisdiction will fail to recognise that a vertical distribution arrangement entails a lessening of intra-brand competition. Thus, under the current interpretation of Article 81, the applicability of 81 (1) to such agreements will easily be recognised by national courts or jurisdictions. There are, however, two possible risks.

364

Working Paper V

The first is that national competition authorities or courts will err in assessing whether or not an agreement contributes to the development of inter-brand competition (that is, it contributes to improving the production or distribution of goods). However, it is very hard to see why one should presume that courts or national authorities will err in this assessment of the competitive impact of an agreement. National courts and/or competition authorities are reasonably competent to assess the impact of practices on competition; otherwise national courts and jurisdiction would be considered unable to assess the conditions of competition on a market and the Commission would be the sole enforcer of Articles 81 (1), 81 (3) and 86. Indeed, a competitive impact analysis is required to enforce all of these provisions. Since no one has argued that national courts or authorities should be deprived of the power to enforce 81(1) and 86, we must presume that the risk of misappraisal of the competitive impact of a practice or an agreement under 81 (3) by national courts or authorities is sufficiently low to be considered tolerable. The second risk is that national competition authorities or national courts might err in assessing whether or not an agreement contributes to the promotion of technical or economic progress (while allowing consumers a fair share of the benefit). Thus, the real risk is that the efficiency benefits of a vertical agreement that involves suppliers with a substantial share of the market, might be overestimated in certain jurisdictions (in which case the agreement will be exempted despite its substantial negative net impact on competition) or may be underestimated (in which case the agreement will not be exempted, even though it should be). One might wonder whether there are many cases of vertical distribution agreements that contribute to technical or economic progress. Usually such agreements are more effective in promoting or reducing competition than as a means of promoting technical progress. However, there are undoubtedly other vertical arrangements that are justified by efficiency reasons and raise the issue of the assessment of technical or economic progress. It is thus possible that some agreements (whether horizontal or vertical) might not be exempted in certain jurisdictions, while being exempted in others because of inconsistent appraisals of their contributions to economic and technical progress. Of course, such an outcome would raise the issue of legal certainty if, and only if, the same competitive conditions prevail in the two countries under consideration; or in other words, if and only if the national markets for the product under consideration are indeed integrated. If the national markets involved are not integrated, there is of course the possibility that the same vertical agreement could be exempt in one country (for example, because inter-brand competition is strong in that country and there is no scarcity of alternative retail outlets for potential or actual competitors) and prohibited in another (for example, if it leads to monopolisation of the retail outlets or if there is weak oligopolistic competition among manufacturers).

2 - Coherence

365

It is therefore fallacious to think that the consistent enforcement of EC law will necessarily lead to similar decisions or, conversely, that (apparently) contradictory decisions or judgments on a particular vertical agreement will necessarily result in inconsistent enforcement. What is important is to have consistency of reasoning (rather than consistency of decisions). The probability that horizontal agreements (which have the potential to provoke the greatest harm to consumers) will give rise to conflicts is relatively small. There is a somewhat larger probability that vertical agreements (which are the least likely to affect competition and the most likely to have efficiency enhancing effects) might give rise to inconsistent decisions.

III. Possible ways to increase consistency of enforcement The mere possibility of such conflicts is not sufficient to establish that the problem is important enough to warrant action. Also the existence of a conflict between national jurisdictions or courts is not necessary to warrant the intervention of the Commission. Indeed, if national courts are consistently wrong in applying Article 81 or 82 to a type of situation, the Commission should have some way to intervene. What is clear is that the likelihood of inconsistent decisions is enhanced by the differences in procedural rules in different European countries. With the possibility of inconsistent decisions comes the possibility of forum shopping. Third parties claiming damages, such as retailers excluded from a selective or exclusive distribution network, will have the choice of attacking a vertical agreement either in the country where the defendant is domiciled or in the country in which the harmful effect occurred (that is, the country in which the exclusion took place). They will have an incentive to carefully consider the standards of proof and the extent to which they can obtain assistance in the discovery procedure in each of the countries considered. Thus, even if there is no inconsistent enforcement of EC law, there might be differences in the intensity of enforcement between the European countries. However, even if it is necessary to obtain consistency of enforcement, harmonisation of evidentiary and procedural rules among European countries is a proposition unlikely to be met with much enthusiasm. Besides harmonisation of procedural and evidentiary rules, a second set of measures to ensure a higher level of consistency of enforcement lies in organising co-operation between national competition authorities or national courts on the one hand and between these national competition authorities or national courts and the Commission and European courts on the other. There is no question that the possibility for national courts or national authorities to consult the Commission or the EC Court of Justice should be facilitated. This is true not only for questions relating to the interpretation of the law but also for questions of fact (such as: what is the structure of competition on the

366

Working Paper V

European market? Do all manufacturers distribute their products on the European market the same way, and is there likely to be a cumulative effect of such arrangements? etc. It would be equally useful to allow national courts and national authorities to co-operate among themselves (for example to exchange information on similar cases being handled simultaneously or on the consistency of the relief they might be tempted to impose on one another). It is clear that there is very little co-operation among European authorities or courts at the present time. This is because of obstacles at the national level (co-operation with a foreign authority—which can result in the sanctioning of domestic firms for violations of the EC law—is often viewed with suspicion by national governments) and because of the limitations on co-operation imposed by European jurisprudence (in particular, the Spanish Banks cases). The EU should thus make a deliberate effort to try to improve the situation by devising (and perhaps imposing) minimum standards of commitment to cooperation among European countries. Incidentally, this approach could set a useful example for the work undertaken at the WTO. However, an issue that would have to be thought about is how to organise cooperation without imposing on the Commission the tremendous cost that it would face if it had to answer all calls for co-operation? Should a screen exist to enable the Commission to refuse to co-operate where it lacks the resources to do so? Should the Commission be given the right to refuse to co-operate (or to limit its co-operation) when it feels that there is insufficient Community interest involved in the case? Should another standard be found? If the Commission declines to co-operate at the first level, does that mean that it would waive its right to intervene at the appellate level? As an example, and to start the discussion on co-operation, it is worth mentioning that Article 26 of the French competition law (Ordinance of 1986 on the Freedom of Price and Competition) organises co-operation between the courts and the domestic competition authority (Conseil de la concurrence) when there are parallel proceedings. On the one hand, the Conseil can petition the courts to obtain access to the affidavits, testimonies and investigative reports in the possession of the courts and that are relevant to the case being investigated by the Conseil (note that the courts are free to grant access or not). On the other hand, courts can ask the Conseil to give an opinion about whether Article 7, 8 and 10 (1) (the equivalent of Articles 81 and 82 of the EU Treaty) apply to practices being examined or under investigation by the Courts. The Conseil (which may use its own wide investigatory powers to investigate a case referred to it by the court) must give its opinion after an adversarial debate, except in instances when it has dealt with the same case in the past and has either emitted an opinion (to another court) or taken a decision. In those instances, the Conseil can give an opinion without a contradictory debate. The opinion of the Conseil can be made public after the court judgment has been issued. In fact, each year courts do refer a few cases to the Conseil (which means that they choose to decide whether there is a violation of Article 7 or 8 without

2 - Coherence

367

asking for an opinion in the majority of cases; thus the Conseil has not been swamped by requests). In most such instances, courts refer cases to the Conseil either because the applicability of Article 7 or 8 is not clear or because they feel that they do not have the time, the procedural means or the expertise to obtain all the relevant information necessary to make an informed judgment. Among the remedies likely to alleviate the problem, the Commission has suggested that the setting up of specialised court: 'would contribute significantly to the efficiency and coherence of the new system'. This indeed is true. Such a step would, in all likelihood, also significantly improve the efficiency of the enforcement of the domestic competition laws, as these national courts would presumably be competent to hear both domestic competition law cases and Article 81 or 82 cases. In addition, this solution would not impose costs on the Commission (contrary to most of the other remedies) and therefore would not limit its ability to focus on the most serious breaches of competition law. Finally, in countries in which the setting up of 'specialised' courts would create a constitutional or political problem, an alternative would be to concentrate the enforcement in the hands of a few judges or a few courts. This solution was adopted in France where a section of the Paris Appeals Court (a non-specialised court) hears all the appeals pertaining to competition cases. Concentration or specialisation of judges is essential to allow the development of a competition culture within the judiciary, and it is a necessity to promote consistent enforcement of EC competition law. Another solution would be to allow the Commission to intervene in proceedings of national authorities or national courts. Two types of intervention have been discussed: the Commission could intervene as amicus curiae, or the Commission could appeal against national decisions in the Community interest. The intervention of the Commission as amicus curiae would have the advantage of minimising the risks of conflicts early on in the judicial process; it would probably also be very costly. Indeed, to know whether it wanted to intervene, the Commission would have to follow numerous cases at the national level. It therefore appears doubtful that the aim of freeing some of the resources of the Commission would be achieved by this means. If the intervention of the Commission were limited to the possibility of appeal against court or national authority decisions whenever it felt that the interest of the Community was at stake, it would seem that the cost of the Commission's intervention would be greatly reduced. Consistency of enforcement could even be obtained faster and at a lower cost if, when the Commission felt the Community interest was at stake, it could appeal a national court or national competition authority decision to the Court of First Instance.

IV. Conclusions When addressing the issue of consistency of enforcement of the EC law, one should not lose sight of the main goal of the EC's competition policy reform.

368

Working Paper V

That goal is to increase the means by which the Commission can fight the most egregious antitrust violations. The reform will necessarily mean that the enforcement of Article 81 will become more complex, both because the enforcement of Article 81 will be decentralised and because a more economic approach to the problem of vertical restraints will be taken. When considering the means of maintaining as much consistency of enforcement of Article 81 as possible, one should follow a cost benefit analysis rather than considering that any means that increases legal certainty for business firms is worth adopting. Because we cannot predict what the scope of the problem will be (that is, what the social cost of inconsistency will be), a cautious approach is recommended. In particular there is the risk that some means of solving the problem might entail more costs than benefits. Measures imposing a high cost on the Commission should be adopted only after there is convincing evidence that the social cost of inconsistent enforcement of Article 81 is sufficiently large to warrant their adoption. From that standpoint, retaining the possibility that firms can notify the Commission of their agreements is definitely not advisable in a first phase. It is also doubtful whether the Commission should be given the means to intervene in national court proceedings at the lower level. It should be given the right to appeal court rulings or decisions by national authorities, preferably to the Court of First Instance. Co-operation between the Commission, national authorities and national courts should be developed, but some (transparent) procedure should allow the Commission and the Court of Justice to decline to co-operate. The step-by-step, progressive approach suggested here is somewhat similar to the approach followed when the Community decided that it would review the EC merger regulation after a few years to decide whether the thresholds should be revised.

VI Anne Willem Kist and Maria Luisa Tierno Centella Coherence and Efficiency in a Decentralised Enforcement of EC Competition Rules: Some Reflections on the White Paper on Modernisation

I. Introduction It is fair to emphasise that the European Commission has led the movement for modernisation in both the form and substance of European competition law, acting as a catalyst for change and challenge, rather than as a brake or hindrance. This creative exercise is an incentive for us, as commentators, to not only spot the shortcomings of their far-reaching proposal, but also to actively contribute to its improvement with positive action and alternative suggestions. This article scrutinises the Commission's White Paper on the modernisation of EC competition law1 by taking coherence, consistency and efficiency as criteria for its assessment. Besides dealing with the systematic internal coherence between the proposals and the objectives set forth in the White Paper, we will further examine the question of coherence if enforcement is decentralised under Article 81. This last aspect can only be tackled usefully by going beyond the sparing, laconic wording of the White Paper on this matter. The truth is that the Commission and the representatives of the Member States involved in the discussions about modernisation have by now progressed further in this debate. Assessing the White Paper demands a realistic, down-to-earth approach in order to make up for its rather theoretical and abstract proposals. It is prudent to balance the Commission's highly valuable creative efforts with at least the following considerations: First, the success of the reform will probably be assessed on its practical consequences for the active investigation, termination and punishment of infringements. Second, only with a concrete text in our hands will we be able to assess whether the means chosen by the Commission really are likely to produce the legal effects it expects from them, and whether the wording of the new Regulation will really prompt the metamorphosis from an authorisation regime into a legal exemption system. We do not want to prejudge the coherence between the wording of the future Regulation and the intention to give direct effect to Article 81 (3), but we will mention some factors that already point in another direction. Third, a good system should also work when conditions are not optimal. The implementing documents or non-papers should not take conditions for granted Commission Program n. 99/027, 1999/C 132/01, OJ 12.5.1999.

370

Working Paper VI

or discuss modernisation exclusively on the fragile basis of the best case scenario where everyone and everything will accurately function like clockwork. That never happens. Common sense and sound management advise us to verify our assumptions beforehand or, where not feasible, to develop alternative scenarios allowing for a margin of error. The point is that even if we fall short of perfection, the change must still be a success. There is a reasonable limit on the number of major changes that can take place, particularly when we expect them to occur simultaneously and when they depend on the development of new skills, the assumption of a new competence or the exercise of new powers. The possibility cannot be excluded that the working group dealing with the White Paper would recommend taking one step at a time and introducing transitional or conditional provisions in the future Regulation. Should that happen, it is essential that the priorities become apparent to us, in order of urgency and importance. We should get a clear idea of which improvements are necessary and which are just desirable or useful; likewise, we should spot the improvements requiring extra preparatory measures. Finally, we must identify which changes belong together and should take place simultaneously.

II. The deficiencies identified by the Commission and the aim of the reform On the word of the Commission, 2 the application of a centralised system of prior control of agreements has become incompatible with one of the objectives enshrined in Art. 83 (former Art. 87): the effective supervision of competition. According to the White Paper, the obstacles to achieve that purpose are a couple of constraints in Regulation 17/62: (a) the Commission's exclusive competence to apply Art. 81 (3) (Art. 9(1) of Regulation 17/62), and (b) the system of preventive control of agreements (Art. 4(1) of Regulation 17/'62). Article 83 EC Treaty is the legal basis for the adoption of Regulation 17/62 and of any new Regulation that might amend or replace it. This provision requires that a Council Regulation 'be designed in particular to lay down detailed rules for the application of Article 81 (3)'. In doing so, account should be taken of the need to ensure effective supervision and to simplify administrative control to the greatest possible extent (Article 83 (2) (b)). 3 Our analysis of the White Paper and the implementing measures 4 should not neglect another 2

White Paper para. 9 (Introduction). Although the English versions of Article 83 and the White Paper state 'simplify administration', the exact wording in French, Dutch, Italian and German is 'simplify administrative control'. Those were the official languages of the 6 Members States who adopted the Treaty of Rome in 1957. That original wording is kept in the respective translations of the White Paper. 4 The White Paper recalls them in the chapter dealing with the options, at para. 52. 3

2 - Coherence

371

objective in Art. 83 EC Treaty: to ensure compliance with the prohibitions laid down in Articles 81 (1) and 82 by making provision for fines and periodic penalty payments. This means that all administrative enforcers of Articles 81 (1) and 82 should be empowered with at least those tools to pursue, terminate and punish infringements. Next to the above mentioned goals, Article 83 assigns a couple of tasks to the Regulation that deserve attention in an expert discussion: • To define the respective functions of the Commission and of the Court of Justice in EC competition matters. If we have to follow the Commission's opinion that the application of Article 81 (3) no longer serves policy directions, the review of its decisions by the Courts in Luxembourg could stop being marginal when Article 81 (3) is concerned. • To determine the relationship between national laws and EC provisions on competition. This has been subsequently settled by the case law of the Court of Justice on subjects such as direct effect, effet utile or the primacy of EC rules over national law. However, the new allocation of powers, the emergence of national laws on competition and the enhanced use of the principle of subsidiarity may bring about some nuances worthy of discussion. One would think that the choice for a decentralised application of the EC's competition law also demands the definition of the respective competencies and functions of the Commission, national competition authorities and, eventually, of national courts. This is not only relevant for the bodies involved, but for the businesses and lawyers as well. For undertakings and for lawyers it matters indeed which national procedural and substantive legislation applies, and which language must be used. These are not tiny details that can be solved once a complaint has been lodged, but relevant facts and issues that have to be explicitly consideredfrom the outset. Regulation 17/62 has not always been inefficient; it has become an ineffective method of supervision of the application of Article 81 (3) due to the new environment and the prospective circumstances. The White Paper addresses the limited administrative means and the material impossibility of formally handling an increasing number of applications for exemption5 within a reasonable time. The Commission wants to concentrate on investigations and on the detection of heavy infringements,6 rather than on issuing 'certificates of compliance'. In our capacity as a national competition authority, we are particularly sensitive to the need to focus on the investigation, termination and sanction of infringements. This approach and determination we want to take up for ourselves. The Commission wants to save the time and resources spent on enforcement. An impartial view of the annual number of new notifications for exemption reveals that the problem is probably not so acute thus far. The total sum of new notifications is reasonably low at the Commission level and is certainly so when 5 6

White Paper paras. 6, 10, 24 and 25. White Paper para. 40.

372

Working Paper VI

we look at the numbers in national notification systems based on the EC regime. Furthermore, the extended scope of the possibility of obtaining exemptions with retroactive effect (the new approach to vertical agreements) and the publication of detailed guidelines are likely to diminish the number of notifications. A decentralisation of competence would also influence the number of cases to be handled by the Commission itself. More convincing for us is the thought that five to ten new States are candidates to accession, and the subsequent increase in the number of preventive applications for exemption could seriously hamper the Commission's ability to play a proactive role in the discovery and investigation of serious infringements. This will be a recurrent situation after each new accession. The Commission advocates more powers of inquiry and investigation. Within the limits of the international conventions and what is permitted by the constitutions of the Member States, we can only welcome that initiative. We miss a reference to an important question for the effectiveness of the regime: a word on the allocation of the burden ofproof. The reform should in no way oblige competition authorities to prove that the conditions in Article 81 (3) are not fulfilled. We will develop this point below. The Commission reminds us further that the formalities required to issue an EC decision (including translations, approvals and advice) are cumbersome. We should learn from this experience not to repeat, spread and multiply the same shortcomings in the future network of authorities. We support the Commission's view that draft decisions should be exchanged electronically in the original language, with just a summary in one or two widely spoken languages. Certainly, if we have to work on the basis of summary exchange of decisions, there is no room for 'approval', but for soft and flexible ways of consultation and protection of consistency, subject to preclusive delays compatible with the obligation of the authorities to handle cases diligently. After all, the whole modernisation exercise must serve the purpose of simplifying administrative control to the largest extent possible. Some of the formalities faced by the Commission in its decision-making process might be burdensome, but they are also indispensable to protect the rights of interested parties. Yet, because the Commission cannot handle all cases formally, it has developed a surrogate for formal decisions: comfort letters. Thereby it has reduced formalities, legal certainty, and transparency altogether. Motivated comfort letters establishing that Article 81 (1) or 82 are probably not infringed, and comfort letters declaring that a block exemption regulation applies, are neutral documents which can even be helpful for other enforcers dealing with the case. Comfort letters on the application of Article 81 (3) are, on the contrary, highly controversial documents. They state that, although 'the agreement appears to contain restrictions of competition', DG COMP 'takes the view that sufficient justification has been provided for finding that the conditions laid down by Article 81 (3) are fulfilled'. Such an 'individual exemption comfort letter' explains that the agreement is probably forbidden under the directly applicable prohibition in Article 81 (1) and it just does not

2 — Coherence

373

grant any individual exemption to redress the situation. A national authority or a national judge can do little with a document that does not protect the agreement against the direct effect of Article 81(1) and 81 (2), because they are themselves unable to grant an exemption instead. This situation is utterly formalistic and impractical. Also for this reason, decentralisation is a more effective way to handle cases for all instances involved. Likewise, decentralisation can be expected to favour a broader application of EC law by national instances, and thereby a more consistent treatment of cases all over the EU. The current monopoly discourages authorities from finding that an agreement may affect trade between Member States, because that finding actually limits their capacity to handle the file efficiently. It can lead to a suspension of proceedings, or even to their closure after an investigation has been set up. The Commission's monopoly does not work efficiently for the undertakings involved either, for it causes delays and rejections of cases by national competition authorities. The Commission's exclusive competence for the application of Article 81 (3) has become contrary to the principle of subsidiarity because it sometimes prevents files from being handled at the most effective level. National competition authorities could fully assess many cases materially, since most national laws already require them to apply a provision largely inspired by Article 81 (3). Only the formal impossibility of doing so, only the legal monopoly, prevents them from fully assessing many files. Consequently, the Commission's monopoly has become a handicap for the effective supervision of competition by all agents involved. We agree with the Commission that this situation cannot last much longer and that it is high time to review the boundaries of its exclusive competence. Against all these reasons for decentralisation, some commentators raise the concern that 'national interests' could guide the decisions of the less independent national competition authorities. We cannot agree. Should that risk exist, it is in no way increased by the decentralisation of the competence to apply Article 81 (3). National competition authorities are already able to apply Articles 81 (1) and 82 taking competition interests into account. Moreover, national competition authorities are also able to apply Article 81 (3) negatively, either by determining that there is no chance that the conditions in Article 81 (3) would be met, or by withdrawing the benefit of the new block exemption for vertical agreements. On top of that, the Commission will retain the evocation mechanism in Article 9 (3) of Regulation 17/62: it will still be able to hear cases from national competition authorities when needed, thereby withdrawing their competence to apply EC law in a certain case. The enforcement of Article 86 (1) against a Member State favouring public undertakings is one of the situations in which the Commission's intervention might be especially appropriate. Finally, the Commission's proposal would probably do no wrong by leading to the increased independence of some national competition authorities. If priorities have to be set in order to reach the objectives of the reform, the decentralisation of jurisdiction to national competition authorities should take precedence, before the decentralisation to national judges. Legal practitioners,

374

Working Paper VI

scholars (even in this forum) and, more significantly, national competition authorities themselves have abundantly pleaded for the decentralisation to administrative enforcers and they have underlined its necessity. As a matter of fact and without raising arguments for or against the empowerment of national judges, no comparable current of opinion has identified a similar problem regarding national courts, and nor has an equivalent demand or urge existed from the side of our judges. Furthermore, the Treaty itself treats national competition authorities and national judges differently: according to Art. 84 (former Art. 88), if Regulation 17/62 were abrogated instead of being modified or replaced, only national competition authorities, and not national courts, would recover their original competence to rule on agreements on the basis of Art. 81 (3). That is a fact. A last reflection over the objectives of the reform takes us to a broader perspective: the enforcement of EC competition law must become more efficient overall, for all parties and authorities involved and not only for the Commission itself. A simple transfer of workload or a higher degree of bureaucracy should be avoided at all costs. The administrative efficiency has to be felt positively by market players.

How does the Commission intend to obtain its objectives? The Commission wants to replace or amend Regulation 17/62 in order to grant competence to apply Article 81 (3) to national competition authorities and national judges, and impose on them the obligation to apply it simultaneously to Article 81 (1). The Commission's purpose is not to share the power to grant exemptions, but to share the competence to declare the whole of Article 81 applicable or not. The only tentative wording made available by the Commission services is as follows: 'All national competition authorities and courts called upon to apply Article 81 (1) should simultaneously apply Article 81 (3)'. We are convinced that this wording is intended to serve a commendable purpose: making full economic assessments of cases. Nonetheless, it raises two major concerns: whether the burden of proof would shift completely to the shoulders of the enforcers, and whether such a text could reach the Commission purpose of establishing a legal exemption system. Another question—and one that is in no way rhetorical—is whether the system could still reach its objectives without amounting to a legal exemption.

III. Who has the burden of proof? The Commission's tentative wording suggests that whenever the applicability of Article 81 (1) is at stake, the enforcing authority should apply Article 81 (3) on

2 - Coherence

375

its own initiative. We should be very careful because that would put the whole burden of proof on the authority. In order to prove the infringement, we would be required to demonstrate also that the conditions in Article 81 (3) had not been met. Until now, the explanations given by the Commission services have not lessened our concern. If Article 81 (3) has to be applied ex officio and bearing the burden of proof, the system may here find its Achilles heel. Such an approach would encumber and delay the administrative procedures to an extent incompatible with the objective of efficiency in the detection, prosecution, termination and punishment of anti-competitive practices. It would also be a deterrent for the instrument that the Commission wants to foster most: complaints. Therefore, such solution would favour neither the effective supervision of competition nor the simplification of administrative control. The efficiency costs would affect the work of each and every enforcer in the network, who would have to handle EC cases according to the same principle. That could create inconsistencies between the treatment of purely national cases, and cases to which EC rules also apply. Moreover, such an approach would be in open contradiction with what the Commission's Legal Service has consistently pleaded before the Courts: The Commission replies that the Court of Justice has consistently held that it is incumbent first and foremost on the undertakings to convince it, on the basis of documentary evidence, that an exemption is justified. (. . .) A fortiori the Commission cannot be required to demonstrate that an exemption is not justified or to indicate what it regards as acceptable'.7

Shifting the burden of proof would challenge well-established case law on the application of Article 81 (3).8 Such case law is the only relevant rule on the burden of proof because Regulation 17/62 is silent about this matter. Such an allocation of the burden of proof is not inherent in a legal exemption system; it can be conveniently adjusted by adding the necessary wording. The French legal exception regime requires the interested parties to prove that they have fulfilled the requirements. Shifting the burden of proof would justify the worries of some commentators that the direct effect of Article 81 (1) and (2) could be jeopardised by the new regime. This would introduce a Trojan horse into the system that would bring about nullity in legal terms and inefficiency in practice.

7 See Joined Cases C-56/64 and C-58/64 Consten and Grundig v Commission [1966] ECR 299, Case C-42/84 Remia v Commission [1985] ECR 2585. 8 See, for instance: Case T-17/93 Matra v Commission; Case T-29/92 Vereniging van Samenwerkende Prijsregelende Organisatie in de Bouwnijverheid v Commission, para. 262; Case T-35/92 John Deere v Commission [1994] ECR 11-957, at para. 105; Joint Cases T-39/92 and T-40/92 Carte Bleu v Commission, para. 114; Case C-45/85 Verband der Sachversicherer v Commission; Joint Cases 43/82 and 63/82 VBVB and VBBB v Commission [1984] ECR 19; Case 258/78 Nungesser v Commission [1982] ECR 2015.

376

Working Paper VI

Our conclusion is that this wording ought to be reviewed or completed in order to put the burden of proof of the application of Article 81 (3) back on the undertakings. This should be clearly codified in the new Regulation. Three further thoughts should be added: First, it is worthwhile exploring whether we should take the opportunity of codifying the bearing of the burden of proof in order to clarify in guidelines which efficiencies have to be analysed under paragraphs 1 and 3 of Article 81. This will help comprehension and consistent enforcement throughout the Union. This is also relevant for the application of Article 81 by national judges, since parties are normally bound by the classical adage: secvndvm allegata probata. Second, once we start codifying the bearing of the burden of proof and clarifying its consequences, a nuance should be introduced: the prevention of the probatio diavolica. In some cases, only one party has access to the means to prove or disprove a fact and should therefore sometimes bear that concrete burden. This clarification, together with the former, could put parties on an equal footing before all enforcers in Europe. Third, the rule that companies invoking the application of Article 81 (3) should prove that the conditions are met is a capital principle of the EC competition law. Moreover, it represents an indispensable legal correction for an exclusively economic approach. To put it in another way, the burden of proof is the limit of the mostflexiblelegal boundary for the application of Article 81 following an economic approach. We support the trend to simultaneously balance economic efficiencies and inefficiencies; however this trend also has to be balanced against the efficiency of the enforcement itself.

IV. Does the proposed wording create a legal exception? Article 81 (3) in itself is not directly applicable, and it grants no absolute right to an exemption even if the four cumulative conditions are met. Article 81 (3) does not create rights for individuals that national courts must safeguard. If that were the case, Article 81 (3) would already have direct effect and the current exemption system would have always been illegal, by imposing an extra requirement upon individuals for the exercise of their rights. Should Article 81 (3) create rights for individuals, judges would not need a Council Regulation to be empowered to apply it. Furthermore, block exemption regulations would just be declaratory instruments, instead of having a constitutive nature. The Commission does not argue that Article 81 (3) is directly applicable, but that it may become directly applicable by way of a Council Regulation. A Regulation based on Article 83 and establishing that 'all national competition authorities and courts called upon to apply Article 81 (1) should simultaneously apply Article 81 (3)' would render Article 81 (3) directly applicable.

2 - Coherence

377

We encourage the Commission to review and improve this wording because we are not fully convinced that this text will put a legal exemption system into place. We also request the Commission to explain thoroughly how the nature of the decisions made in application of Article 81 (3) will change from constitutive into declaratory. Let us explain both concerns. In our view, the efficacy of the tentative wording to perform such a transformation in the nature of decisions is rather limited. This sentence merely attributes competence to several instances and establishes a condition for its exercise. It does not lead to the conclusion that the competent instances will no longer be granting an exemption. The fact that the competent authority or court will have to apply paragraphs 1 and 3 of Art. 81 simultaneously is no novelty, because every time the Commission has granted an exemption it has applied both paragraphs simultaneously. Does this provision go further than a 'simple' decentralisation of competence to national competition authorities and judges? It does, but only in one way: it eliminates the requirement of preventive notification for exemption. It removes the need to act upon a request to apply Article 81 (3). In the Commission's wording, all national competition authorities and courts called upon to apply Article 81 (1) should simultaneously apply Article 81 (3); in a system of preventive notification, all national competition authorities and courts called upon to apply Article 81 (3) should simultaneously apply Article 81 (1). With the wording proposed by the Commission, the only consequence is that the exemption may, and eventually must, be granted without preventive notification, either in an ex officio proceeding or upon the request of a complaint. However, 'exemption' and 'notification' are not synonyms. One can grant an exemption without notification by just assessing the case when a complaint arises or when a procedure is heard on the Commission's own motion. There can also be notification without exemption, as proposed by Germany in the last meetings of the Workgroup and explained by Dr. Ulf Boge in his contribution to the present volume. An exemption can already be granted on the occasion of a complaint and it can have retroactive effect counting from the very moment when the agreement was made. This scenario is possible when Article 4 (2) of Regulation 17/62 applies, and it is absolutely compatible with the traditional interpretation of the relationship between paragraphs 1 and 3 of Art. 81. Still, we talk about an 'exemption', or a constitutive decision. More generally, the constitutive value of the decision taken in application of Article 81 (3)—the 'exemption'—does not vanish because Article 81 (3) and 81 (1) are applied together. The competent authority always applies Article 81 (3) in conjunction with Article 81 (1) and, in doing so, it takes a constitutive decision granting an exemption. The difference between the current situation and the one implicit in the alternative text is the following: as of now, there are decisions whereby only Article 81 (1) is enforced either because the conditions in Article 81 (3) not being met or because the parties have not notified authorities of the agreement (these are

378

Working Paper VI

declaratory decisions). Next to them, there are also decisions where both Article 81 (1) and Article 81 (3) are applied (these are constitutive decisions or 'exemptions'). The Commission has never formulated an exemption decision declaring that an agreement is compatible with Article 81 (as a whole), but actually, it could have.9 In an individual exemption decision, the Commission infers that the benefits of the agreement—under Art. 81 (3)—offset its restrictive effect— under Art. 81 (1)—and, as a consequence, the legal nullity in 81 (2) no longer applies. The nullity may be eluded for the past or/and for the future. Validation is occasionally granted with retroactive effect from the very moment when the agreement was concluded.10 More often the retroactive effect only goes back to the date of the application for exemption. An exemption subject to conditions can eventually have effect for the future only {ex nunc). Whatever the time limit of their effects, validation decisions are and remain constitutive in nature. Therefore, under the current system a decision applying Article 81 (3) retroactively, applicable from the very instant of the conclusion of the agreement (what the Commission proposes as general rule for the future) is no less constitutive than any other exemption. As for the future, it seems that all decisions will deal with both paragraphs, without previous notification. In our view, it has still to be shown that this will change the nature of the decision made. The new scenario is as follows:. Sometimes the conditions in paragraph 3 will still not be met, and only Article 81 (1) [and (2)] will apply (exactly as today). The Commission will formulate it by saying that the agreement is incompatible with Article 81 (as a whole). Other times, both paragraphs 1 and 3 will apply (but not the nullity in paragraph 2, exactly as today) and then the Commission will write that the agreement is compatible with Article 81 (as a whole). So far, this is no different from an exemption. Obviously, without a preventive notification system, the validation could only have either a retroactive effect counting from the date of conclusion of the agreement {ex tune) or a future effect from the time of the decision {ex nunc). An economic approach and the principles of equity and legal certainty indicate that the correct option is the first one. It would not be fair to abolish the possibility to 'notify' and grant exemptions only for the future. Yet, the tentative text does not suffice to validate agreements retroactively, while the current Article 4 (2) of Regulation 17/62 does. This nuance should be explicitly stated somewhere in the new Regulation. The consequences of finding that the new Regulation might not put a legal exemption in place are not so dramatic. The reform could still work because the 9

In the same way, the Commission could have formulated prohibition decisions by writing that the agreements are incompatible with Article 81 (as a whole), at least when the conditions in Article 81 (3) are not met. 10 In application of Article 4 (2) of Regulation 17/62.

2 — Coherence

379

two undesired constraints in Regulation 17/62—preventive notification and the Commission's monopoly in applying Article 81 (3)—would be abolished. Besides, the validation of agreements would have full retroactive effect and production joint ventures could be consistently treated under another regime without being an exception to an exception. Furthermore, regulations applying Article 81 (3) would logically remain as (block) 'exemptions' and competition authorities could use fine-tuning when granting an individual exemption (conditions and duration). This interpretation is not even incompatible with empowering the judges (although they would only grant exemptions for the past), but it allows the possibility of empowering them once the necessary means have been put in place. Hoping to have served the healthy purpose of giving rise to some intellectual polemics, and to a change in the future wording of the new Regulation, we turn to another question: coherence in the decentralised enforcement of EC rules.

V. A decentralised competition enforcement system at work The consequences of the reform proposed by the Commission are manifold and profound from the perspective of all national bodies called upon to enforce EC competition rules as well as for companies expected to comply with them. The involvement of national competition authorities will be particularly intense, as they will have to: • learn to co-operate with each other closer than ever before, to an extent not yet known, sharing experience, means, knowledge and information with new technical means and a common language; and • improve their communication and contacts with national courts. It should be reminded, though, that the NCAs of about half of the current Member States will become competent to apply Articles 81 and 82 for the first time. In addition, NCAs will be given a new competence, namely to apply Article 81 (3), and they will also be empowered to withdraw the benefit of block exemptions in individual cases. We would not be realistic enough if we did not acknowledge that national competition authorities do not apply EC rules as much as they could, and that their co-operation with the Commission and, in particular, with other competition authorities could be improved. Decentralising more competencies without increasing vertical and horizontal co-operation and without encouraging the use of EC provisions in the resolution of cases would be useless. All effectiveness would be lost. That is the reason why the mechanisms for consistency have to be enhanced; not only for the sake of the application of Article 81 (3), but also for the enforcement of the prohibition provisions.

380

Working Paper VI

Rising to this challenge takes more than our desire and enthusiasm; it requires analysis, realism and commitment.

1. Decentralisation of policy or of enforcement? Any rational appraisal of the risks and challenges of decentralisation must start with a clear view of what exactly will be decentralised, what competencies will be shared, and with whom. Let us face it: competition policy as such is not going to be decentralised, only its enforcement will be. Article 81 (3) has traditionally been a policy instrument by way of block exemption regulations or individual decisions. Nowadays, some Commission officials insist on saying that there is a right to an exemption whenever the four conditions in Art. 81 (3) are met. In fact, the Commission enjoys such a wide margin of discretion1' for the assessment of those conditions that it has always been able to defend, in its view, the fact that this or that requirement was not fulfilled, without having to declare that it has chosen not to grant an exemption although all the conditions of Article 81 (3) were fulfilled. In legal workshops and papers, however, the Commission has traditionally defended the fact that it has no obligation to grant an exemption, even if the four conditions are met. One of the most clear and interesting examples in this line is to be found in Emil Paulis' contribution to the 1996 European Competition Law Annual: 'Article 81 (3) leaves a wide margin of discretion to the Commission, because it does not state that the Commission 'shall', but that the Commission 'may' declare 85(1) inapplicable if certain conditions are fulfilled. Thus, there is no absolute right to an exemption. Article 81 (3) is different in this respect from Article 65 of the ECSC Treaty and Article 92 (2) of the EC Treaty'.

The wording of most national competition statutes across the EU inspired from Article 81 EC Treaty shows that this was certainly the interpretation given by national legislators to this provision. Yet if we think that the future decentralisation relates mainly to enforcement, it is for two reasons:

1 ' The Court of Justice and the Court of First Instance carry out only a marginal control of decisions applying Article 81 (3), while their review is complete under Article 81 (1) and 82. The margin of discretion is accordingly broader: 'The review carried out by the Community judicature of the complex economic assessments undertaken by the Commission in the exercise of the discretion conferred on it by Article 85 (3) of the Treaty in relation to each of the four conditions laid down therein must... be limited to ascertaining whether the procedural rules have been complied with, whether proper reasons have been provided, whether the facts have been accurately stated and whether there has been any manifest error of appraisal or misuse of powers' (Case T-29/92 SPO and Others v Commission [1995] ECR 11-289, at para. 288, upheld on appeal in case C-137/95 P [1996] ECR 1-1612).

2 - Coherence

381

First, future policy will be set basically with block exemption regulations, detailed guidelines and some selected decisions. The room left for policy in the application of Article 81 (3) by national competition authorities will, in the end, be comparable to the one we have in the application of Article 81(1) and 82. Therefore, the risks for the consistent application of Article 81 throughout the European Union are comparable to the risks of an inconsistent application of Article 82. Second, unity and consistency of decentralised enforcement across the EU requires a minimum homogeneity: whenever the parties justify the applicability of Article 81 (3), the exemption rule ought to be applied.

2. Mechanisms to ensure coherence and consistency in the enforcement of EC competition law within the network of competition authorities The White Paper discusses preventive and corrective mechanisms to ensure consistency within the network of competition authorities. Preventive mechanisms are less traumatic, and therefore preferable, since they limit the involvement of the Commission as to what is strictly necessary to safeguard consistency, most of the time without it having to implement its own decision procedure. Besides, preventive mechanisms do not change the Commission into a court of first instance to which undertakings may appeal, dejure or de facto, against national competition authorities. Article 9 (3) of Regulation 17/62 will still be available for the Commission to withdraw the competence of a national competition authority to apply EC competition law in a concrete case. It should be used sparingly. Commission decisions will remain subject to consultation within the Advisory Committee, as they are now. A database with the cases already handled by competition authorities could prevent contradictory and superfluous decisions. Authorities should insert their new cases in the database, in a way similar to the notifications they get from the Commission about new applications. Those electronic means would simplify the task of identifying overlaps and ensuring consistency. Another question is what draft decisions or draft statement of objections should be exchanged within the network (if any), and with what purpose. Are we just talking about consultation and reallocation or also approval? For as long as the legal effects of a draft decision remain limited to the territory of a Member State, and especially if that territory forms a distinct market within the EU, draft decisions applying EC rules should be sent to the Commission for information. The Commission would be granted a delay for reaction. The same procedure would be followed to involve other national competition authorities 12 12

As the Commission proposes: with a summary in 'a commonly spoken language'.

382

Working Paper VI

on a voluntary basis and, when more authorities are handling parallel cases, compulsorily. 'Compulsory consultation' does not mean 'application for approval or authorisation'. In principle we would only see the need for electronic consultation, with the possibility of convening an advisory committee only upon request. This soft harmonisation would be less burdensome for the members of the network. The advantage is that we do not reproduce the formalities and the bureaucracy that we were trying to correct. Moreover, a soft harmonisation would encourage the enforcement of community law by making it much more appealing for national competition authorities to simply apply national law. Those exchanges with a view to consistency would not justify the exchange of confidential information. Only draft public versions should be disseminated, subject to an obligation to maintain secrecy. Confidential information in general should only be exchanged with those authorities who need it in order to handle a case applying EC competition law13 and subject to the obligation to observe secrecy (except when essential to prove an infringement). There are still some complications to be solved. In countries where the national authority is actually a real judicial body (Ireland), preventive mechanisms would not be plausible. In some Member States (France, Belgium, Portugal and Spain, for instance) the deciding authority is a deliberating Council. In those cases, we could still explore the possibility of using a preventive mechanism by taking notice of preparatory documents made by the Council rapporteur. Corrective mechanisms are improper means between administrative authorities, particularly when they are not connected hierarchically and when one of them—the Commission—could have withdrawn the competence of the rest (Article 9.3 of Reg. 17). The competence of national competition authorities to apply Article 81 (3) is in no way delegated by the Commission, but is directly based on the future Regulation, and even on Article 84 of the EC Treaty. Correction should be left to the judicial power. We have to ponder the possibility of informing the parties about the advice of the network, which could be of use to help their arguments before the judge.

3. Should the structure, powers and procedures of national competition authorities be harmonised? In a system without preventive applications for exemption, the effective and homogeneous enforcement of EC law depends on having the largest possible number of national competition authorities taking part in the future network. 13

Another question is whether a multilateral agreement or a regulation based on Article 11 EC should ensure the same kind of arrangement for the application of national competition law.

2 — Coherence

383

It is of central importance that all national competition authorities are empowered by their national acts with: • the competence to apply Article 81(1) and (3), together with Article 82 EC; • the legal means to settle cases, including the power to investigate infringements, impose fines, periodic penalty payments and remedies and to order interim measures; • the adequate material resources in terms of manpower and equipment; and • the possibility of exchanging and using information with other national competition authorities, subject to the necessary legal requirements. The second and third criteria have already been mentioned in the Commission Notice on co-operation between national competition authorities and the EC as conditions for the Commission to send a case to a national authority.14 This is only consistent with Article 83, which requires that compliance with the prohibitions laid down in Articles 81 (1) and 82 should be ensured by making provision for fines and periodic penalty payments. It is also consistent with the case law on rejection of complaints for lack of Community interest. This is only possible when appropriate remedies are available at a national level.15 Those criteria comprise the common denominator required if we want to work together. National competition authorities can be very different from each other: depending on whether their competencies apply for all sectors of the economy, on who takes the decisions, etc. Some Member States have an integrated authority encompassing the powers of investigation and decision, while others have a body for investigation and a collegial body (a council) makes the decisions, or even a court of justice. On top of that, competition law is subject to criminal sanctions in some countries and to administrative or civil sanctions in other. A common denominator has, thus, to be found. A certain harmonisation of the level and kind of sanctions (only fines) would be suitable to prevent forum shopping and to facilitate the exchange of evidence. Community competition law only uses administrative fines to punish infringements. Information can only be exchanged to the extent that the authority receiving it would use it to eventually impose fines, (not the sanction of imprisonment).

14 '25. A national authority having sufficient resources in terms of manpower and equipment and having had the requisite powers conferred on it, also needs to be able to deal effectively with any cases covered by the Community rules which it proposes to take on. The effectiveness of a national authority's action is dependent on its powers of investigation, the legal means it has at its disposal for settling a case—including the power to order interim measures in an emergency—and the penalties it is empowered to impose on businesses found guilty of infringing the competition rules. Differences between the rules of procedure applicable in the various Member States should not, in the Commission's view, lead to outcomes which differ in their effectiveness when similar cases are being dealt with'. 15 See Case T-24/90 Automec IIECR 11-2223.

384

Working Paper VI

Leniency programmes should be harmonised accordingly if they are to work at all. The authorities offering leniency would not obtain evidence if they could then hand it over to another authority that applies no leniency program. The possibility of holding individuals liable for infringements has to be explored further, as it could be a good complement to leniency programs. National procedures could remain in place without harmonisation, except for the points already mentioned. 4. Repercussions of the reform proposals on national competition laws 4.1. The theoretical possibility of maintaining the burden of notification at national level The Commission services have told national representatives that 'there is no legal obligation for Member States to align national notification systems to the new EC model [no problem for purely national cases]. However, in cases affecting trade between Member States, the primacy of EC law must be respected'. Accordingly, the Commission suggests that it would be legally possible, and even desirable, to maintain a notification system for purely national cases ('a notification system may prove useful in the initial phase of new competition legislation, as was the case for Regulation 17/62 in the early sixties'). On the contrary, the national notification system would not apply to cases involving the application of both national and community law. It would be highly inefficient to choose the route of a national notification system under those conditions. Such choice would discriminate between operators depending on their capacities to affect trade between Member States even for the purpose of applying national law. Besides, the ability to affect trade between Member States is not always obvious, and unnecessary notifications would be lodged. Furthermore, if Member States will finally endorse the Commission's proposal, they would have been convinced that there are more gains in the new system than in a notification regime and, in the context of that conviction, it would be contradictory maintaining a notification system at the national level. Multiplying procedures to handle the same kind of infractions and sectors would also work inefficiently for case handlers, undertakings and legal practitioners. It is true that some Member States could need or prefer a notification system, particularly new Member States getting adapted to EC competition law. However, if the possibility of prior notification were to remain, then they should be allowed to use it also for the application of EC law, and that has no purpose in a system of legal exemption, but only in a decentralised system of authorisation with full retroactive effect.

2 - Coherence

385

4.2. Consequences of the primacy of a directly applicable Article 81 (3) The Commission services have advanced a disturbing interpretation of the consequences of the reform for national law. Under the scenario proposed by the Commission national competition authorities would find themselves in the impossibility to prohibit anti-competitive practices falling under 81 (3). This scenario amounts to establishing that national competition authorities and national courts called upon to apply national competition law should simultaneously apply Article 81 (3), even if the parties have not invoked this provision. However, national competition authorities cannot always exclude the possibility that an agreement might fall under Article 81 (3). If the burden of proof remains where it should, it is up to the companies claiming the applicability of Article 81 (3) to actively demonstrate it in order to elude the application of a national (or EC) prohibition. If we were to follow another opinion, national competition authorities would have to prove that the conditions of Article 81 (3), are not met every time they make a prohibition decision (even under national law). As we have seen, such an allocation of the burden of proof does not even exist in real legal exemption systems, such as in France.

VI. Conclusions There is a long way to go before all the theoretical and practical questions will be adjusted, but we should walk all the way without haste. Let us take the time for it. This exercise is one of the most stimulating and inspiring tasks for European competition experts and its implementation must result in the reinvigoration of competition enforcement. Let us not jeopardise a good project by rushing its progress or by not carefully pondering each point at stake. The best contribution we can make at this stage of conception is to immunise the project with early constructive criticisms and to feed it with all improvements we can already think of. Let us not forget that all parties involved must have clarity about a number of subjects, including: • The rules on allocation of cases between the Commission and national competition authorities, as well as among the national competition authorities. • A clarification and codification of the criteria on the basis of which it is determined when a case has an EC dimension (or effect on trade between Member States). • The Commission's role regarding policy making, enforcement and coordination. • The rules on the burden of proof, i.e. who bears the burden of proof and to what extent. In this sense, a clarification on the competitive advantages to be examined under paragraph 1 and 3 of Article 81 would be welcomed.

386

Working Paper VI

• The way in which national courts will be able to obtain guidance and assistance on how to apply the EC competition rules, on which matters (factual, legal), and the value and effect of that guidance and assistance. Some ambition should be encouraged to set minimum requirements on procedure and on the concomitant institutional framework: those necessities to make enforcers more efficient individually and collectively (we have already mentioned leniency programs, exchange of confidential information, use of evidence, autonomy, etc). We should not lose sight of the variety of institutions involved (courts, councils, integrated authorities, independent and less independent administrative bodies), and the effects of these differences on the co-operation. We are bound to foster mutual trust in order to co-operate and to apply the same set of rules. Therefore, we should rather rely on preventive control than on corrective means. Finally, we suggest that the Commission should review its current tentative text for the reasons explained in this paper, in order to meet the high expectations nourished hitherto.

VII Gheorghe Oprescu The Modernisation of EC Competition Law: The Case of an Associated Country

Abstract Section I of this article describes the provisions of the Romanian competition legislation relevant in the context of the debate on the modernisation of EC competition law enforcement. The Romanian competition statute (Law No. 21/1996) is largely based on the Community competition rules. It covers cartels, abuse of dominant position and mergers. The law sets up a procedure by which undertakings may, on criteria balancing the anti-competitive and positive effects on consumers and the national economy, seek exemptions from the legal prohibition (this applies to both individual and group exemptions). At present the Competition Council has to be notified of all agreements, and exemptions are granted on a case-by-case basis. Notification is necessary for the time being because there is not enough experience of the application of competition rules. Section II evaluates several of the provisions of the Association Agreements that, in our opinion, do not show a perfectly coherent philosophy. This is due to some lack of concrete detail and to several ambiguities regarding the enforcement process. Section III analyses the impact that the Commission's reform proposal could have on the enforcement of competition law in Romania. It concludes that— though a problem in the short-term—it will not be a cause for concern in the medium-term.

I. The Romanian competition enforcement system Articles 81 and 82 EC Treaty find their correspondent in Articles 5 and 6 of the Romanian competition statute (Law 21/1996, which entered into force on 1 February 1997). Article 5 prohibits 'any express or tacit agreements between undertakings or association of undertakings and any partnership decision1 or concerted practices which have as their object or may have as their effect the 1

The relevant part of the analogous paragraph of Article 81 EC Treaty reads: 'The following shall be prohibited as incompatible with the common market: all agreements between undertakings, decisions by associations of undertakings and concerted practices which may affect...' (emphasis added).

388

Working Paper VII

restriction, prevention or distortion of competition on the Romanian market or on a part of it, shall be prohibited . . .' Article 5 (2) provides that agreements or concerted practices may be exempted from the interdiction of paragraph 1 if they fulfil the first four cumulative conditions (below 'a' to 'd'), and one out of other five listed in the last condition ('e'): 'a) the positive effects prevail over the negative ones or are sufficient to compensate the restriction of competition caused by the respective agreement, partnership decision or concerted practice; b) beneficiaries or consumers are allowed an advantage corresponding to that realised by the parties involved in the respective agreement; c) possible restrictions of competition are indispensable for the attainment of the foreseen advantages, and the agreement or concerted practice does not impose upon the undertakings concerned restrictions that are not necessary for the fulfilment of the objectives mentioned in letter (e); d) the respective agreement, partnership decision or concerted practice does not afford the undertakings the possibility of eliminating competition in respect of a substantial part of the product or service market in question; e) the agreement contributes, or may contribute, significantly to: —improving the production or distribution of goods or services; —promoting technical or economic progress, improving the quality of goods or services; —consolidating the competitive position of small and medium-sized enterprises on the domestic market; —increasing the competitiveness of Romanian goods and services on the international market; —charging, over the long run, substantial and durable lower consumer prices.' The provisions in (b),2 (c), (d) and the first two lines of (e) are quite similar to the EC Treaty provisions. Provision (a) though exceeds the Community framework, suggesting it leaves less room for granting exemptions if judging by its quasi—quantitative approach. The last three objectives in (e) also represent an 'originality' of the Romanian legislation; while the desire to strengthen the sector comprised of small and medium-sized enterprises might be understandable,3 the indirect support (by means of exemption) for increasing the competitiveness of Romanian exports might cause serious concerns in the framework of the accession process. At the same time, the last objective—refer2

Even if the advantages for the consumers—and 'beneficiaries', whatever this means—should be more precisely evaluated in order to be declared as 'corresponding' to those of the undertakings involved, as opposed to 'a fair share of the resulting benefit' in the EC Treaty. 3 However, the same objective is dealt with by Article 8 (1), which provides a 'de minimis' rule for agreements of minor importance.

2 - Coherence

389

ring to substantially lower consumer prices—is very vague (it is unclear whether it refers to nominal or to real prices, or whether or not it takes quality improvements into consideration) and is reminiscent of the periods when prices were strictly controlled by the state. The Romanian competition statute also provides that the exemptions mentioned in paragraph (2) shall be granted through Competition Council decisions in cases involving individual agreements, partnership decisions or concerted practices, while exemptions for certain other categories of agreements (block exemptions) shall be granted through Competition Council regulations. The Competition Council issued block exemption regulations within the following categories of agreements: exclusive distribution, exclusive purchasing, research and development, specialisation, technology transfer, franchise, motor vehicle distribution and servicing, and insurance. The undertakings requesting individual exemptions from the Competition Council shall prove that the conditions of paragraph (2) are met. At the same time—and this is the main difference by respect to the current EC legal framework—agreements or concerted practices that as a matter of principle fall under the provisions of block exemption regulations must also be 'notified' to the Competition Council, which will verify whether they comply with the criteria and conditions for block exemption. Therefore, the current Romanian competition legislation imposes an obligation to notify even in the case of agreements falling under block exemption regulations, despite the existence of specific regulations for most of the categories of agreements that are also block-exempted in the European legislation. This intermediate system (i.e. requiring compulsory notification despite the existence of the block exemption regulations, which would normally allow for a 'declaration of inapplicability') is to be explained by the fact that the Romanian administration and business community are not yet very familiar with the concepts of free market and free enterprise, which are the main components of a so-called 'competition culture'. Any decision taken by the Competition Council may be appealed to the Court of Appeals. To summarise, at present the Romanian competition enforcement system is centralised and based on prior notification and authorisation (including in the case of agreements falling under block exemption regulations). In this system, the administrative authority plays an important role, whereas the involvement of the courts is moderate. In some cases, the intervention of courts is indispensable in order to ensure compliance with the decisions of the competition authority that aim to correct restrictive practices or abuses of dominance. In theory, the courts also have a role to play if undertakings challenge a decision by the Competition Council, though so far this has never occurred in practice. We may nevertheless imagine a situation in which the court reverses a decision of the Competition Council. This means that, even if the undertakings cannot go directly to court, the final decision may nevertheless be in the hands of the judiciary. In a context where the judiciary has so far had a low level of exposure

390

Working Paper VII

to competition enforcement issues, there is a risk of lack of coherence in the competition enforcement process. In Romania the prior notification and authorisation system did not result so far in an overwhelming number of notifications. Most of the notifications referred to the block exemptions, where the analysis was more or less formal, involving the examination of the restrictive clauses exempted (this was the list of other obligations that normally do not restrict competition, and the 'black list' that disqualifies agreements from the exemptions). At the same time, the number of notifications and applications for an individual exemption—that requires a careful analysis of complex economic arguments—was insignificant:

Non—intervention requests —Granted —Prohibited —Case stock at the end of the year Requests for individual exemption —Granted —Prohibited —Case stock at the end of the year Notifications for block exemptions —Granted —Prohibited —Case stock at the end of the year

1997

1998

1999

6 4 1 1 1 — 1 — 9 4 2 3

8 5 1 2 1 — 1 — 22 17 3 2

10 7 — 3 2 — — 2 37 30 — 7

The current system of compulsory notification to obtain a block exemption is quite close to a 'stamp of approval' for agreements that do not raise concern, otherwise they would not have been 'notified' to the Commission. Moreover, in one case, this centralised authorisation system was used to block a private action before national competition authorities, when the Competition Council had given a block exemption to an agreement that was investigated by the Competition Office. During the period 1997-1999, the Competition Council analysed 877 cases: 171 cases in 1997, 231 cases in 1998, and 475 cases in 1999. Over this period the cases related to anti-competitive agreements prevail (231 cases), although they consistently decreased over time: from 40% of the total cases in 1997, to 32% of the total cases in 1998, and to 19% of the total cases in 1999. Simultaneously, the number of cases addressing economic concentrations increased continuously. In 1999 they reached 171 cases, compared to 10 cases in 1997 and 50 cases in 1998. Their weight increased from 6% in 1997 to 22% in 1998, reaching 36% in 1999. Thus, if we take into consideration the compulsory notification system in cases of economic concentration, we can conclude that the administrative work implied by the notification system diverted the Competition Council from a more active approach oriented towards an in-depth analysis of restrictive

2-Coherence

391

agreements and abuses of dominance. In 1997 and 1998, there were as no cases related to Articles 5 and 6 (81 and 82 EC Treaty) started by the Competition Council ex officio.

II. The Association Agreement The political and economic transformation of the countries of Central and Eastern Europe prompted the European Community to modify its sorts of commercial relationships with this region and to conclude free-trade agreements with the individual countries. The main instrument—for Romania, as for the other countries—is the European Agreement that establishes, on a bilateral basis, the new commercial framework between the European Communities and the Member States on one side, and each of the associated countries on the other. The rules on competition are provided in Article 64 (1): 'The following are incompatible with the of the Agreement when they may affect trade between Romania and the Community: —all agreements between undertakings, decisions by associations of undertakings, and concerted practices that have as their object or effect, the prevention, restriction or distortion of competition; —any abuse by one or more undertakings of a dominant position within Romania or the Community, or in a substantial part of it.'

The Association Agreement provides that these rules and forbidden behaviours should be interpreted in accordance with Articles 81 and 82 EC Treaty. In addition, the associated countries should observe the case law of the Commission and the jurisprudence of the European Court of Justice. However, on a more detailed analysis, the provisions of the Association Agreement referring to competition do not show a perfectly coherent philosophy, due to some lack of concrete details and several ambiguities regarding the enforcement process. The Association Agreement might be included in the category of free-trade agreements (although an incomplete one, because trade liberalisation in agricultural goods is not a firm objective and does not establish a precise timeframe within which to be enforced). From this perspective—that of a free-trade agreement limited to industrial goods—the requirements related to the protection of competition appear to be tougher when compared with other free-trade agreements, such as NAFTA. However, we could imagine a couple of possible explanations for the introduction of the rules on competition into the Agreement: • the Agreements were negotiated at a time when competition policy started to be seen as a crucial element for the liberalisation of international trade and as a potential substitute for traditional commercial policy measures; • the Community was concerned with the risks of participating in a free-trade area including countries without a fully-fledged market economy; such a

392

Working Paper VII

concern might have been the reason for 'adding' competition rules to the normal commercial defence instruments. According to the wording of the Association Agreement, specific anticompetitive agreements and behaviours are 'incompatible' with the proper functioning of the Agreement when they might affect trade between Romania and the Community, but are not 'forbidden'. Therefore there are opinions in the sense that it is improbable that the competition rules in the European Agreements have the precise and unconditional character necessary to have a direct effect. As a matter of consequence, it is doubtful whether a private entity would invoke these rules in court. The Commission's White Paper on the Preparation of the associated countries of Central and Eastern Europe for integration into the Internal Market of the Union showed that there is no binding obligation for the Member States to adapt their legislation to that of the Community in areas that are of the Commission's competence. At the same time, the associated countries are required to adopt a competition enforcement system similar to that of the Community. Therefore, several conclusions can be drawn: • the Europe Agreements provide that the associated countries should adopt rules concerning competition that are similar to those of the Community. Such obligation is stricter than that falling upon the existing Member States; • applicant countries should harmonise their legislative framework with EC competition law in as much detail as possible; • the harmonisation obligation includes the jurisprudence of the European Court of Justice. Or, competition rules are enforced in cases that affect the domestic undertakings only, but also in those where bilateral trade is affected. One may draw the conclusion that the national competition authorities may ignore Community law only when bilateral trade is not affected. This means that a 'double enforcement standard' could appear in function of whether a Community undertaking is involved/affected or not. There is a certain degree of confusion between the obligations that the accession countries should observe in order to comply with the provisions of the European Agreements and those that should be observed from the perspective of a (future) Member State. In fact, the natural order seems to have been reversed: the competition law systems of the applicant countries are evaluated exclusively from the perspective of their conformity with those of the Member States, while some necessary qualifications are made only subsequently. Yet, because the respective country is not yet a Member State, it does not have the institutional framework to apply the Community rules. The institutional framework established by the Europe Agreements also has its own degree of ambiguity, and is rather 'weak' when compared with the mechanisms in place in the case of other agreements concluded between the Community and third countries. The European Economic Area Agreement for instance provides for a full exchange of information between the EFTA

2 - Coherence

393

Surveillance Authority (ESA) and the Commission in cases concerning both the EFTA countries and the Community. The ESA and the Commission are entitled to formulate opinions in proceedings brought before the other authority. This was made possible by adding a Protocol to the EEA Agreement allowing for derogation from the Community rules on confidentiality. The Europe Agreements instead provide that the Commission and the competition authority of each associated country undertake, in particular, to notify one another of the cases they are handling that also concern the other authority. However, there is no obligation to communicate information where this would be prohibited by law or would be incompatible with the parties' interests. Furthermore, each of the authorities undertakes to take into account the observations of the other, and to seek mutually acceptable solutions. In summary, one of the few advanced elements in the Agreement is the provision for 'positive comity', while functional co-operation does not include the possibility for the other competition authority—before a final decision is taken by the other party—to make any observations it considers appropriate. The exchange of information does not include the sharing of information protected by confidentiality rules. The limits to co-operation between the applicant countries and the Community are probably not due mainly to the current legislation of the applicant countries (which is by now more or less harmonised with EC law), but to the differences in terms of enforcement between the two parties. These differences are probably the reasons why, despite the adoption of competition rules similar to those of the EU in the associated countries, the Europe Agreements did not make it possible to abandon trade protection instruments. Moreover, the settlement of disputes between the parties is mainly a political negotiation: the Association Council can make 'recommendations', while the parties have the possibility of applying 'adequate measures'.

III. Romania and the proposed reform of the rules implementing Articles 81 and 82 EC Treaty In Romania, the main problem with the existing notification and prior authorisation system is that it strains the enforcement capacity of the competent authorities. As already mentioned before, current Romanian competition rules ask for compulsory notification in order to obtain even exemptions for agreements falling under block exemption regulations. The Romanian competition authorities—like, probably, the competition authorities of other transition economies—have limited budgets and limited experience. It is almost certain that, if all transactions that should be 'notified' under the current regime were in fact 'notified' (including those qualifying for exemption by categories), the competition authorities would be overwhelmed. In such circumstances, it would be hard for the competition authorities to weed out the 'good' cases from

394

Working Paper VII

the 'bad' ones. Even as the case stands, the scarce resources of the competition authorities are diverted from the really important enforcement matters. Current experience shows that many agreements should be 'notified' but are not. This is mainly due to ignorance, and most of the agreements that are not notified are not really problematic. Yet the current system makes law-breakers out of the parties to the agreements if they do not notify to the authorities their agreements. In this case, the authorities have two choices: • To prosecute for not notifying—yet this option is not always attractive, especially in the case of agreements whose effects are 'neutral' from the competition point of view, or are even pro-competitive; or • Not to prosecute—yet this is not a positive attitude either, especially when one of the main objectives is to promote the rule of law. To the contrary, experience has shown that most of the agreements that are notified are not problematic. In an environment where ignorance of the competition law is common, those few undertakings that notify the authorities are those who have consulted knowledgeable counsels. It is to be expected that lawyers would have advised them not to notify an illegal agreement. If very few 'notified' agreements are problematic, then the costs of the notification system probably exceed the benefits it confers. Economies in transition are, in some ways, in the position experienced by Western Europe in the 1960s. Most governments and businesses do not understand or appreciate competition law. Competition authorities therefore need to be free to address that issue, rather than being diverted from that important task by a flood of paperwork.

1. Some comments on the options ahead From a certain point of view, the decentralisation of the application of 81 (3) is likely to lead to the re-nationalisation of competition policy. However, we should not forget that in the accession countries the enforcement procedure is already decentralised. In the short-term it would be unreasonable to expect that the relatively new and inexperienced national competition authorities of the accession countries will apply competition rules uniformly and with an eye to Community law. In all likelihood, they will look at competition issues first from the point of view of national law, rather than Community law. If this is a serious risk in the existing Member States, it will be furthermore so in the applicant countries. At the same time, the decentralised enforcement system proposed by the Commission in the White paper represents a shift from a regime of 'what is not authorised is forbidden' to one where 'what is not forbidden is authorised'. The current enforcement system corresponds in a certain sense with the type of regulation prevalent under the old communist regimes, and which still retains a

2 - Coherence

395

strong hold in the post-communist mindset. The old system of applying 81 (3) simply reinforces that mindset, and this is not what is needed in order to develop a vigorous economy. We have had enough of that mindset coming from within as it was. Given the relative lack of 'sophistication' of the competition authorities in many accession countries, the current prior authorisation system presents for them a higher risk that anti-competitive agreements will end up by receiving a 'stamp of approval' than in the current Member States. When the anticompetitive nature of the agreement eventually becomes apparent, this damage will not be so easily undone. Removing the 'stamps of approval' makes it easier to challenge an agreement ex post facto, once its anti-competitive nature becomes apparent.

2. A decentralised enforcement system The abandoning of the notification system opens the question of how can national authorities (and the EU) learn of anti-competitive agreements. Of course, the really 'bad' agreements—naked cartels, for example—are not notified anyway, and legal action against them relies on private complaints and court litigation. However, in the East undertakings are still even more reluctant to complain against the illegal behaviour of their competitors than in the West. This is probably even more of a problem in applicant countries than in the West, because it is combined with a low notification rate. Moreover, apparently the threat that an illegal agreement will be declared void ab initio is not a serious deterrent of anti-competitive behaviour either. The expectation that the notification system would lead to more agreements being notified—and therefore reviewed—has not worked in practice. The new regime is probably better. It places the responsibility to inform the authorities of an illegal agreement on the parties that are injured by it. Once clear rules will been developed in the EU, they can be imported wholesale into the applicant countries, and there is little risk of uncertainty on the part of business. This is clearly the approach of the Commission: 'The translation .. . should ideally be done using the same concepts and terminology as in the EC. Every attempt to re-formulate or re-interpret the notions used by the Treaty and by secondary legislation is likely to raise dispute in the future .. .'4 One of our priorities should be to educate the business community in the applicant countries about competition law. Competition authorities will need to seek out opportunities, through the media and otherwise, and explain what is legal and what is not. 4

Roland Kobia, 'The admissibility and monitoring of State Aid—The competencies and role of the State aid monitoring authority', Fourth Competition Conference of the Central and Eastern European Countries and the European Commission, Bratislava, 24-26 May 1998.

396

Working Paper VII

3. Enforcement under the new regime A certain degree of caution about empowering national courts in the applicant countries to enforce competition rules is justified. Judges in the accession countries are still inexperienced in applying business law, and the resources available to them are limited (in many cases they have no law libraries, no Internet access and no training in competition law). Moreover, while the civil law system in the Member States has gradually evolved towards a 'common law' approach which enables them to consider and even apply legal precedents from EU courts and the national courts of other Member States, in most accession countries such evolution has not begun. Consequently there is a serious risk that national courts in applicant countries will apply the law in a rigid, mechanistic way, without regard to maintaining consistency with earlier decisions adopted in the EU (there is already some evidence of such trend in Romania). Not the same caution should be adopted, however, with respect to competition authorities in the accession countries. While admittedly there are differences between the competition authorities of the accession countries, in general it can be said that they all are quite well acquainted with EU competition law and have the resources to apply it. However, they have a long way ahead in order to assimilate the practice and the case law of the Commission and Court of Justice. In Romania, for instance, only a few administrative and judiciary bodies are aware of this obligation, and fewer started to do anything about it. Moreover, there are no mechanisms of co-operation between the courts and the national competition authority/authorities (in the case of Romania, the Competition Council). To give just one example, the staff of the Romanian Competition Council is forbidden by law to provide expert testimony before the courts. Of course, the courts of the applicant countries will face difficulties when it comes to establishing the relationship between national and Community competition rules. However, such problems can only diminish in the context of the new enforcement system (whereby the law may be actually applied) as opposed to the old notification system (which blocked the actual application of the law). The possibility of an appeal (when based on Community rather than national law) to the EU Courts will go a long way towards reducing the possibility of inconsistencies in enforcement. The White Paper does not seem to allow for the possibility of an appeal to the Commission, and perhaps it should. The national competition laws of most applicant countries provide for the possibility to appeal the decisions of the competition authorities to national courts, and, for the reasons described above, these national courts might not be fully competent to handle the appeals. To conclude, the reform proposed by the Commission will help the Romanian Competition Council to re-focus its enforcement activity and ease the administrative burden on undertakings. Though from the point of view of consistency with the Community competition rules the new enforcement sys-

2 - Coherence

397

tern might be a problem in the applicant countries in the short term, it should nevertheless not be a cause of concern in the medium-term if 'in the context of pre-accession strategy, the Commission will devote particular attention to the development of competition in the candidate countries and will provide their competition authorities with increased assistance'.5 This is probably the reason why Romania accepts the acquis in full under the Chapter on Competition Policy and 'will not apply for special treatment or transitional periods in this field.'6

5

European Commission: White Paper on Modernisation of the Rules Implementing Articles 85 and 86 of the EC Treaty, OJ C132 of 12.5.1999, at p. 37. 6 Romania, Screening Results: Chapter 6—Competition Policy.

VIII Emil Paulis1 Coherent Application of EC Competition Rules in a System of Parallel Competencies

I. The importance of the issue The tasks of the Community are to establish a common market, to approximate the economic policies of Member States and to promote the harmonious development of economic activities throughout its territory (Article 2 EC Treaty). The rules of competition play a central role in achieving these goals, as they seek to ensure that competition in the common market is not distorted (Article 3 EC Treaty). Distortions of competition jeopardise the functioning of the common market by preventing the harmonious development of economic activities. Articles 81 and 82 EC Treaty are the rules of substance applicable to undertakings designed to ensure an undistorted market. Incoherent application of these rules can provoke a distortion of competition in two ways: first, by applying these rules to some companies but not to others, with the result that some are forced to obey the law while others are not; second, by applying the rules differently to companies operating in the same market. Both situations involve discrimination that can lead to a distortion of competition on the market and thus put at risk the proper functioning of the common market. In a narrower sense, incoherent application means that the same company is faced with diverging decisions that it cannot implement because they contain conflicting obligations and rights. In their replies to the White Paper, commentators have mainly focused on this type of incoherence, which they claim would be increased in a system of full parallel competencies where the Commission, national competition authorities and national courts would all have the power to apply Articles 81 and 82 EC Treaty. All of these types of incoherent application of EC competition law exist in the present system because Articles 81 (1) and 82 and Block Exemption Regulations are already directly applicable by national courts and national competition authorities. However, it is said that at present the Commission's monopoly over the application of Article 81 (3) guarantees coherent application at least in the field of Article 81. This is the most difficult Article, because it contains a negative rule and a positive rule, and it thus contains a greater potential for conflicting decisions than Article 82. Those who voice these concerns request the maintenance of the Commission's exemption monopoly. They All opinions expressed in this article are personal.

400

Working Paper VIII

think that its abandonment will lead to a re-nationalisation of competition law, whereby the Commission would no longer have the means to achieve the Treaty objective of ensuring a system of undistorted competition.

II. The discrimination created by the Commission's exemption monopoly In the early days of the Community there was no competition culture, no internal market and no body of case law by the European Court of Justice and the Commission. The risk of incoherent application of the law was therefore considerable. At that time, the Council made the right choice in entrusting the Commission with the exclusive power to exempt restrictive agreements. This system of enforcement worked well for the initial phase of the Community by creating an EC competition culture and by guaranteeing the coherent application of the EC competition rules. However, over time the exemption monopoly has started to handicap, rather than benefit, the efficient enforcement of EC competition rules. Today, the exemption monopoly still serves the coherent application of the rules, but it is also the reason for under-enforcement of the rules and unequal treatment in the application of the rules. Therefore, we must ask ourselves the question: what is the value of a system that guarantees coherence if that system allows anti-competitive cartels to be maintained and benefits those who violate the law to the detriment of consumers? The present system creates distortions of competition and incoherence through the unequal application of Article 81. There is coherence only for those who get an Article 81 (1) or Article 81 (3) decision (this concerns few cases per year), but not for those who are the victims of unpunished infringements or for those who cannot afford to notify and wait for an Article 81 (3) exemption decision. The present system leads to discrimination in respect of those who get no redress for violations and in respect of those who have no means of getting the benefit of an Article 81 (3) decision. This double discrimination goes more against the public interest than does the possible risk of incoherence in the application of the law (incoherence in the narrow sense). If the exemption monopoly is abandoned, a certain price might be paid in respect of the latter type of incoherence, but this will benefit the stronger enforcement against anti-competitive cartels and will benefit the improved civil enforceability of all those contracts which fulfil the exemption conditions under Article 81 (3). That is a very positive contribution to the maintenance of undistorted competition and to the equal treatment of companies operating in the common market. There is no perfect reform. In all reforms, the risks have to be balanced against the efficiencies engendered by the reform. In the new system of legal exception and parallel competencies, the risk of incoherent application of the EC competition law is in certain respects increased compared to the present sit-

2 - Coherence

401

uation of an exemption monopoly. However, the new system will provide more equal treatment in the application of the rules in three ways: (i) A more efficient enforcement of the prohibition rule against cartels and abuses of dominant positions through strengthened public enforcement by the Commission and 15 national competition authorities, and through private enforcement in national courts; (ii) The direct applicability of the exemption rule of Article 81 (3) will legalise, without the need of a Commission exemption decision, all those agreements that are pro-competitive on balance; and (iii) Strengthening the effect of the rule of the primacy of EC law over national competition laws will create greater convergence in the application of national and Community competition rules to the benefit of companies. All three effects contribute to a more coherent application of EC and national competition laws by ensuring undistorted competition in the common market.

III. The broader context of the enforcement of Community law The issue of the coherent application of EC competition law in a system of parallel enforcement competencies has to be seen in the context of the general system of enforcement applicable to Community law. The whole Community legal order is based on the application of Community law by national authorities and courts. This is the case for Treaty provisions (such as, rules on the free circulation of goods, services, persons and capital) and for secondary legislation (for example, all the internal market legislation). From the very beginning, the Treaty has opted for the decentralised application of Community law and has accepted a certain risk of divergent application resulting from decentralised enforcement. In Van Gend & Loos,2 the Court of Justice founded the direct applicability of Treaty provisions on reasons of efficiency in the enforcement of Community law: public and private enforcement should be combined to ensure the most efficient application of EC law. Possible divergent applications resulting from decentralised enforcement can be prevented or corrected by two mechanisms foreseen in the Treaty: • The mechanism of Article 234 (preliminary questions), which is not a Community appeal system but a system of co-operation between national courts and the European Court of Justice; • Infringement actions against Member States under Article 226 (ex-Article 169) which can be brought by the Commission.

Joined Cases 98/83 and 230/83 Van Gend & Loos v Commission [1984] ECR 3763.

402

Working Paper VIII

While the Court of Justice has the task of ensuring the correct interpretation and application of the provisions of the Treaty and other Community legislation in the decentralised enforcement of Community law (Article 220 EC Treaty), the Commission is the guardian of the Treaty and has been entrusted with the task of ensuring that Community law is applied (Article 211 EC Treaty). In certain fields of Community law, the Commission has also been granted the power to apply Community law, either alone or in parallel with Member State authorities and courts. This is particularly the case in the field of competition law (see Articles 81, 82 and 86 EC treaty). Under the ECSC Treaty the Commission has exclusive competence in the field of competition (Articles 65 and 66 ECSC Treaty). In the EC Treaty applying to all sectors of the economy, this system was abandoned. From the beginning, national courts, national authorities and the Commission were competent to apply Articles 81 and 82. It was only Regulation 17/62 that granted the Commission exclusive power for providing exemptions under Article 81 (3), thereby excluding national courts and authorities. While this exclusive power was the right solution for shaping policy and ensuring coherent application at the beginning of the enforcement of Article 81, after forty years of enforcement and after the considerable growth of the Community, this exclusive power has become an obstacle to efficient enforcement. It has led to under-enforcement and to the fact that Member States apply national competition laws and not Community competition law. This situation is likely to lead to distortions of competition inside the common market and to affect the coherent application of Community competition law. This evolution is incompatible with the development of the internal market. Thus, the decentralisation of Article 81 (3) has to be seen in the broader context of the general system of enforcement applicable to Community law where the existing exemption monopoly of the Commission is rather an anomaly and where the maintenance of that monopoly promotes the application of national competition laws. This system leads to a re-nationalisation of competition law, with a higher risk of incoherence than a system of parallel competencies promoting the application of EC law to cases affecting trade between Member States. National laws still diverge in substance and are not subject to a uniform interpretation under Article 234 EC Treaty. In addition, under the present system national law is only constrained by the rule of primacy if the Commission adopts a formal decision. We are therefore convinced that, on balance, the new system will lead to more coherence in the overall system of national and Community competition rules than does the present enforcement system. It will certainly be in line with the development of the Community towards a more integrated market and with the need of companies to trade on a more level playing field.

2 - Coherence

403

IV. General framework to enhance coherence in the new enforcement system There remains the risk of conflicting decisions in the same case (incoherence in the narrow sense). In a system of full parallel competencies, the risk of conflicting decisions would be likely to increase. As regards Article 82 EC Treaty nothing changes, because that provision has always been applied concurrently by the Commission, by national competition authorities and by national courts. Forty years of application of that provision has shown that conflicts are far less frequent in practice than textbooks would suggest. However, Article 82 has the advantage of being a straightforward prohibition rule without the possibility of exemptions that can conflict with prohibition decisions. Under Article 82, conflicts can only arise as to the finding or non-finding of an infringement. The potential for conflicting decisions is thus more limited and less visible. Article 81 contains a prohibition rule and an exemption rule. Its potential for conflicting decisions is therefore higher than for Article 82. The hitherto wide interpretation of the scope for the application of Article 81 (1), combined with the exemption monopoly of the Commission, has severely constrained decentralised enforcement and thereby reduced the risk of conflicting decisions under this provision. However, there have been conflicts between prohibition decisions on the one hand, and negative clearance decisions on the other hand, and under Article 81(1) or exemption decisions under Article'81 (3). There have also been conflicts between decisions based on EC law and decisions based on national law. The possibility of these conflicts occurring in the same legal order, or between two different legal orders, cannot altogether be excluded when there is parallel application of the same law by several decision-makers, or when there is parallel application of different laws to the same subject matter. The issue is how to reduce this risk to an acceptable level. In the past conflicts have been resolved through co-operation and the rule of primacy. The whole Community legal order is based on the principle of cooperation enshrined in Article 10 EC Treaty (co-operation for the achievement of the Community's objectives) and on the principle of primacy of Community law over national law. Co-operation and primacy are two fundamental principles designed to ensure the coherent application of Community law. These two principles should also resolve possible conflicts in the new system where Article 81 (3) would become fully applicable by national competition authorities and courts. In addition to the principles of co-operation and primacy, the new enforcement system will create a general framework that enhances coherent application. This framework contains the following elements:

404

Working Paper VIII

1. Widest possible application of EC competition law Application of the same substantive law is the best basis for coherent treatment of competition cases. The risk of incoherent outcomes is, by definition, greater in a system where different substantive rules are applied. Coherence is greatly facilitated if all decision-makers apply the same law or, at least, laws that pursue the same objectives. In the Community of today, the conditions for coherence are much better than they were at the beginning. Indeed, most Member States and all candidate countries for enlargement have already adopted national competition laws modelled on the rules of Articles 81 and 82. However, there remain differences, such as in the field of vertical restraints, and there is a risk of divergent applications of national competition laws since they are not subject to the Article 234 mechanism. The more national competition authorities and national courts can and do apply EC competition law to cases affecting trade between Member States, the better coherence will be ensured for the benefit of the Community. For these reasons, we should not narrow the scope of application of Articles 81 and 82 but continue to adopt a wide interpretation of the condition that trade between Member States should be affected. It is in the interest of coherence to keep the widest possible scope of application to the Community competition rules. To the extent that Member States, in spite of having been empowered to fully apply EC rules, continue to apply national competition law in cases affecting trade between Member States, the rule of primacy should be applied to the widest extent possible in order to bring about convergence in the application of EC and national competition laws. We will return to this issue in Section VII.

2. The rules on substance must be made as clear as possible The clearer the rules on substance, the lower the risk of incoherent application. The Commission is engaged in a comprehensive revision of the EC rules on substance. It has introduced a new generation of Block Exemption Regulations that exempt, up to a given market share threshold, all restrictions with the exemption of a limited list of hardcore restrictions and conditions. The approach of the new Block Exemptions is no longer to prohibit all restrictions except those that are explicitly exempted, but rather to exempt all restrictions except those that are explicitly prohibited. This new approach focuses on what is prohibited. This makes the rules clearer and simpler. The Commission also issues guidelines for the assessment of cases not covered by Block Exemption Regulations. The Commission will further clarify the rules by adopting individual decisions that will also focus on what is prohibited. As regards Article 81 (3) more specifically, the Commission monopoly has largely prevented the development of a body of case law on the interpretation

2 - Coherence

405

of that provision. In the new enforcement system, the wide margin of appreciation left to the Commission under article 81 (3) will necessarily be narrowed down for two reasons: first, because there will be more decision-makers, Article 81 (3) will become 'justiciable' like any other directly applicable Treaty provision. There will be stricter judicial control over the application of that provision in the same way as the Court of Justice controls the application of Articles 81 (1) and 82. Second, there will no longer be constitutive exemption decisions with effect for the future which involve a prediction of potential effects for which the margin of appreciation is always wider than for declaratory decisions applying the law to existing or past facts. Also, other objectives of the Treaty can only be taken into account if they can be subsumed under the conditions of Article 81 (3). A number of commentators, the ECOSOC and certain Member States, have asked the Commission to publish a set of guidelines on the criteria of application of the four conditions of Article 81 (3). In my view, such guidelines would greatly assist national authorities, national courts and companies in applying a provision that hitherto has only been applied by the Commission. Clarification of the law is an essential element in ensuring coherent application, but we must take care not to fall back into a situation where we define restrictions of competition in a legalistic form, irrespective of their negative or positive effects in the market. Competition law is not tax law. One can work with a limited number of per se prohibitions (hardcore restrictions) and per se permissions in the form of Block Exemption Regulations. In between these two categories, a correct assessment requires a case-by-case approach that focuses on the economic effects on the market.

3. Commission keeps power to enforce itself the rules In most fields of Community law, only Member States and national courts have the power to apply the EC rules and the Commission has only the role of guardian of the Treaty with the power to bring infringement actions against Member States under Article 226. In the field of competition, the Treaty and the Council have granted the Commission the power to itself apply the rules to companies and Member States (see Articles 85, 86, 87-89 and the present Regulation 17/62). This autonomous power of direct enforcement is an essential tool to establish competition policy and to contribute to a coherent application of the EC's competition rules. The Commission maintains the power to adopt decisions that have a direct effect on companies. These decisions will develop all the effects of Community acts as set out in the Treaty and in the jurisprudence of the Court of Justice, in particular in the Walt Wilhelm case3 as regards primacy and in the Delimitis case4 3 4

Case 14/48 Walt Wilhelm v Bundeskatellamt [1969] ECR-1. See also Case 48/72 Brasserie de Haecht II [I973] ECR 355.

406

Working Paper VIII

as regards the obligation of national courts to avoid conflicts with envisaged or existing Commission decisions. We will return to these principles in sections VII and VIII below. History has shown that Commission decisions have a strong precedent value beyond the individual case involved. Combined with the existing case law built up over 40 years, the power of direct enforcement by the Commission will play a major role in the maintenance of a coherent application of the rules. While the European Court of Justice is the ultimate guarantor of coherence in the Community legal order, it is not the role of the Court to determine competition policy. The Commission will have to set policy by adopting horizontal policy instruments as well as individual decisions. The tasks of the European Court of Justice and the Commission are in fact complementary. Both carry the specific responsibility of ensuring a uniform application of the EC competition rules.

4. The Commission's monopoly over the application of Article 81 (3) will be replaced by co-operation The exemption monopoly is not the only system that can ensure uniform application of EC competition law. As already explained, decentralised enforcement by national courts and authorities, combined with an obligation of cooperation and the rule of primacy, is the common system chosen by the EC Treaty for the enforcement of all other Community provisions. As regards national courts, the main co-operation mechanism is in place: it is the 234 procedure by which national courts may and, if deciding as a court of last instance, have an obligation to co-operate with the Court of Justice for questions about the interpretation and validity of Community law. The Commission intends to add to that mechanism the possibility for national courts to also co-operate with the Commission; it intends to hold itself available to assist national courts with non binding opinions on questions of law or fact relevant to the assessment of a particular case.5 The Commission cannot provide authoritative and binding answers to legal questions as that is the prerogative of the Court of Justice. However, since the European Court of Justice only replies to questions of law, and since competition law is heavily fact-based, an opinion from the Commission can constitute a useful complement to the 234 procedure, particularly for non-specialised judges. Horizontal co-operation between national courts in applying Community law will follow the general rules established by legal instruments such as the Conventions of The Hague and the Convention of Brussels, regulating judicial 5 Such co-operation exists already under the present enforcement system (see European Commission: Notice on cooperation between national courts and the Commission in the application of Articles 85 and 86 of the EEC Treaty, OJ C 39 of 13.2.1993).

2 - Coherence

407

co-operation in the field of civil matters as regards assistance, and recognition and enforcement of court decisions. In future it might become necessary to further promote horizontal co-operation between courts of different Member States. However, it is not justified to undertake this work only for the transfer to national courts of the competence to apply Article 81 (3). With regard to national competition authorities, close and constant cooperation between national authorities and the Commission is already operating under Regulation 17/62 and the 1997 Notice on Co-operation with National Competition Authorities.6 In the new system of full parallel competence, this co-operation will be further intensified. It is the intention of the Commission to create a network supported by an Intranet of National Competition Authorities that should allow an instant and simultaneousflowof information between EC competition authorities. Inside that network, regular information, consultation and discussion about cases at the level of case handlers and bosses is essential to create a common competition culture. This is the best way to guarantee a coherent application of the Community rules. The Advisory Committee on Restrictive Practices and Dominant Positions will no longer operate only for vertical consultation, but will also become a forum for horizontal co-ordination of policy. In the establishment of these co-operation mechanisms, it is important to achieve the right balance between the co-operation necessary for the proper functioning of the new system and the need to avoid excessive bureaucracy. Too much bureaucracy could prevent the Commission and national authorities from concentrating on the prosecution of the most serious infringements and could incite national authorities to apply national competition law in order to circumvent the obligations connected with the application of Community law. The intense co-operation foreseen for the application of Articles 81 and 82 does not exist in most otherfieldsof Community law. It will play a vital role for a coherent and efficient enforcement of the EC competition rules. Co-operation is not just a faculty, but an obligation under the EU Treaty. Article 10 of the Treaty contains an obligation of co-operation between the Member States and the institutions of the Community to achieve the objectives of the Treaty. One such objective is that of ensuring that competition is not distorted on the common market. To achieve this objective with the co-operation of the Member States, the Commission has been entrusted with a special responsibility set out in Article 211 of the Treaty: 'In order to ensure the proper functioning and development of the common market, the Commission shall. .. ensure that the provisions of this Treaty and the measures taken by the institutions pursuant thereto are applied'. The Commission is thus obliged to ensure that the provisions of Articles 81 and 82, and decisions adopted pursuant to these articles, are applied 6

European Commission: Notice on cooperation between national competition authorities and the Commission in handling cases falling under the scope of Articles 85 and 86 of the EEC Treaty, OJ C 313 of 15.10.1997.

408

Working Paper VIII

in an effective and coherent manner. To that effect, the Commission has the power to adopt decisions applying Articles 81 and 82, and it has the power to bring actions against Member States for non-compliance with Treaty provisions or measures taken pursuant thereto.

5. Preventive and corrective mechanisms to avoid or solve conflicts of decisions The Commission will not give up its exemption monopoly without creating certain preventive and corrective mechanisms to avoid or solve conflicts arising from decisions. These mechanisms (explained in Sections V and VI below) are part of the overall framework under which the Commission proposes to fully associate national courts and authorities in the application of Article 81. In this respect, it will be important to find the right balance between measures that are necessary for the good functioning of the new system, and the need to avoid putting national jurisdictions and authorities under excessive supervision by the Commission. This could be incompatible with the independence of judges and with the objective of stimulating the application of Community law to cases affecting trade between Member States. While coherent application is vital, some degree of competition between the decisionmakers under the ultimate control of the Court of Justice could be beneficial for the development of the law.

6. Elimination of constitutive exemption decisions In the new legal exception system it is essential that neither courts nor competition authorities should be empowered to adopt constitutive exemption decisions which would be based on ex ante control following requests by companies. Constitutive exemption decisions create rights for the past and the future, irrespective of changes in the facts over time. Such decisions can be opposed to all third parties and thus have erga omnes effects. Creating rights for the future with erga omnes effects creates a high potential for conflict with prohibition decisions adopted by other decision-makers in a system of parallel competencies. There could frequently be diverging decisions creating conflicting rights and obligations. A system of constitutive exemption decisions is only workable with an exemption monopoly or with such strong co-ordination by the Commission that there would no longer be real decentralisation. National exemption decisions of a constitutive nature would pose major threats to a coherent application of EC competition law: • They entail the risk of being abused to protect national champions and, unlike prohibition decisions, they are not likely to be challenged in court; to

2 - Coherence

409

make constitutive exemption decisions consistent could create considerable tensions between Member States and the Commission; • Given their erga omnes effects, any inconsistency between a constitutive exemption decision adopted by one national authority and a prohibition decision adopted by another national authority would be much more detrimental to the uniform application of EC law than a rejection of a complaint in one country and a prohibition decision in another country; a rejection of a complaint has only an effect rebus sic stantibus for the competition authority which has rejected the complaint. • If national exemption decisions were given a Community-wide effect (provided that would be politically acceptable), then there would be a problem of association for other Member States in the decision-making procedure at the national level. The same problem would exist in a system of mutual recognition of national exemption decisions. Extra-territorial effect or mutual recognition would force the Commission to control each national exemption decision, the effect of which would be that the Commission would continue in the business of exemption decisions and decentralisation would loose much of its benefits. The number of exemption decisions adopted by national authorities could be considerable, so that the co-ordination work imposed on the Commission would be increased accordingly. • If national exemption decisions did not have EU-wide effect, they would be of little value and companies would have to notify their agreements in several Member States. This would lead to multiple notifications and to a considerable risk of conflicting decisions. On the contrary, prohibition decisions do not cause the same problem because their de facto extraterritorial effect is usually sufficient to solve the competition problem without conflicting with a formal exemption decision adopted by another cartel authority. Thus the system chosen by the Commission in the White Paper (excluding constitutive exemption decisions) considerably lowers the risk of incoherent application only of Article 81. The system of legal exception that gives Member States the power to adopt prohibition decisions or non-binding declaratory decisions is, in our view, more suitable for a coherent application of EC law than a system where Member States could adopt constitutive exemption decisions with erga omnes effect. The fact that national courts can also apply Article 81 (3) bears its own risk of incoherent applications in light of the number of national courts. But this risk is more limited in scope because: • National courts are not given any power to adopt constitutive exemption decisions with effect for the future; they only apply the law to past and present situations. • National courts have no power to decide with erga omnes effect; their decisions only have effect inter partes, even if executed in other Member States. • National courts, unlike national authorities, are under the unifying control of the European Court of Justice (Article 234 mechanism).

410

Working Paper VIII

Also, it would be wrong to exaggerate the risk of incoherence in connection with courts because each lower national court is controlled by the Supreme Court of its own country and the Supreme Court has an obligation to refer questions of interpretation on Community law to the European Court of Justice. Therefore, there are ultimately only 15 Supreme Courts to be coordinated, which is not an insurmountable task. Article 82 has been applied by the Commission, the national competition authorities and national courts for over 40 years without any major problem of coherence. The same is true for Article 86, that contains a rule of prohibition and a rule of exemption like Article 81. This shows that the legal exception system can also work for Article 81 provided that no authority and no court can adopt a constitutive exemption decision.

7. A more reasonable application of Article 81(1) The interpretation of Community competition rules is undergoing a shift from a legalistic to a more economic based approach. For non-hardcore restrictions, this will limit the application of Article 81 (1) to companies holding a certain degree of market power. The economic approach will lead to a relaxation of the prohibition rule with regard to companies holding no market power while it will entail a stricter application for companies with market power. The number of Article 81 (1) cases will be reduced accordingly, and hence the number of cases where Article 81 (3) becomes relevant will also be reduced. The economic approach will therefore lower the risk of incoherent applications of Article 81 (3). In turn, the direct application of Article 81 (3) can also help to lower the risk of incoherent applications of Article 81 (1) by eliminating the effect of the rule of prohibition where such a rule would have been applied in too legalistic a manner. All in all, the new system should lead to more reasonable and more consistent outcomes by combining the analyses under Article 81(1) and (3).

V. Specific measures applicable to national competition authorities In addition to the general principles explained above, there are four specific measures that will help ensure the coherent application of the Community competition rules by national competition authorities. 1. Allocation of cases avoiding multiple control and forum shopping The network must work out a system of information and allocation of cases that achieves two goals: cases should always be dealt with by the authority best

2 - Coherence

411

placed to effectively solve the competition problem at issue, and cases should normally only be subject to the control of one authority. Multiple control and forum shopping should be avoided so as to reduce control costs to companies and authorities, and to eliminate the risk of conflicting decisions on the same subject matter. Parallel action by several authorities should only take place if such action is more efficient for the protection of competition and if the risk of conflicting decisions is minimal. Parallel action could be efficient when a clear infringement (such as, a cartel) is prosecuted by two or three neighbouring competition authorities with a view to adequately punishing the infringing companies. As regards other cases, only the best placed authority should act, determined by the extent of the geographic market and the capacity to effectively terminate and to appropriately punish the infringement at issue. Other authorities approached by complainants with the same subject matter should be empowered to suspend their proceedings or to reject the complaint on the ground that another authority is dealing with the same complaint.

2. Obligation to consult the Commission on prohibition decisions National competition authorities should be obliged to consult the Commission about proceedings that could lead to prohibition decisions, decisions accepting commitments or decisions withdrawing the benefit of a Block Exemption Regulation. These decisions are the most important tools of enforcement in the new system and have the strongest effects on companies. The network should not be overloaded by consultations on other decisions such as rejections of complaints or the closing of an ex officio infringement proceeding. These latter decisions only bind the acting competition authority and do not constitute a decision of compatibility in favour of the company concerned. Therefore, there is no real problem of coherence in respect of these decisions. National competition authorities should, however, always have the ability to request an opinion from the Commission in the same way as national courts can address the Commission with a request for an opinion (see in particular Article 211 EC Treaty).

3. The possibility for the Commission to withdraw a case In order to solve conflicts over the allocation of a case or in order to prevent an incoherent national decision, the Commission should keep the right to withdraw a case from national competition authorities, as it already can do under Article 9 (3) of the present Regulation 17/62. The Commission should only use this instrument in exceptional circumstances, but it is an indispensable tool to ensure the proper functioning of the network. However, it is important that the

412

Working Paper VIII

network functions on the basis of mutual trust and not on the basis of the Commission systematically playing the role of policeman or controller.

4. Possibility for the Commission to itself adopt a decision The Commission keeps its own power of enforcement. Since it has a special responsibility to maintain the coherent application of EC competition law (see Article 211 EC Treaty), it might have to apply the law itself by adopting a prohibition decision in a case where a national authority has rejected a complaint or closed an ex officio infringement proceeding. This might be a useful way to counteract a possible lack of enforcement by certain national competition authorities. Where a national authority intends to prohibit a practice that the Commission considers compatible with Articles 81 and 82, the Commission could either apply the withdrawal mechanism or leave the parties to challenge the national prohibition decision before the competent courts. The specific measures applicable to national competition authorities are thus principally preventive measures and only exceptionally corrective measures. We consider that the success of the network depends on obligations of consultation and on mutual trust. It is important that the network becomes a family sharing the same objectives and with every member pulling in the same direction.

VI. Specific measures applicable to national courts National courts applying Community law are co-ordinated by the European Court of Justice via the co-operation mechanism of Article 234. The Commission has no jurisdiction over national courts. However, the Commission has the power to bring actions against Member States for non-compliance with the Treaty (see Article 226). Such actions could be brought against Member States whose jurisdictions would systematically violate Treaty provisions or other Community measures adopted pursuant to Treaty provisions. The Commission also has its own decision-making authority to itself apply the rules of Articles 81 and 82. Since it is not proposed to give the Commission an Article 9 (3) power with regard to national courts, there might be more reasons for the Commission to itself apply the rules of Articles 81 and 82 in order to influence a coherent application of these rules by national courts. The effects of Commission decisions on national courts will be examined in the following sections. In addition to adopting decisions itself, the Commission also proposes to assist national courts in two principal ways: • First, either the national competition authority or the Commission itself will intervene in a national court proceeding as amicus curiae in order to present its view on a particular competition policy matter;

2 - Coherence

413

• Second, a national judge should have the possibility of requesting a written opinion from the Commission on questions of fact or law raised in proceedings relating to the application of Articles 81 and 82. National competition authorities could also be involved in such assistance work. Both measures are of a preventive nature and are aimed at contributing to a coherent application of the EC rules by national courts.

VII. Conflicts between two different legal orders: Walt Wilhelm and the principle of primacy Since national competition laws can also be applied to cases affecting trade between Member States, conflicts can arise between decisions based on national laws and decisions based on Community law. Article 83 (2) (e) empowers the Council to regulate the relationship between Community and national competition laws. The existing Regulation 17/62 has not regulated that relationship. In the absence of such regulation, the Court of Justice applied the principle ofprimacy of EC law over national laws in the 1969 case of Walt Wilhelm v Bundeskartellamt. However, since it was determined, the Walt Wilhelm decision has been the subject of diverging interpretations as a result of which there remain a number of uncertainties for companies, authorities and courts. In the Walt Wilhelm case, the Court of Justice explicitly referred to the possibility of the Council regulating the relationship between these two different legal orders in a Regulation based on Article 83 (2) (e). In our view, the new Regulation should be considered to set out the basic principles regulating the relationship between EC and national competition laws by giving answers to the following questions: a) Can national law authorise what is prohibited under EC law? b) Can national law prohibit or add conditions and obligations to an agreement that falls within Article 81(1) but that fulfils the conditions of Article 81 (3)? c) Can national law prohibit or add conditions and obligations to an agreement that falls within Article 81 (1) but that fulfils the conditions of a Block Exemption Regulation? d) Can national law prohibit an agreement that does not violate Article 81 (1) but that affects trade between Member States? These are difficult, but important questions, the answers to which impact directly on the uniform application of EC competition law throughout the Community and on the creation of a level playingfieldfor economic operators in the common market. If the application of EC competition rules could be blocked or modified by the application of national competition laws, this could

414

Working Paper VIII

jeopardise the proper functioning of the internal market and lead to a renationalisation of competition law. The goal of the Treaty—to create a common market where competition is not distorted—would be put at risk. By making the whole of Article 81 directly applicable and by giving the national authorities and courts the full power to apply Articles 81 and 82, the Commission wants to create an incentive for national authorities and courts to apply more EC competition law and to promote convergence between EC and national competition law. The more we all pull the same string, the stronger the effect of competition rules will be. The new system will bring primacy to bear not only where the Commission has adopted a decision—as is the case under the present system of constitutive exemption decisions—but it will also create the effect of primacy through the direct application of Article 81 (3). In future, agreements will be exempted by direct effect without any Commission decision being required, and this exemption will prevail over contrary national law. This will be the same situation as that of a prohibition under Article 81 (1), which has direct effect without any prior Commission decision declaring an agreement incompatible with that provision. The operation of this direct effect presupposes, however, that companies invoke the benefit of EC law before national courts and national competition authorities. Where national competition authorities apply national competition law to cases affecting trade between Member States, they have an interest in applying EC law in conjunction with national competition law so as to avoid any incoherent application. In Walt Wilhelm, the Court of Justice formulated a general principle of the primacy of EC competition law over national competition law. According to this principle, Member States may not adopt measures that would affect the 'effet utile' of Community competition law. To make the principle of primacy operational, the Court of Justice devised certain rules for national proceedings: • Where the Commission has adopted a decision, national authorities have an obligation to respect the effects of the decision by not adopting incompatible decisions. • Where the Commission has not yet adopted a decision but where a conflict of decisions could arise, national authorities must take the necessary measures to avoid a possible conflict; although the Court did not indicate which measures these would be, a national authority may suspend proceedings in order to avoid a conflict. It could, of course, also ask for an opinion from the Commission or, in the case of a court, refer a 234 question to the Court of Justice. While leaving national authorities and courts the freedom as to which measures to choose to avoid conflicts with Community law, the Court has clearly formulated an obligation to avoid conflicts with Community law. The Walt Wilhelm case concerned two parallel proceedings prohibiting the same cartel and imposingfines,one before the Commission and one before the Bundeskartellamt. There were thus two proceedings going in the same direction

2-Coherence

415

but with the risk of imposing double fines. On that issue, the Court held that parallel infringement proceedings are permissible in relation to the same subject matter but that the last deciding authority has an obligation to take any earlier decision into account in determining the level of its fines (principle of equity). Although this case involved two parallel prohibition proceedings, the Court also stated that Article 81 (3) permitted the Commission to exercise certain positive action in order to promote the harmonious development of economic activities in accordance with Article 2 of the EC Treaty throughout the Community. Since this particular statement of the Court related to Commission's action through constitutive exemption decisions, some commentators have doubts as to whether the Walt Wilhelm rule of primacy would still hold in a legal exception system where Article 81 (3) would be directly applicable and only have a declaratory nature. In my opinion there is no doubt that in a legal exception system the Walt Wilhelm rule of primacy will continue to apply to Article 81 (3). In Walt Wilhelm, the Court formulated a general principle of primacy that was based on the need to avoid that Community law produces different effects in different Member States depending on national measures which would jeopardise the proper functioning of the common market. The Court has applied its reasoning not only to Commission decisions with a constitutive nature but also to Treaty provisions producing a direct effect (see in particular paragraph 6 of the Walt Wilhelm decision). Its statement on "private action" under Article 81 (3) was an obiter dictum that was certainly not meant to exclude Treaty provisions with direct effect from the principle of primacy. The fact that Article 81 (3) would be given a direct effect in the future puts it at the same level as other Treaty provisions having a direct effect, in particular Articles 81(1) and 82. The mention of 'positive action' referred to the present Regulation 17/62 where only a constitutive Commission decision under article 81 (3) can render Article 81 (1) inapplicable. In the new system, if the Commission has adopted a declaratory decision under Article 81 (1) or 81 (3), companies will be able to directly invoke such decision against contrary national decisions. In the absence of such a decision, they can directly invoke the application of Article 81 (1) and 81 (3) before national courts and national competition authorities in order to set aside applications of national competition laws which would conflict with the direct effect of EC law. The burden of proof will be on the party invoking the primacy of EC law over national law. This is not different from a company bearing the burden of proof under Article 81 (3) or a complainant bearing the burden of proof under Article 81 (1) in a Commission proceeding. In order to remove any doubt about the primacy effect under the new system in which Article 81 (3) becomes directly applicable, it could be envisaged to clarify the primacy effect in the new Council Regulation. It could also be made clear that the burden of proof in relation to the fulfilment of the conditions of

416

Working Paper VIII

Article 81 (3) remains with the company invoking the benefit of that provision before a national court or national competition authority. How far should the rule of primacy be applied in the new system? The issue of primacy was debated very intensively before the adoption of the Walt Wilhelm decision (debate on Zweischrankentheorie v Einschrankentheorie). In the Community of the 1960s, that debate was influenced by the fact that Community law was still new and the common market was in its early stages. In the Community of 2000, the rule of primacy of Community law seems a more natural thing for a fully integrated market where distortions of competition are less tolerable and where national authorities and courts will be empowered to fully apply the EC competition rules. In the internal market of today, coherent application of national and Community competition law requires not only that directly conflicting decisions are avoided (the companies subject to national and Community decisions should not be confronted with conflicting rights and obligations). The Community of today requires also the avoidance of diverging decisions, which would create distortions of competition if applied to companies in the same market. Take a Community-wide market where the same agreement would be treated more severely in one Member State than in the rest of the Community. If the agreement could be executed in one part of the Community and not in another, this would create a distortion of competition, even if the addressee of these diverging decisions could technically execute both orders. If, in my example, the agreement were in addition indivisible (such as a specialisation agreement), the result would be that the most severe national competition law would de facto take priority over all other national laws and over Community law. This outcome would be paradoxical and would create distortions of competition on the Community market. Such situations are hardly compatible with a well-functioning common market. Where a case appreciably affects trade between Member States, two or more national competition laws frequently apply in addition to EC competition law. In that situation, if one Member State can prohibit an agreement that is not prohibited by EC law, it can make the execution of that agreement impossible in other Member States; or, if not impossible, it can make the execution of the agreement less efficient and prevent companies from fully exploiting the advantages resulting from the internal market. In an economically integrated market, companies need a level playing field not only in relation to what is prohibited but also in relation to what is permitted. That is the reason why the EC Treaty has foreseen a common competition policy. It is not a defence for the stricter Member State to say that its law will bring more competition in its own territory. What might be good for one Member State is not necessarily good for the harmonious development of economic activities inside the Community. If the market is wider than the territory of one Member State, companies should be able to operate as they might in an inter-

2-Coherence

417

nal market, i.e. either with one set of rules which would be EC law or with national rules which are fully convergent with EC law. There is therefore a case for at least a strong rule of primacy that does not only eliminate direct contradictions between national and Community law measures (such as national prohibitions and Community exemptions under Article 81 (3)), but that also excludes unequal treatment and distortions of competition in the internal market. The maximalist approach advocated by some commentators would go as far as excluding the application of national law from all cases affecting trade between Member States. Such an approach would avoid conflicts between EC law and national law, but the condition of effect on trade is probably too vague to serve as a basis for a system of exclusive jurisdictions. It is, however, a criterion that could work for the application of a strengthened rule of primacy. In my view, national competition laws should only continue to apply to cases affecting trade between Member States, if, and to the extent, that their applications do not interfere with the uniform application of EC law and does not distort competition on the common market. The rule of primacy should promote the widest possible convergence between EC law and national laws. This means that, for all cases affecting trade between Member States, the rule of primacy should at least: • prevent national law from authorising what is prohibited by EC law; • prevent national law from prohibiting or treating more severely what is exempted under Article 81 (3); and • prevent national law from prohibiting or treating more severely what is exempted by Block Exemption Regulations. It should further be debated whether or not national law can prohibit an agreement affecting interstate trade that does not violate Article 81 (1), particularly if the geographic market is wider than the territory of one Member State. This debate raises difficult new issues and should not be avoided, even if there will be a lot of resistance. The following questions can be asked: a) Why should one Member State be allowed to prohibit an agreement that might produce positive effects in other Member States if the agreement affects trade between Member States, in particular if it relates to a geographic market that is wider than one Member State? and b) Why can an agreement be treated more severely under national law if it does not restrict competition under Article 81 (1), compared to the same agreement that restricts competition but fulfils the conditions for an exemption under Article 81 (3)? Take a selective distribution system of a purely qualitative nature applied in several Member States which has a negative impact on prices but it does intensify competition between brands through better service. Compare this to another quantitative selective distribution system that violates Article 81 (1) but fulfils the exemption conditions of Article 81 (3). Under the Walt Wilhelm

418

Working Paper VIII

rule of primacy that applies only if Article 81 (1) is violated, the first system could be prohibited by national law; the second one could not. That is a rather paradoxical outcome. Companies would have an incentive to conclude restrictive agreements falling under Article 81 (1) but that are exempted, and would have an incentive to plead the application of Article 81 (1) in courts and before competition authorities in order to get the benefit of the rule of primacy that protects them against stricter national law. It will be objected that a strong rule of primacy including non-infringement of Article 81 (1) would eliminate national competition laws. In reality, this effect might not be so dramatic. Indeed, it is likely that the majority of cases presently examined under national competition laws are cases where the market is not wider than the territory of one Member State. In most other cases, notification to the Commission blocks national action. In addition, most Member States already have provisions similar to Articles 81 and 82, which facilitate a convergent application of national law and EC law. Certain Member States have an obligation—in applying their national versions of Articles 81 and 82—to take account of the case law and decision practice of the European Court of Justice/CFI and the Commission (see for instance Section 60 of the UK Competition Bill). Such rules amount to a strong rule of primacy but are incorporated in national legislation. Last, Member States will get full competence to apply EC competition law and are thus no longer obliged to systematically apply national competition law to cases affecting markets going beyond their Member State. The extension of the primacy rule to non-infringements of Article 81 (1) in cases affecting a geographic market that is wider than the territory of a Member State would make a major contribution to a coherent application of competition law in the common market: • It would help to eliminate conflicts between national and Community decisions in individual cases; • It would lead to a common de minimis rule for cross-border cases; • It would increase the value of Community Block Exemption Regulations that could not be set aside on the ground that Article 81 (1) is not violated even though the case affects cross-border trade; • Where a national competition authority brings a case affecting cross-border trade under national competition law, with the consequence that the Commission cannot withdraw the case under the rule of Article 9 (3) of Regulation 17/62, the national authority would be forced to take account of EC competition law when applying national competition law. An imperfect alternative to the extended rule of primacy would be to apply the Article 9 (3) mechanism to national competition law cases that affect a geographic market that is wider than the territory of a Member State. All these issues of primacy could, of course, be avoided to the extent that national competition authorities apply EC law directly, possibly in conjunction with national law, to cases affecting trade between Member States.

2 - Coherence

419

VIII. Conflicts in the same legal order: Delimitis1 and the obligation of co-operation on national courts Given the fundamental importance of ensuring a coherent application of the Community's competition rules, the Court of Justice has developed rules applicable to national proceedings not only to avoid conflicts between national and Community competition laws (which was the issue in Walt Wilhelm), but also to avoid conflicts between national and Community decisions both applying EC competition law. The latter relationship is the subject of the 1991 Delimitis judgment. In the same legal order, conflicts between decisions can arise either because several bodies share the competence to apply the same law (parallel competencies) or because one body has exclusive competence over part of the applicable law. Under the existing Regulation 17, both types of conflict can arise. Under the new system, conflicts can arise from the parallel competence for the application of Articles 81 and 82. In the Delimitis case, the Court of Justice created an obligation of cooperation for national courts, which basically aimed at the same result as the rule of primacy set out in Walt Wilhelm: • The Court starts by underlining that the Commission is responsible for the implementation and orientation of Community competition policy.8 • The Court then recalls that while the Commission holds exclusive competence for the application of Article 81 (3), it shares competence with particular national courts for the application of Articles 81 (1) and 82. National courts have an obligation to apply Articles 81 (1) and 82, which are directly applicable provisions. • The Court examines the consequences of this sharing of competencies, given the risk that national courts take decisions which might conflict with existing Commission decisions or with decisions envisaged by the Commission when applying Articles 81 (1), 82 and 81 (3). Such conflicting decisions would be contrary to the general principle of legal certainty. The Court states that conflicting decisions must be avoided. • The Court accordingly devised the following rules for national courts to reconcile their obligation to avoid conflicting decisions with their obligation to decide on cases submitted to them: —if it is manifest that Article 81(1) does not apply and that there is therefore no risk that the Commission will adopt a different decision, the national judge can decide;

7 8

Case C-234/89 Delimitis v Henninger Brciu [1991] ECR 1-935. See also Case C-l 19/97 P UFEX& Others v Commission [1999] ECR 1-1341, par.

420

Working Paper VIII

—if it is manifest that Article 81(1) applies and that the agreement concerned can under no circumstances benefit from an exemption under Article 81 (3), the national judge can decide; —if the national court finds that an agreement could possibly benefit from Article 81 (3) or if there is a risk of diverging decisions under Article 81 (1) or 82, the national judge may either suspend its proceedings pending the outcome before the Commission, or it may adopt interim measures. • In situations of doubt as to the application of Articles 81 (1), 82 or 81 (3), the Court retains the option that a national court might ask for an opinion from the Commission, in which case the Commission is obliged to reply (obligation of co-operation based on Article 10 of the Treaty). The national court may also ask a preliminary question to the Court of Justice under Article 234. Finally, in Delimitis the Court concluded that a national court could only pronounce the nullity of an agreement in application of Article 81 (2) if it was certain that the contract could not benefit from Article 81 (3). We conclude from this judgment that national courts have an obligation not to adopt decisions applying Community competition law that would be contrary to an existing or contemplated Commission decision. The Court leaves national courts free as to the measures to be taken to achieve that result. The fact that the Court does not impose on national courts an obligation to suspend proceedings pending the outcome of a contemplated Commission decision does not change the obligation imposed on national courts to avoid conflicts with adopted or contemplated Commission decisions. Also, the fact that the Delimitis case related to an 81 (3) issue, for which the Commission currently has exclusive competence, was not material in the reasoning of the Court of Justice. Delimitis and Walt Wilhelm follow the same logic, based on the need to have a coherent legal system. The rule of primacy applies to the relationship between two different legal orders: national law may not contradict or eliminate the 'effet utile' of Community law. As regards the same legal order, the need for uniform application and legal certainty requires national courts not to contradict or eliminate, directly or indirectly, the effectiveness of a Community act. In both situations, the national court has to take account of an envisaged or already adopted Commission decision as long as that decision has not been successfully challenged before the Court of Justice. This is logical because the Commission has the special responsibility of implementing and orientating Community competition policy. Even if national courts act as Community courts when applying Community law, they do not have the role of the Commission. Their decisions remain national in scope, deciding a case with inter panes effect; the Commission's decisions—whether negative or positive— have a Community-wide effect resulting from Article 249 (previously Article 189) of the Treaty. It has been argued that a national court can not be bound by a Commission decision because the Commission is only an administrative body (principle of separation of powers). The principle of separation of powers that applies

2 - Coherence

421

within the same legal order cannot be applied, as such, to the relationship between the Community legal order and the national legal orders. If this principle were to be applied to that relationship, even the rule of primacy which is based on the application of Community law would be placed in jeopardy because a national judge could disregard envisaged or adopted Commission decisions that were 'only' administrative decisions. Prohibition decisions of the Commission could also be ignored by national courts. This would deny the very special role of the Commission and detract from the fundamental requirement of uniform application of Community law and of legal certainty. It would also violate Article 10 of the Treaty that contains an obligation for Member States, including national courts, to facilitate the achievement of the Community's tasks and to abstain from any measure that could jeopardise the attainment of the objectives of the Treaty. Article 10 covers obligations arising out of the Treaty or resulting from action taken by the institutions of the Community. The principle of separation of powers was developed by Montesquieu (L'Esprit des Lois) in order to ensure an equilibrium of powers, particularly between the executive and the judicial. Executive power has to be controlled by the judicial power and the judiciary may not have the power to execute its own decisions. However, assuming that the Commission is an ordinary body of executive power, then it is controlled by the European Court of Justice. Its decisions cannot oppress national courts because Member States can always test the legality of Commission decisions before the European Court of Justice (Article 230). Once Commission decisions have acquired definitive validity inside the Community legal order, they cannot and do not need to be further controlled by national judges who have no jurisdiction over the Commission, a Community institution.9 There are no legal grounds and no reasons for national courts to control the 'administrative' activity of the Commission. Equally, there is no valid legal or political reason why national courts should not submit to Community acts that have acquired validity inside the Community legal order. There is, however, a compelling reason why national judges should respect definitive Community acts: if each national court could go its own way, the effect of adopted Commission decisions would vary from Member State to Member State. This would jeopardise the whole Community legal order. That is also the reason for the specific obligation of co-operation enshrined in Article 10 of the Treaty. This Article is of fundamental importance for the functioning of the Community because it creates not only a faculty of co-operation, but also an obligation to co-operate to achieve the goals of the Treaty. However, the Commission has no jurisdiction over national courts. Not even the Court of Justice has jurisdiction over national courts. The Community legal 9 This is without prejudice to the possibility that third parties can raise the question of validity of Community acts before national courts, which must then refer a preliminary question to the Court of Justice, see Case C-l88/92 TWD [1994] ECR 1-833.

422

Working Paper VIII

order is characterised by a system of co-operation where, for instance, a national court can have an obligation to refer a question on the interpretation of Community law to the Court of Justice, but the latter has no power to annul contravening national laws or to annul a national court decision. However, national courts are subject to the obligation to follow the interpretation given by the European Court of Justice in respect of Community law. The European Court of Justice has only the power to annul Community decisions. It follows that the competence to solve conflicts between national and Community judgments lies with national courts, but they are subject to certain fundamental rules of law and to certain obligations of co-operation with the Court of Justice and the Commission. National courts themselves decide about the application of the rule of primacy and the application of EC law, but they have an obligation to avoid conflicts with envisaged or definitive Commission decisions. They have three options: they can either follow the decision of the Commission, suspend proceedings, or refer an Article 234 question to the European Court of Justice. If the European Court of Justice does not annul the Commission's decision, then the decision must be respected by national courts. This obligation holds, of course, only rebus sic stantibus. In the Community legal order where the Court of Justice has no power to annul national court decisions and where national courts cannot put into question the validity of Community acts, there must be a rule of law that creates a bridge between these two legal orders: it is the rule of co-operation and the obligation not to adopt decisions that would conflict with Community acts that form this link. These are binding legal rules for national courts. Non-respect of these rules could give rise to an Article 226 action initiated by the Commission. This is not excessive interference in the independence of national judiciaries. The Delimitis obligation not to contradict Commission decisions should a fortiori apply to national competition authorities. This follows from Article 10 of the Treaty.

IX. Effects of Commission decisions In the new system, all findings of an infringement or non-infringement will have a declaratory effect, that is, they will state the law applicable to a given set of facts {rebus sic stantibus). Material changes to the facts can lead to different findings. This is already the legal situation as regards Articles 81 (1) and 82. A restriction of competition or an abuse exists only as long as the condition of appreciability or dominance is given as a matter of fact. If these conditions are fulfilled, the Commission, a national competition authority or a national court can find an infringement and draw the legal consequences of the infringement. Commission decision under Article 81 (1) contains two parts: first, the finding of the infringement; second, an order of termination. If the factual situation materially changes, the first part of the decision (which is the indis-

2 - Coherence

423

pensable basis for the second part) could disappear and thus also lead to the end of the order of termination. The company can start implementing the same agreement or adopt the same practice if it is no longer infringing Article 81(1) or no longer committing an abuse. The same holds for a finding of noninfringement under the new system. Such a finding is only valid as long as the facts have not materially changed as to the non-application of Article 81 (1) or as to the application of Article 81 (3). It is important to distinguish this declaratory effect of Commission decisions in the new system from the constitutive effect attached to Article 81 (3) exemption decisions under the existing Regulation 17/62. Under the present regime of Article 81 (3), the Commission has the power to create rights in the form of an exemption decision having effects for the future irrespective of any change in facts (see Article 8 of Regulation 17/62). This constitutive effect goes with an exclusive power of exemption granted to the Commission under Article 9 (1) of the present Regulation. Such decisions with a constitutive effect are of course incompatible with a legal exception system where all decisions, whether negative or positive, have a declaratory nature. The orders attached to declaratory decisions hold only as long as the finding of infringement or the finding of noninfringement holds. What then will be the exact effects of future declaratory decisions of the Commission? The effects will depend on the nature of the decision. A Commission decision has the effects provided for in Article 249 EC Treaty: 'a decision shall be binding in its entirety upon those to whom it is addressed'. Parties with a right to challenge a Commission decision under Article 230 (a right that must be exercised within a two month period) cannot subsequently indirectly challenge the validity of the decision before a national court. Third parties who are not entitled to appeal against a decision under Article 230 can put into question the validity of the Commission decision before a national judge. In such a case, the national court must refer a preliminary question to the Court of Justice under Article 234 (see Fotofrost10). If the material facts are the same, a national court can not depart from a prohibition decision—in spite of its declaratory nature—since such departure would directly or indirectly put into question the validity of a Community act over which the national court has no competence. For instance, if a national court were to order the execution of an agreement that the Commission had prohibited, there would be two conflicting orders. A legal system could not survive without a rule to solve that conflict. This is the Delimitis rule that obliges national courts not to adopt contradicting decisions, and the Fotofrost rule that prevents national courts from putting the validity of Community acts into question. These two rules are complementary. A national court can only decide differently if the facts have changed materially or if it has referred an Article 234 question and the Court of Justice has declared the Commission decision invalid. 10

Case 314/85 Fotofrost [1987] ECR 4199.

424

Working Paper VIII

As regards prohibition decisions, once the Commission has made a decision the national court must respect the finding of infringement and the order of termination. This is the only guarantee that the addressee of the Commission decision will execute the Commission decision and that a third party can oppose the execution of the illegal agreement and claim damages. The Commission has no power to execute its own decisions and it cannot draw the civil law consequences resulting from an infringement. It fully relies on national courts for the enforcement of its decisions. The only exception to this rule is the execution of Commission decisions imposing a pecuniary obligation (see Article 256). Non-infringement decisions, as foreseen in paragraphs 88 and 89 of the Commission's White Paper, will also have a declaratory effect. If these decisions are not mere decisions of non-intervention (such as the present Article 2 negative clearance decisions) but are decisions finding with effect rebus sic stantibus that either Article 81 (1) is not violated or that 81 (3) applies, they will have the same effects as prohibition decisions. That is, a national court cannot depart from a non-infringement decision in spite of its declaratory nature, since such departure would directly or indirectly put into question the validity of a Community decision over which the national court has no competence. It can only decide differently if the facts have materially changed or if it has referred an Article 234 question to the Court of Justice. This is the only guarantee that the addressee of the non-infringement decision has the right to execute its agreement and that the contracting party has (or third parties have) an obligation to respect the execution of the agreement ('opposabilite'). In the new system there would only be non-infringement decisions in exceptional circumstances. There would be no possibility for companies to request a non-infringement decision upon notification with ex ante control. Neither the Commission nor national competition authorities will have the power to adopt non-infringement decisions for the purposes of certifying to companies the compatibility of their agreements with Articles 81 or 82. Non-infringement decisions should only be taken by the Commission for reasons of public Community interest. These are: • to provide the market with guidance regarding the Commission's approach to new competition issues, or • to exercise the Commission's own power of decision-making in the enforcement of the Treaty competition provisions with a view to influencing the orientation of Community competition law in a system of parallel competencies. This power of decision-making by the Commission should not be actionable under Article 232 of the Treaty. In the new system, national courts have full power to apply Articles 81 and 82 and can draw all civil law consequences deriving thereof. National courts are no longer hindered in their powers of decisionmaking by the exemption monopoly of the Commission. The effects of Commission decisions can thus be summarised as follows:

425

2 - Coherence

Commission decision

Effect on addressee

Effect on 3rd parries

Effect on national courts

Prohibition decision

Obligation to terminate infringement (rebus sic stantibus)

Right to oppose execution of agreement

Walt Wilhelm; Delimitis; Fotofrost (subject to res judicata)

Decision accepting commitments

Obligation to respect commitments {rebus sic stantibus)

Right to request compliance with commitments

Walt Wilhelm; Delimitis; Fotofrost (subject to res judicata)

Non-infringement decision

Right to execute agreement (rebus sic stantibus)

Obligation to respect execution of agreement

Walt Wilhelm; Delimitis; Fotofrost (subject to res judicata)

National competition authorities would only adopt 'positive' decisions in the context of rejecting complaints or terminating ex officio infringement proceedings. These decisions should only bind the acting national competition authority. They would not have any binding effect on the Commission or other national competition authorities. Finally, the Commission envisages that it will issue a limited number of nonbinding opinions in respect of issues that are not sufficiently covered by existing horizontal measures or individual decisions. Although not binding on national courts or competition authorities, they are likely to assist companies in situations of real doubt about the interpretation of the law. They will also make some contribution to the coherent application of EC competition law because such opinions would be motivated and published.

X. Harmonisation of national procedures and sanctions At the national level, Community law will be applied with national procedures that vary from country to country. The rules on evidence, rights of defence, decision-making, appeals, time limits, sanctions and remedies are not harmonised. Therefore, a number of commentators have expressed the fear that it will be impossible to apply EC competition rules in a uniform manner throughout the Community. They also point towards the risk of forum shopping that could be provoked by the existing differences in procedures and sanctions. The ideal situation would indeed be the application of the same rules of substance and procedures by all decision-makers. However, Europe cannot be

426

Working Paper VIII

created in one day. We do not consider that a harmonisation of national procedures is indispensable for the decentralisation of Article 81 (3) to courts and national competition authorities. They have already applied Articles 81 (1) and 82 with national procedures and sanctions for over 40 years, and the same is true for the application of other Community provisions of the Treaty or for secondary legislation. Yet it must be recognised that national courts, in particular, have made a major contribution to the efficient application of EC law throughout the Community without putting at risk the functioning of the internal market. In the field of procedures it is important to strike the right balance between the principle of subsidiarity on the one hand and the principle of efficiency and coherence on the other hand. In the Protocol on the application of the principles of subsidiarity and proportionality annexed to the Treaty of Amsterdam it is stated that '[t]he Community shall legislate only to the extent necessary . . . While respecting Community law, care should be taken to respect well established national arrangements and the organisation and working of Member States' legal systems.' Under Community law in general, procedural law falls within the sphere of competence of the Member States. However, the Court of Justice has held that this competence is subject to certain limits imposed by Community law. According to the principle of effectiveness, national procedural rules must not render virtually impossible or excessively difficult the exercise of the rights conferred by Community law. This implies that Member States might have an obligation to change their rules on procedures and sanctions in order to ensure an 'effet utile' to Community law as a result of Article 10 of the Treaty (which creates an obligation of co-operation to facilitate the achievement of the Community's tasks). The Court of Justice has further established the principle of equivalence, according to which national rules on procedures and sanctions concerning the enforcement of Community law must not be less favourable than those governing similar domestic law, in this case national competition law. Finally, the Court of Justice has developed the principle of equity, obliging courts and authorities to take account of sanctions imposed by other authorities/courts in relation to the same infringement (see Walt Wilhelm). Thus, excessive punishments must be avoided. In spite of these principles applied by the Court of Justice, there are and will remain differences between the Member States—sometimes substantial ones— which probably means that the harmonisation of certain procedural rules is desirable in the medium to long term. As long as there are differences in procedures and sanctions, there will be a certain risk of forum shopping. However, this is a risk that is not specific to competition law, it applies to all Community law that is implemented by national authorities and courts. Also, forum shopping is limited by the Convention of Brussels as regards courts, and by the elimination of constitutive exemption decisions and rules about the allocation of cases inside the network as regards national competition authorities. Therefore, there is no urgency to harmonise national rules on procedures and sanctions for the implementation of the reform proposed by the Commission.

2 - Coherence

427

However, absent harmonisation of national rules on procedures and sanctions it might be necessary to put certain limits on the use of information exchanged between legal systems that do not contain the same protections of rights. Such limits would not force Member States to change or harmonise their national rules but would restrict, for instance, the use of evidence received from another national authority. Over time, such limits will put pressure on national systems and possibly lead to a gradual convergence in the field of procedural rules and rules on sanctions, including damages. Application of the same substantive law and benchmarking will bring national legal systems closer to each other. At a later stage, however, it might be desirable to adopt a Regulation harmonising certain aspects of procedures and sanctions in connection with the application of Community competition rules.

XI. Concluding remarks The decentralised application of Community competition rules and the abandonment of the prior authorisation system will not put at risk the coherent application of Community law and will not lead to a re-nationalisation of competition law, provided that: a. The principle of primacy of Community law over national law prevents the application of national law from undermining the uniform application of EC law. b. The principle of co-operation with the European Court of Justice (Article 234) and with the Commission (Article 10) is implemented efficiently under the control of the European Court of Justice. c. Co-operation with national competition authorities is intensified on the basis of a relationship of trust within an efficient network that should develop a Community competition culture. d. The Commission continues to play a central role in the formulation of competition policy via horizontal instruments and individual decisions. e. Community competition law will be applied more and more to cases affecting trade between Member States.

IX James F. Rill* Recent Trends and Developments in the Role of Federal and State Authorities in US Antitrust Enforcement: Is Cooperation Working?

I. Introduction The United States has a long-standing history of apportioning governing authority between the national government and the states. This sharing of power also extends to the enforcement of antitrust laws—where state authorities have jurisdiction over certain matters together with the federal antitrust agencies. As the European Commission contemplates relinquishing certain competition policy oversight functions to national competition authorities, it may be instructive to review and evaluate the United States' system of dual enforcement authority between the federal antitrust agencies and the individual state authorities. Read in conjunction with a paper I presented before this conference several years ago on federalism in U.S. antitrust enforcement,1 this paper provides an update on recent activities within the United States pertaining to federal-state cooperation or conflict and offers additional insights as to the effectiveness of this system of dual enforcement. As was discussed in the 1996 paper, while federal authorities have maintained primary responsibility for the enforcement of antitrust laws, certain occurrences during the early and mid-1980s prompted the state authorities to increasingly turn their attention to antitrust matters and to enforcement of both federal and state antitrust laws. While some of the initial reasons for enhanced state activity in the antitrust arena may have dissipated or been resolved, the states' role in antitrust enforcement has remained—and possibly increased—during recent years. As this paper will discuss, however, although state involvement in antitrust matters continues, instances of conflict generated by inconsistent federal and state decisions appear to have decreased. We will examine some of the causes for this reduced conflict and will analyse those instances where conflict has not been eliminated from the dual enforcement system. As the European Commission considers launching a new procedural regime for certain competition-related * I acknowledge with gratitude the indispensable assistance of my colleague Sarah E. Bauers in the preparation of this paper. 1 For a comprehensive background on the genesis of U.S. federalism in the antitrust context and an overview of significant developments in this arena, see Rill, J.F. and Chambers, C. S. (1997): 'Federalism in Antitrust Enforcement: The United States' Experience with a Dual Enforcement Regime', in Ehlermann, C.-D. and Laudati, L. eds., Robert Schuman Centre Annual on Competition Law 1996, pp. 205-225.

430

Working Paper IX

matters, the EC and European businesses and their advisors may find useful an evaluation of the advantages and disadvantages of the federal-state enforcement regime within the United States.

II. Recent trends toward cooperation The role of both federal and state regulators in the enforcement of antitrust laws has become headline news in the United States through several recent high-profile cases involving antitrust issues.2 These well-publicised cases have served to further underscore the extent to which state attorneys general have maintained a presence in the antitrust enforcement arena. Unlike many actions of the past, many cases today pursued by state attorneys general are now brought in conjunction or cooperation with similar federal action. This trend has produced somewhat greater harmonisation in the outcomes of both federal and state actions and has eliminated some of the friction historically associated with dual enforcement. Nevertheless, despite the increased level of cooperation, it is evident that the unavoidable tensions inherent in such an enforcement regime still appear, albeit in a more limited number of instances.

1. Mechanisms for promoting greater cooperation As discussed more extensively in Rill and Chambers (1997), a number of mechanisms have been created to facilitate cooperation between the federal and state antitrust officials. These mechanisms include the Voluntary Pre-Merger Disclosure Compact,3 the Executive Working Group for Antitrust,4 and the 2 The state attorneys general also were involved in a very high-profile settlement with four major tobacco companies in the U.S. While the settlement did raise some discreet competition-related issues, this paper will not attempt to address the matter since a review of the tobacco settlement would be well beyond the scope and purpose of this discussion. 3 4 Trade Reg. Rep. (CCH) % 13,410 (Apr. 5, 1994). See Rill and Chambers (1997) for a complete analysis of these initiatives for enhanced cooperation. 4 See Assistant Attorney James F. Rill, Antitrust Division, U.S. Department of Justice, Antitrust Enforcement: An Agenda for the 1990s, Address Before the TwentyThird Annual New England Antitrust Conference (Nov. 3, 1989), at pp. 10-11. Other building blocks developed during the Administration of President Bush include indepth consultation between the Department of Justice and the National Association of Attorneys General in connection with the development of the DOJ-FTC Horizontal Merger Guidelines (1992) and the NAAG Merger Guidelines (1993) and the joint investigation and consultation on remedies in the first Primestar matter. See United States v Primestar Partners, L.P., 1994-1 Trade Cas. (CCH) 170,562 (S.D.N.Y. 1994); New York v Primestar Partners, L.P., 1993-2 Trade Cas. (CCH)H 70,404 (S.D.N.Y. 1993).

2 - Coherence

431

Protocol For Increased State Prosecution of Criminal Antitrust Offenses.5 Since 1996, additional formal efforts have been undertaken to further enhance practical levels of cooperation between federal and state antitrust authorities. For example, the state attorneys general, the Federal Trade Commission, and the Department of Justice jointly released the Protocol for Coordination in Merger Investigations {^Merger ProtocoF) in 1998.6 The Merger Protocol provides a framework for cooperation between federal and state authorities during merger investigations, with the aim of minimising friction between the agencies and reducing the transaction costs for the merging parties. In order to achieve these results, the Merger Protocol addresses procedural issues that might arise during any coordination of efforts, such as the need to ensure the confidentiality of relevant information7 and specific guidelines for strategic planning and evidence gathering during merger investigations. Officials noted that the Merger Protocol merely formalises procedures and guidelines that had previously been used by federal and state authorities. Federal-state cooperation within the United States also occurs through the submission of amicus curiae briefs. Through this mechanism, either the federal or state regulators are afforded the opportunity to present their views during the litigation process without generating the burden associated with initiating a separate proceeding. Likewise, the business community needs only to respond to one agency and the ensuing costs of a single litigation process. In several recent large cases, state attorneys general have supported actions undertaken by one of the federal antitrust agencies through the submission of amicus curiae briefs. In most of these cases, a contingent of state attorneys general voiced their support for the federal agency's position.8 The amicus initiative finds its recent and most significant roots in the Department of Justice's submission of a brief on behalf of the states in the 5 See Department of Justice Press Release, Justice Department Issues New Protocol for Increased State Prosecution of Criminal Antitrust Offenses (Mar. 26, 1996). 6 Attorneys General and Federal Antitrust Agencies Announce Protocol for Joint State/Federal Merger Investigations, 25 NAAG Antitrust Rep. 1 (Mar./Apr. 1998). 7 The Merger Protocol suggests that either agency that receives confidential information from the other agency should take all steps necessary to maintain confidentiality, including: (1) provide timely notification to the originating agency of any requests for information; (2) assert any and all privileges or exemptions from disclosure claimed by the originating agency; (3) intervene or provide assistance to the originating agency in the intervention in legal proceedings to assert relevant privileges or exemptions; and (4) comply with any conditions imposed by an agency that shares information it deems to be confidential. See Merger Protocol, § 1. 8 See Amicus curiae Brief filed in FTC v Staples, Inc., 970 F. Supp. 1066 (D.D.C. 1997); Amicus curiae Brief filed in FTC v Butterworth Health Corp., 1997-2 Trade Cas. (CCH) K 71,863 (6th Cir. 1997); Amicus curiae Brief filed in FTC v Cardinal Health, Inc., 12 F. Supp. 2d 34 (D.D.C. 1998); Amicus curiae Brief filed in Toys R' Us v FTC, No. 984107 (7th Cir. 1999).

432

Working Paper IX

Insurance Antitrust Litigation.9 The federal government supported the states' position that the insurance-company defendants' alleged conduct was not exempted by the McCarran-Ferguson Act10 nor, in the case of Lloyds, protected by international comity principles.'' The Department's assertion of this position was the subject of both vigorous lobbying by the defendants and equally active support from the states. As a result of this filing, moreover, a spirit of mutual confidence and cooperation developed in the federal-state relationship.

2. Cooperation in practice The dual jurisdiction over antitrust enforcement by both federal antitrust agencies and state attorneys general has been illustrated by one of the most significant antitrust cases on record. In the anti-competitive business practices case against Microsoft Corp.,12 19 state attorneys general joined sides with the U.S. Department of Justice in the prosecution and settlement aspects of the investigation. The joint action against Microsoft, which is proceeding in federal court, has garnered overwhelming attention from the news media, and has propelled the role of the state attorneys general in the antitrust field into the headlines as well. While the coordination between the Antitrust Division and the states has been notable, there are indications that some frictions may have emerged during the process. The joint case against Microsoft originated following parallel, but separate investigations undertaken by the Antitrust Division and several state attorneys general. Reports speculate that the state attorneys general initiated their own separate investigation because of their uncertainty as to whether the Department of Justice would bring a case against Microsoft.13 While federal and state authorities filed separate complaints against Microsoft in federal court, the court consolidated the two cases shortly after the process began in May 1998 and the process of cooperation was underway. Once the cases were consolidated, cooperation between the plaintiffs was imperative in order to bring a cohesive and uniform position to the litigation process. During the trial, both federal and state officials were permitted to appear before the court. The judge, however, imposed strict trial rules and man9

In re Insurance Antitrust Litig., 723 F. Supp. 464 (N.D. Cal. 1989), rev'd and remanded, 938 F.2d 919 (9th Cir. 1991), affdinpart, rev'd in part, sub relnom., Hartford Fire Ins. Co. v California, 509 U.S. 764 (1993). 10 15U.S.C. %\Q\\etseq. 11 In re Insurance Antitrust Litig., 723 F. Supp. at 795-800. 12 United States v Microsoft Corp., No. 98-1232 (D.D.C. filed May 18, 1998). 13 See James V Grimaldi, Microsoft Trial Puts States in Backseat; Out for Blood, They Must Follow the Feds' Lead, Seattle Times, at Al (Dec. 26,1999). Some believe that the states' decision to proceed with a case against Microsoft helped to convince the Justice Department to officially charge Microsoft with anti-competitive conduct. See id.

2 - Coherence

433

dated that each side allocate only one attorney to each witness. This order, in effect, precluded the states from examining the vast majority of witnesses and, according to some press reports, generated tensions between the federal and state authorities.14 Reports of friction increased following the court's initial ruling. Both sides were required to submit recommendations for possible remedies to be issued by the court. Presiding Judge Thomas Penfield Jackson ordered that only one proposal be tendered by the plaintiffs, forcing both the federal and state authorities to reach a consensus on the appropriate relief. Throughout the course of the litigation, news reports indicated that the federal and state officials held differing opinions as to the appropriate remedy, with supposition that the Department of Justice favoured conduct-based relief and the state attorneys general favored structural relief.15 In the end, both the Department of Justice and the 19 state attorneys general submitted a joint recommendation for relief requesting a structural division of Microsoft.16 While most public statements by the relevant parties indicated a cohesive process of negotiation, private comments hinted at widespread dissension among the Antitrust Division and state attorneys general. As one senior aide to a state attorney general was quoted as saying: 'There's more tension between the states and Justice on remedies than there was during the trial. During the trial, we were mostly just passengers, but now there are two different trains running on parallel tracks.'17 In drawing conclusions, one must be careful not to rely faithfully on press reports on how well cooperation has worked in this case. In addition, given their differing orientations, it is to be expected that the federal antitrust authorities and the state attorneys general may have different perceptions regarding appropriate relief, even when they agree on the merits of the case against Microsoft. An overall theme of coordination between federal and state authorities also pervaded the high-profile investigation and settlement of the world's largest merger involving the oil giants Exxon and Mobil. The Federal Trade Commission, 13 states, and the District of Columbia entered into settlement agreements with the merging companies following lengthy negotiations that required the divestiture of over 2,400 gas stations nationwide.18 In addition to the consent agreement accepted by the FTC, the merging parties entered into separate settlement agreements with four states that addressed the local impact 14

See id. Joel Brinkley, Antitrust Alliance Not Quite Solid, New York Times, at A18 (Apr. 3, 2000). 16 Reports indicate that the transformation of the Antitrust Division's stance on appropriate remedy was based primarily on the strongly-worded opinion against Microsoft by the court—not because of any influence directed by the states. See Jube Shiver, Jr., Some States Wavering on Microsoft Breakup, Los Angeles Times, at Cl (Jan. 26, 2000). 17 Joel Brinkley, U.S. and State Officials Weigh Microsoft Remedies, New York Times, at Cl (Nov. 19, 1999). 18 In the Matter of Exxon Corp. and Mobil Corp., Docket No. C-3907 (Nov. 30,1999) (consent order). 15

434

Working Paper IX

of the merger on consumers.19 The interested states and the FTC coordinated their investigations and remedy proposals throughout the entire process. Because of the direct relevance of the merger to local conditions in individual states, state attorneys general asserted substantial influence in the settlement process. Despite this involvement, however, one state refused to accept the proposed settlement agreements, citing a post-merger threat of increased prices and lower quality of service for consumers. Connecticut Attorney General Richard Blumenthal argued that by accepting the consent order, the state would relinquish its right to challenge anti-competitive conduct of the mergedfirmat a later date.20 Attorney General Blumenthal acknowledged that the refusal to accept the consent order was principally a statement against the merger, as one state 'standing alone would face formidable obstacles in [efforts to block] the merger.'21 In addition to the federal-state cooperative efforts that have reached the front pages of the newspapers, cooperation on a less publicised level continues to occur on a regular basis. Much of this cooperation centers on joint action in the merger area, particularly for proposed transactions that involve local markets.22 Many of these cases involve coordinated investigations and joint consent orders issued by both the federal agency and one or more of the state attorneys general. From the perspective of the agencies, such coordination can be more efficient as agencies may be able to pool resources and share the responsibilities for gathering necessary evidence. In addition, joint investigations and concurrent remedies can minimise the transaction costs for businesses, as well as the legal uncertainties generated by several investigations and the legal aspects of abiding by different, although not necessarily inconsistent, decrees.

3. Reasons for increased cooperation This increase in cooperation between both federal and state authorities may be attributable to a combination of factors. To start, one of the initial reasons for the rise of state involvement in antitrust matters is often linked to the less 19 States, FTC, Approve Exxon/Mobil Merger With Divestitures, 26 N A A G Antitrust Rep. 4 (Nov./Dec. 1999). 20 See S u s a n E . K i n s m a n , Big Oil Merger Making Waves; Blumenthal Balks at Agreement, Hartford C o u r a n t , at E l ( D e c . 1, 1999). 21 Rejection of Exxon-Mobil Merger Deal, Connecticut Attorney General's Office Press Release (Nov. 30, 1999). 22 See, e.g., United States v Allied Waste Indus., Inc., 1999-2 Trade Cas. ( C C H ) H 72,606 (D.D.C. 1999) (consent decree filed b y D O J , Illinois, a n d Missouri to remedy alleged anti-competitive effects o f a n acquisition in the waste management industry); FTC v Tenet Health Care Corp., 117 F. Supp. 2d 937 (E.D. M o . 1998), rev'd, 86 F.3d 1045 (8th Cir. 1999) (consolidated lawsuit b r o u g h t by the F T C a n d Missouri against a proposed hospital merger in which t h e 8th Circuit issued an order enjoining the transaction); United States v Cargill, Inc., 1997-2 T r a d e Cas. ( C C H ) \ 71,893 (W.D.N.Y. 1997) (consent order filed by D O J , N e w York, Pennsylvania, and Ohio requiring divestiture in

an acquisition in the commercial salt industry).

2 - Coherence

435

aggressive enforcement policies of the federal antitrust agencies during the early and mid-1980s. Today, state attorneys general appear to generally support the overall direction of federal enforcement policies for antitrust matters.23 While the states' involvement in antitrust enforcement may not be a trend that the states wish to reverse, cooperation with federal officials who appear generally to be on the same page in terms of enforcement direction seems to provide an acceptable balance for all parties. In addition, cooperative efforts help to relieve some of the pressure felt by both federal and state authorities regarding scarce prosecutorial resources. State antitrust agencies, in particular, are often thinly staffed because of budgetary constraints.24 Moreover, the increasingly complex nature of some of the high-tech industries currently being investigated for possible antitrust violations results in a combination of resources with the federal agencies.

III. Inherent inconsistencies resulting from dual enforcement Although significant benefits may be achieved by federal and state authorities when coordination is pursued in antitrust enforcement, there are inherent frictions associated with a dual enforcement regime. As demonstrated by the wellpublicised cooperation in the Microsoft case, tension and duplication of efforts can still arise in instances where cooperation is undertaken. While overlapping review of antitrust matters holds the potential to waste scarce prosecutorial resources, the greater threat to effective antitrust enforcement is the possibility that the federal and state authorities will reach inconsistent outcomes. Although the incidence of inconsistent outcomes has been limited over the past several years, the potential for such a result is still a serious concern. One area of particular concern has been the states' more aggressive stance in the remedy phase of merger investigations—a position that sometimes has conflicted with the federal antitrust agencies' determinations in merger cases.25

23

See Joel I. Klein Confirmed as Head of Antitrust Division, Attorneys General Support Nomination, 24 N A A G Antitrust Rep. 1, 2 - 3 (July/Aug. 1997) (statements by several state attorneys general voicing support for the direction of federal antitrust enforcement under the leadership of Assistant Attorney General Klein). 24 Approximately 130 deputy state attorneys general are assigned to h a n d l e antitrust issues. However, at least 21 states have either one or n o antitrust lawyers o n staff See Myron Levin, Teaming Up to Aid Consumers; With Activism and Sheer Numbers, State Attorneys General Emerge as Powers in Regulating Business, Los Angeles Times, at A l (July 6, 1999). 25 It should be noted, however, that sometimes the reverse can be true. F o r instance, a recent merger between two hospitals in N e w York drew a court challenge by t h e Department of Justice, while, following an investigation, the N e w York state attorney

general opted to neither pursue its own case nor join the Antitrust Division in its challenge. United States v Long Island Jewish Med. Ctr, 983 F. Supp. 121 (E.D.N.Y.

436

Working Paper IX

The potential for disparate outcomes is heightened by the fact that states sometimes apply their antitrust laws in order to accomplish objectives unrelated to competition principles.26 While the incidence of this type of application of antitrust laws appears to be declining, it still occurs. In a recent action in Alaska regarding the merger of two local supermarket chains for example, the state attorney general entered into a consent order that required several divestitures in order to remedy potential anti-competitive harm.27 In addition, however, the consent order contains several provisions aimed at insuring the supermarkets' continued involvement in the local Alaskan community. For instance, the consent order requires Safeway to continue to provide the opportunity for Alaskan companies 'to sell their products and services to Safeway, hire Alaskan employees, maintain in-state Alaskan management, and continue [the acquired chain's] support for local schools, charities and other community groups.'28 Occasionally, federal and state authorities can have diametrically opposed conclusions regarding the potential anti-competitive effects of a proposed transaction. Such was the case in the litigation surrounding the merger between BP Amoco and Atlantic Richfield Co. (ARCO). The FTC and three states filed suit against the merger, alleging that the transaction would lessen competition in Alaska North Slope crude exploration, production and sales to West Coast refineries.29 While the FTC and state attorneys general from California, Oregon, and Washington coordinated their efforts opposing the merger, Alaska's governor Tony Knowles directed State Attorney General Bruce Botelho to intervene in the proceeding and directly oppose the FTC's attempt to challenge the transaction. Calling the FTC's reasoning 'flawed,' Governor Knowles argued that the merger between BP and ARCO was imperative for the Alaskan economy, con1997). In fact, the Justice D e p a r t m e n t lost t h e m o t i o n for a permanent injunction. In the court's opinion allowing the merger t o proceed, the j u d g e noted the New York Attorney General's decision n o t to join the Antitrust Division ' i n the prosecution of this lawsuit.' See id. at 135. 26 See, e.g., Commonwealth of Pennsylvania v Russell Stover Candies, Inc., 1993-1 Trade Cas. (CCH) H 70,224 (E.D. Pa. 1993); see also Pennsylvania Settles Charges From Whitman!Russell Stover Merger, 20 NAAG Antitrust Rep. 8 (Sept./Oct. 1993). 27 In addition to the divestiture requirements, the consent order contains a unique clause that seeks to guarantee that prices will not increase during the period the divestitures are occurring. To this end, the consent order imposes a ceiling on the supermarket's gross profit margin in each relevant market. The formula mandates that the gross profit margin can be no higher than one percentage point above the combined gross profit margin for the two stores during the previous year. If the margin violates this formula, the store must return the amount of revenue realised through the violation to the State. See Alaska v Safeway, Inc. and Carr-Gottstein Foods Co., Case No. 3AN-99-4371 (Sup. Ct. Alaska 1999) (consent order). 28 Alaska v Safeway, Inc. and Carr-Gottstein Foods Co., C a s e N o . 3AN-99-4371 (Sup. Ct. Alaska 1999) (consent order). 29 Complaint, FTC v BP Amoco, pic, Case No. 000415SI (N.D. Cal. Feb. 4, 2000).

2 - Coherence

437

sumers, and workers. The FTC challenged Alaska's attempt to intervene in the preliminary injunction proceedings, claiming the only role the state should play in the litigation would be as amicus curiae.30 While the FTC and the three states opposed to the merger eventually resolved their concerns through a consent order with the merging parties,31 the litigation between federal regulators and Alaska over which authority has the ultimate right to challenge or approve a merger provides a glimpse into the potential problems associated with a dual enforcement regime system.32 The ramifications of potential frictions associated with concurrent federal and state merger authority become particularly troublesome as international cross-border transactions increase. While few, if any, instances exist where state authorities have asserted their jurisdiction in international mergers, the possibility of such action remains. States traditionally have become involved in mergers concerning markets that are particularly local in scope (i.e., retail industries). As globalisation spreads, however, international companies are more likely to enter what were historically local markets through merger or acquisition. If states were to exercise their jurisdiction in such a situation, the inherent frictions associated with concurrent review would only be magnified. By adding another additional layer of review (in addition to both the federal antitrust agency and any number of international competition agencies that may have jurisdiction), the transaction costs for business only increase. International companies could potentially find themselves subject to multiple requests for relief (within the United States alone), and the time delay and expense associated with multiple reviews could impose significant burdens. Disparate state action also holds the potential to create conflict with existing comity or other international policy considerations established by the federal agencies. Since state involvement in international mergers has not yet become a significant occurrence, the full extent of possible concerns remains unclear. In this regard, it may prove beneficial for federal and state authorities to continue to coordinate joint actions in order to increase communication and understanding of each other's respective policy considerations. Such a dialogue may be important if state attorneys general begin to exercise their jurisdiction over international cross-border transactions.

30

Opposition of Plaintiff Federal Trade Commission to Motion of State of A l a s k a to Intervene, FTCvBP Amoco, Co., Case N o . 000416SI (N.D. Cal. Feb. 22, 2000). 31 In the Matter of BP Amoco, pic, File N o . 991-0192 (Apr. 13, 2000) (consent o r d e r ) . 32 The state of Alaska ran into additional criticism following the settlement agreement with the FTC when BP Amoco reimbursed the state for its legal expenses on behalf of the company's attempt to acquire ARCO. While the state argued that it is common practice for states to recover the cost of the merger investigation from the parties, others criticised the state for a potential conflict of interest. See Paul Queary, Knowles Administration Defends Taking Money From BP Amoco, Associated Press (Feb. 10, 2000).

438

Working Paper IX

IV. Proposals for modification to the dual enforcement regime While the tradition of dual enforcement in the United States is steeped in the legal and philosophical history of the country's development, some observers recently have called for modifications to the system in order to minimise duplication of resources, unduly burdensome requests on the business community, and the potential for inconsistent outcomes. In a recently released report, the International Competition Policy Advisory Committee33 ('ICPAC or the 'Advisory Committee') addresses the role of state authorities within the merger context. The Advisory Committee offers several recommendations in order to clarify the states' involvement in merger proceedings, specifically those mergers with an international dimension. The Report suggests that state authorities refrain from using antitrust laws in order to accomplish non-competition related objectives.34 Further, the Advisory Committee encourages federal antitrust agencies to file amicus curiae briefs in matters where the state attorneys general may decide to challenge a transaction with an international dimension where the federal agency has previously conducted a review. In this regard, the Advisory Committee sees value in informing the court of potential comity or international policy considerations that may be in play with a specific cross-border transaction. Third, the Advisory Committee recommends further study and discussion of overlapping merger review for both federal and state regulators. In a submission presented to the Advisory Committee, members of an American Bar Association ('ABA') committee focusing on international antitrust issues discussed some concerns with the dual enforcement regime structure. While the comments express varying levels of concern regarding concurrent review in the merger context, the discussion highlights this issue's importance within the legal community, particularly as the world's economies become more intertwined. One committee member calls for new legislation designed to preempt state laws dealing with cross-border transactions involving foreign parties, citing the need for the federal government to speak with one voice in matters of international relevance.35 While other members of the ABA 33 The International Competition Policy Advisory Committee was established by the U.S. D e p a r t m e n t of Justice in November, 1997, t o study and make recommendations regarding international antitrust issues within the framework of increasing globalisation. T h e Advisory Committee focused o n three main substantive areas: (1) multijurisdictional merger review; (2) the interface between trade and competition policy; and (3) enforcement cooperation in international cartel activity. The Committee was composed of twelve members from a cross-section of the legal, academic, and business communities a n d was co-chaired by James F. Rill a n d Paula Stern. The Committee released its final report in February 2000. 34 International Competition Policy Advisory Committee Final Report, at 153 (Feb. 28, 2000). 35 Presentation by Members of the International Antitrust Law Committee of the Section of International Law and Practice Before ICPAC Hearings, Comments of Daryl Libow, at 9 (Apr. 22, 1999).

2 - Coherence

439

committee do not, at this time, see the need for legislation to limit the role of state attorneys general, a general consensus exists that maximum coordination between federal and state authorities must continue 'to reduce the risk of duplicate inquiries or inconsistent remedies.'36 Concern also was expressed that any remedy required by state authorities in international matters should be based on competition-related considerations, since the United States is often viewed as an example for nations currently developing or instituting competition regimes. Some observers have noted that the states' emergence as an active antitrust enforcement body has provided additional resources in the overall enforcement of the federal antitrust laws. While this may be true, news reports and testimony before congressional hearings constantly deliver the message that enforcers do not always have the necessary personnel resources to effectively and thoroughly investigate and prosecute violations of the antitrust laws. In light of this, it may seem particularly troublesome to some when scarce resources are used for duplicative or concurrent prosecution or review of a matter. In a recent article published in the newsletter of the ABA's State Antitrust Enforcement Committee, Robert Langer acknowledges this issue and calls for an effort to eliminate the occurrences of duplicative review of a single transaction.37 Mr. Langer's proposal suggests establishment of a new protocol between the Justice Department, FTC, and the states whereby a system for allocation of all antitrust matters would be developed in an attempt to best utilize the limited resources available for antitrust enforcement. Even the author acknowledges the certain difficulties, however, in establishing such an arrangement where antitrust authorities may need to cede jurisdiction over a matter which could be of particular relevance to them. As outlined in some proposals currently circulating regarding such issues in the international context, perhaps a less formal arrangement could be considered in which 'work-sharing' between federal and state authorities could be employed for matters that affect both parties and where one party could adequately resolve the concerns of the other party.38 36

Presentation by Members of the International Antitrust Law Committee of the Section of International Law a n d Practice Before ICPAC Hearings, Comments of Donald I. Baker, at 12 (Apr. 22, 1999). 37 Robert M . Langer, Should The Antitrust Division, the FTC and State Attorneys General Formally Allocate the Market for Antitrust Enforcement, ABA Section of Antitrust Law, State Antitrust Enforcement Committee Newsletter (Apr. 1999). Mr Langer was formerly Assistant Attorney General of the antitrust division for the State of Connecticut and Chairman of the Multistate Antitrust Task Force of the National Association of Attorneys General. 38 In the recent ICPAC report, the Advisory Committee recommends a system of work sharing in cross-border merger investigations where one jurisdiction may refrain from investigating a merger if another jurisdiction sufficiently investigates and remedies all relevant anti-competitive harm affecting both jurisdictions. In such an arrangement, however, each jurisdiction would retain its right t o pursue additional action if t h e remedy o r final outcome did n o t satisfy its concerns. See International Competition Policy Advisory Committee Final Report, at 76-82 (Feb. 28, 2000).

440

Working Paper IX

V. Conclusions The United States' system of a dual antitrust enforcement regime appears to be a permanent landmark in the antitrust enforcement landscape. While inherent conflicts appear in any such system where multiple authorities share jurisdiction, it appears that over the past several years, both the federal and state antitrust authorities in the U.S. have attempted to minimise these frictions through increased cooperation. This cooperation is occurring on many different levels, from enhanced dialogue on a daily basis to full-scale coordination in high-profile litigation matters. Cooperative efforts can significantly aid in reducing conflicts for both the enforcement agencies and the business community. It remains clear, however, that cooperation is not always a panacea for all conflicts that may arise. Continued dialogue and efforts at enhancing current levels of cooperation appear to be useful mechanisms in efforts to avoid undue frictions associated with concurrent federal and state antitrust enforcement. Additionally, consideration should be paid to the possible establishment of work-sharing arrangements in situations where one party has the ability and the will to adequately resolve the concerns of another party. Specific attention also should be paid to the potential involvement of state attorneys general in international cross-border transactions, as these matters hold increased relevance within the framework of general policy and comity considerations. To date, however, concerns in this area have not yet materialised through actual actions by the state attorneys general. As the European Commission and interested parties debate proposals currently under consideration to modify policy and provide EU Member States with jurisdiction over specific competition practices, the experiences of the United States' system of a dual enforcement regime may provide useful background. While there are differences between the United States' dual enforcement regime and the European Union's proposed dual enforcement regime as embodied in the White Paper on Modernisation, it appears that the similarities are great enough to render the experiences of the United States instructive as the European Union devolves more power to the Member States. Specifically, we believe that there are at least five basic lessons that may be drawn from the United States' experience in this arena. Respectfully, it seems desirable that: The manner in which responsibility for specific competition matters will be allocated between the European Commission and the Member States be clearly and formally articulated, so as to provide certainty to businesses and to minimise friction over jurisdictional issues among the various competition authorities. A uniform substantive approach to merger review and other legal areas be articulated through the use of guidelines.

2 - Coherence

441

The European Commission and the Member States consider the routine use of work-sharing arrangements similar to those employed by the Department of Justice and the state attorneys general in Primestar39 and WorldCom/MCI.40 Mechanisms could be established to insure open lines of communication and to facilitate cooperation between the European Commission and the Member States, as, for example, in the proposal of the White Paper on Modernisation to strengthen the role of the Advisory Committee on Restrictive Practices and Dominant Positions. Retention of the European Commission's authority to intervene in national court cases may be useful, so as to further insure uniformity in the application of European competition laws. As noted in this paper, the filing of amicus briefs by state attorneys general in cases brought by the federal antitrust authorities, and vice versa, has been an important form of cooperation in the United States' dual enforcement regime. It is hoped that, if the EC decides to implement the proposals outlined in the White Paper on Modernisation, both the successes and the shortcomings of the U.S. system may provide beneficial examples of mechanisms to identify and minimise the inherent conflicts and frictions associated with any system of dual authority.

39

United States v Primestar Pamters, L.P., 1994-1 Trade Cas. (CCH) \ 70,562 (S.D.N.Y. 1994); New Yorlcv Primestar Partners, L.P., 1993-2 Trade Cas. (CCH) \ 70,404 (S.D.N.Y. 1993). 40 'Commission Clears WorldCom and MCI Merger Subject to Conditions,' European Commission Press Release (July 8, 1998); 'Justice Department Clears WorldCom/MCI Merger After M C I Agrees to Sell Its Internet Business,' U.S. Department of Justice Press Release (July 15, 1998).

X Mario Siragusax The Modernisation of EC Competition Law: Risks of Inconsistency and Forum Shopping

I. Introduction The Commission's proposal to end the exemption and notification system established by Regulation 17/622—as laid down in the White Paper on Modernization of the Rules Implementing Articles 81 and 82 EC Treaty,3 hereafter 'the White Paper'—raises a number of concerns. Under the proposed enforcement system, the Commission would loose its sole jurisdiction to apply Article 81 (3), and this provision would become directly applicable. Consequently, parties to agreements, decisions or concerted practices could invoke Article 81 (3) before a national competition authority (hereafter 'NCA') or national court, 4 and would no longer be entitled to notify the Commission of their agreements to seek an exemption decision or a negative clearance.5 One of the main concerns is that the proposed reform might lead to inconsistent applications of EC law.6 Another concern is that the new rules might encourage forum shopping between different Member States. Both concerns are based on the assumption that (i) the reform will lead to a more decentralised application of Article 81 by NCAs and that (ii) no clear and binding rules to allocate cases between competition authorities and to prevent conflicting

1 The author wishes to acknowledge the invaluable contribution of his colleague Filippo Amato. 2 Council Regulation No. 17/62, OJ 13 (February 21, 1962) 204 (Special Edition 1959-62, p. 87), as amended. 3 OJ 1999 C 132/1. Although the White Paper uses the old numbering of the Treaty establishing the European Community (the 'Treaty'), all quotations in the present text are updated to the new numbering resulting from the entry into force of the Treaty of Amsterdam. 4 Thus, in a way, the reform would end the division between Article 81(1) and (3) EC in terms of market analysis, since it would no longer matter under which paragraph this analysis was carried out. 5 The principal declared purposes of the reform are the reduction of the workload created by notifications to allow the Commission 'to use its resources to combat cartels, particularly in concentrated markets and in markets which are being liberalised' (White Paper, para. 45), and the decentralisation of the application of EC competition law. The relief of administrative constraints on undertakings is also mentioned as an objective of the reform (White Paper, para. 42), but this seems to tie in more with a reduction in the Commission's workload than being an objective in itself. 6 This concern is raised by the Commission itself (White Paper paras. 101-103).

444

Working Paper X

decisions based on divergent interpretations of Article 81 (3) will be established.7 It is doubtful, however, that the proposed reform will lead to a more decentralised application of Article 81 by NCAs. In any event, very simple procedural safeguards might be provided to avoid the risk that NCAs open parallel proceedings or adopt conflicting decisions based on Article 81. It is also doubtful that the reform would lead to an increased number of Article 81 decisions by national courts. On the other hand, the reform will allow national courts to apply Article 81 (3). The real issue is therefore how to ensure that, under the new system, national courts will apply Article 81 (3) in a uniform and consistent way. Although a notification system would be compatible with a system whereby Article 81 (3) is directly applicable, and would also ensure a better degree of legal certainty without frustrating the objectives of the proposed reform,8 this paper discusses the concerns mentioned above and suggests possible solutions on the assumption that no notification system will be maintained under the new enforcement rules.

II. Decentralised application of EC competition law 1. Application of EC competition law by NCAs As explained above, the core of the Commission's proposal is to render the provision laid down in Article 81 (3) directly applicable. According to the Commission, this would improve the decentralisation of EC competition law since NCAs 'could investigate cases of application of Community law, in response to complaints or on their own initiative, without fear that the parties might invoke a notification pending before the Commission.'9 Since at present only eight NCAs are specifically empowered under their national laws to apply Articles 81 and 82, l0 the Commission acknowledges that to achieve a real decentralisation 'the seven Member States that have not yet done so will have to empower their competition authorities to apply Community law.'11 These statements, however, do not reflect reality for a number of reasons. First, to create the minimum conditions for a decentralised application of EC competition law by NCAs, all fifteen Member States (and not only seven, 7

See Ehlermann (2000), at 50-51. See in this regard Siragusa (1999) and Ehlermann (2000). 9 White Paper para. 93. 10 Belgium, France, Germany, Greece, Italy, the Netherlands, Portugal, and Spain. Denmark and Ireland do not consider specific empowerment necessary. See Siragusa (1998, 16). 1 ' White Paper para 94. 8

2 - Coherence

445

as the Commission seems to claim), will have to adopt new legislation empowering NCAs to apply Article 81 in its entirety. This is due to the fact that even the eight Member States that have already empowered their NCAs to apply Article 81 have done so only with regard to paragraph 1 of that Article. Second, assuming that all NCAs were empowered to apply Article 81 (3), it is doubtful whether they would apply Article 81 more than they currently do. It is in fact extremely likely that, regardless of whether the reform includes specific rules on the allocation of cases between the Commission and NCAs, the latter would still tend to investigate alleged infringements producing effects only or mainly within the territory of their own State.12 In these circumstances, it is highly unlikely that NCAs would prefer to apply Article 81, rather than their own national laws, simply because Article 81 (3) is a directly applicable provision.13 On the contrary, a number of reasons militate against increased instances of application of Article 81 by NCAs. Under the proposed system, in fact, an NCA that wished to apply Article 81 (3) would have to accept not only the judicial review of its decision by national courts, but also possible interference by the Commission,14 by other NCAs, and, last but not least, by the Court of Justice. Moreover, if the alleged infringement affected more than one Member State, the NCA that opened proceedings under Article 81 would also need to co-operate with other NCAs, notwithstanding the technical and practical problems (exchange of confidential information, linguistic barriers, conflicting views, etc.) that such co-operation might imply. Finally, the White Paper does not foresee any specific measure to increase the application of Article 82 by NCAs or to involve such authorities to a greater extent, for instance through an extended use of Article 9 of the EC Merger Regulation, in the review of concentrations having a Community dimension.15 In light of the above it is therefore very improbable that, if implemented, the White Paper's proposals would lead to a more decentralised application of EC competition law.16 NCAs would still prefer to apply their own national laws, unless in exceptional circumstances it appeared more advantageous to apply EC law, such as for example the possibility to condemn (pursuant to Articles 81 12

See Bruzzone and Todino (2000, 18). This is confirmed, for instance, by the fact that although Article 82 has always been directly applicable, it has not been applied by NCAs more often than Article 81 (1). 14 One of the main reasons that currently leads NCAs to apply their own national law rather than Articles 81 (1) and 82 is that, pursuant to Article 9 (3) of Regulation 17/62, the Commission may at any time take up the case and deprive them of the power to apply those provisions. However, the Commission intends not only to maintain a mechanism equivalent to that in Article 9 (3) of Regulation 17/62 (White Paper, para. 105), but even to reserve to itself the right to prohibit agreements or practices that NCAs might have already considered, by decisions no longer open to appeal, to satisfy the conditions of Article 81 (3) (White Paper para. 102). 15 Council Regulation No. 4064/89, OJ L 395/1, as amended. See Bruzzone and Todino (2000, 26-32) with regard to the need to decentralise the control of concentrations having Community dimension. 16 Bruzzone and Todino (2000, 18). 13

446

Working Paper X

or 82), agreements or practices that it would not be possible to prosecute under national competition law. Consequently, the risks of inconsistent applications of EC competition law by competition authorities, which might potentially result from the implementation of the proposed reform, have probably been overestimated.

2. Application of EC competition law by national courts The White Paper states that the direct applicability of Article 81 (3) would improve decentralisation of EC competition law since complainants, with a view to obtaining more quickly obtaining the declaration of nullity and/or damages where they are victims of illegal agreements, would more often invoke Article 81 before national courts.17 However, it is unlikely that Article 81 proceedings before national courts would become quicker or would increase in number, merely because national courts obtain the power to apply Article 81 (3).18 Most national courts might not feel at ease in applying a provision whose interpretation focuses more on policy and economic considerations rather than general principles of law and might prefer to seek advice from the Commission,19 to refer the matter to the Court of Justice pursuant to Article 234 EC,20 or—if the Commission or an NCA have already initiated investigations concerning the same contested practice—to stay the proceedings until the outcome of such investigations.

III. Risks of inconsistent application of EC competition law 1. Risk of conflicts between competition authorities A substantial increase in the application of EC competition law by NCAs as a consequence of the proposed reform is doubtful. However the possibility still remains that—in cases having a Community dimension and requiring action by the Commission—an NCA might decide to open an investigation under EC or national law, or two or more NCAs might decide to open parallel investigations either under national law, EC competition law, or both laws. 17

White Paper p a r a . 100. D o u b t s as t o w h e t h e r t h e direct applicability of Article 81 (3) might lead to an increased application of Article 81 b y national courts are raised by m a n y authors. See, inter alia, K o n (1999) a n d B r u z z o n e a n d Todino (2000, 19-20). 19 Such a possibility h a s been expressly provided by the C o u r t of Justice in Case C - 2 3 4 / 8 9 Delimitis v Henninger Brdu [1991] E C R 1-935, para. 53, a n d confirmed by the C o m m i s s i o n in its Notice on co-operation between national courts and the Commission in applying Article 81 and 82 EEC Treaty, O J C 39/6 1993, p a r a s 3 7 ^ 0 . 20 Ex-Article 177 E E C Treaty. 18

2 - Coherence

If an NCA's decision is based on national competition law, there is a risk of conflict with previous or subsequent decisions adopted pursuant to EC competition law by either the Commission or another NCA. The following principles would probably apply. If the Commission or another NCA takes the view that an agreement or practice does not fall within Article 81 (1), any NCA may still declare that the same agreement or practice infringes national competition law.21 More problematic is determining the effect on national competition law of a decision of the Commission or an NCA declaring that, although an agreement or practice restricts competition pursuant to Article 81 (1), such agreement or practice satisfies the conditions of Article 81 (3). Although there is no clear case law on the issue,22 under the current system it is generally thought that a Commission's exemption decision may not be questioned under national competition laws.23 In our opinion the same principle should also apply in a system where Article 81 (3) is a directly applicable provision. However, in such a system, a risk of inconsistent application of EC law would also arise when (1.1.) an NCA adopts a decision pursuant to Article 81, and the Commission or another NCA intend to adopt a different decision based on the same provision, or when (1.2.) the Commission adopts a decision based on Article 81 and an NCA intends to adopt a different decision based on the same Article. 1.1. Risks of conflicting decisions in cases of alleged infringement of Article 81 which an NCA deals with first The White Paper suggests that the Commission should always be allowed to adopt a prohibition decision if it disagrees with an NCA's non-prohibition decision, regardless of whether the latter decision is still open to appeal or is already definitive according to the relevant national law.24 On the contrary, when an agreement has been prohibited by an NCA, the White Paper states that the Commission 'should not normally seek to intervene', regardless of whether it 21 See Joined Cases 253/78 a n d 1 to 3/79 Giry and Guerlain [1980] E C R 2327, p a r a 18, where the Court stated that ' T h e fact that a practice has been held by the C o m m i s s i o n not to fall within the ambit of the prohibition contained in Article 85 (1) a n d ( 2 ) , t h e scope of which is limited to agreements capable of affecting trade between m e m b e r States, in n o way prevents that practice from being considered by the national a u t h o r i ties from the point of view of the restrictive effects which it may produce nationally'. 22 E u r o p e a n Commission: Notice on co-operation between national competition authorities and the Commission, O J C 313/3 1997, para 19. 23 This conclusion is generally drawn from the j u d g m e n t in Case 14/68 Walt Wilhelm v Bundeskartellamt [1969] E C R 1, para. 6, where the Court stated that '. . . conflicts between the rules of the Community and national rules in the mater of the law o n c a r tels must be resolved by applying the principle that Community law takes precedence.' See Korah (1997, 24-25). Some a u t h o r s are, however, of a different view—see, for instance Doherty (1994). 24 White Paper paras. 102(2) a n d (4).

448

Working Paper X

deems that such an agreement could (either as it stands, or subject to certain amendments or undertakings given by the parties to the agreement) fall outside the scope of Article 81 (1) or fulfil the conditions of Article 81 (3).25 These rules are highly unsatisfactory for a number of reasons. First, if implemented, they would in practice transform the Commission into an appellate body against NCA's non-prohibition decisions adopted under EC law, a role for which the Commission is certainly not suited.26 Second, they would unfairly discriminate between complainants and parties suspected of a restrictive practice, since while the former could always ask the Commission to re-open proceedings and to reverse an NCA's non-prohibition decision, the parties to an agreement prohibited by an NCA would not enjoy a similar option, although the Commission could probably be persuaded that the agreement as such, or subject to certain modifications or undertakings given by the parties, may satisfy the conditions of Article 81 (3). Third, although such rules intend to solve possible conflicts between different bodies applying EC competition law,27 they in fact do exactly the contrary: in a system where Article 81 (3) becomes directly applicable, an NCA's decision declaring that such a provision applies to a restrictive agreement would have exactly the same value of as a conflicting decision by the Commission unless a clear hierarchy between such decisions was established by law. Fourth, the White Paper does not propose any rules to solve conflicts between different NCAs. Finally, it would certainly be better, as the White Paper recognises, to establish a mechanism based on the avoidance of conflicts rather than a mechanism based on their solution.28 In our opinion, a relatively simple mechanism for avoiding conflicting decisions under a system of the direct applicability of Article 81 (3) would be to have a provision similar to that of Article 9(3) of Regulation 17/62, according to which NCAs may apply EC competition law as long as the Commission has not initiated any procedure, combined with an obligation on NCAs to inform the Advisory Committee on Restrictive Practices (the 'Advisory Committee') of any decision they intend to adopt pursuant to EC competition law.29 The Advisory Committee would become a forum in which all other proposed decisions by NCAs based on EC competition law would be discussed to ensure that other competition authorities (the Commission or other NCAs) that might potentially adopt a conflicting decision, could reach a common agreement. When such a common agreement is reached, a prohibition decision adopted by an NCA pursuant to Article 81 should not be susceptible to being overruled by another NCA on the basis of the same provision. Similarly, neither the Commission nor any NCA entitled to apply Article 81 should be able to ques25

White Paper para. 102(3). See Ehlermann (2000, 59). 27 White Paper para. 103. 28 White Paper, ibidem. 29 Similar rules are also suggested by the Commission, although no proposals are made on how they should work in practice (White Paper para. 105). 26

2 - Coherence

449

tion non-prohibition decisions adopted by NCAs pursuant to Article 81 (3), subject to a certain time limit30 and except in exceptional circumstances (such as a dramatic change of market conditions or a finding that the decision was based on wrongful information). If, on the contrary, agreement is not reached within the Advisory Committee, or if two or more NCAs have opened parallel proceedings and a real risk of conflicting decisions is envisaged, the Commission should take up the case and decide it by way of a formal decision.31 Of course, the Commission's prohibition or non-prohibition decisions would also be binding on all NCAs for a specified period, except in exceptional circumstances. 1.2. Risks of conflictive decisions in cases which the Commission deals with first The White Paper provides that in cases where the Commission has initiated proceedings, or where it has adopted a definitive decision, NCAs should be bound to avoid conflictive decisions, if necessary, by suspending the proceedings before them.32 These principles are in line with the rules proposed above. That is, with the rule according to which the Commission should always be entitled, when an agreement on a NCA's planned decision is not reached within the Advisory Committee, to take up the case and decide it itself, and with the rule according to which a competition authority's decision based on Article 81 should bind the other competition authorities potentially entitled to apply the same provision.

2. Risk of conflicts between competition authorities and national courts Under the present system, where the Commission is already seized of a matter, the national court has a duty under Article 10 EC to avoid inconsistent decisions and can suspend proceedings before it and request information from the Commission.33 This principle would certainly apply if Article 81 (3) becomes a directly applicable provision. Another fundamental principle of the current enforcement system is that individual exemption decisions adopted by the Commission are binding on 30 Similarly to exemption decisions, non-prohibition decisions based on Article 81 (3) should also be issued for a specified period of time, to allow the possibility to reexamine, after a certain period, whether the conditions of the market still justify the application of Article 81 (3). 31 See, for a very similar proposal, Temple-Lang (1999, 13). 32 White Paper para. 102(1). 33 See Delimitis v Henninger Bru, cit., para. 52.

450

Working Paper X

national courts.34 This principle has traditionally been based on the consideration that, pursuant to Article 9(1) of Regulation 17/62, the Commission has exclusive jurisdiction to apply Article 81 (3).35 One might therefore wonder whether, under a system of the direct applicability of Article 81 (3), national courts would be bound (2.1.) by a formal Commission decision or (2.2.) by an NCA's decision declaring that an agreement or practice fulfils the conditions of Article 81 (3).36 2.1. Effect on national courts of Commission decisions based on Article 81 The White Paper seems to suggest that under the new system—unlike individual exemption decisions—formal Commission decisions declaring that Article 81 (3) applies to a certain agreement would not be binding upon national courts.37 Accordingly, any national court could declare the same agreement void under Article 81(2). This would be an undesirable result not only because of the potential for inconsistency in the application of EC competition law, but also from the point of view of legal certainty for the undertakings involved. However, the White Paper's position does not take into account the fact that Community acts, even those based on directly applicable EC provisions, cannot be disregarded by national courts. In fact, by disregarding a Community act (irrespective of whether or not it is based on a directly applicable provision), a national court would implicitly hold such an act invalid, contrary to the principle according to which Community acts may not be declared invalid by national courts.38 According to such a principle, when a national court has doubts about the validity of a Community act (such as a Commission decision), that court may not declare such an act invalid, but must refer it to the Court of Justice for a preliminary ruling on the validity of the act.39 In fact, pursuant to the EC Treaty, only the Community Courts (Court of Justice and 34

See Commission Notice on co-operation between national courts and the Commission, cit., para. 25(a). 35 For the same reason, the Commission's formal negative clearance decisions are generally considered not t o bind national courts, since Article 81 (1) is a directly applicable provision. See, in this sense, the Opinion of Advocate General Reischl in Case 37/79 Marty v Lander [ 1980] E C R 2481, at para. 2507; Opinion of M r Kirschner in Case T-51/89 Tetra Pak v Commission [1990] E C R 11-309, at p. 345 para. 104; Opinion of Advocate General Van Gerven in Case C-313/93 Banks [1994] E C R 1-1279, at p. 1264, para. 59. 36 The question, of course, regards only formal Commission decisions. Comfort letters d o not bind national courts, but only constitute an element of fact which national courts and national competition authorities may take into account (see Case C-99/79 SA Lancome and Cosparfranee Nederland BVx Etos BVand Albert Heyn Supermarket 5K[1980] E C R 2511). 37 White Paper para. 89. 38 Case 314/85 Fotofrost [1987] E C R 4199. 39 Fotofrost, cit., para. 17.

2-Coherence

451

Court of First Instance) have jurisdiction to declare Community acts invalid. If national courts had such a power, '. . . divergences between courts in the Member States as to the validity of Community acts would be liable to place in jeopardy the very unity of the Community legal order . . .'40 Therefore there might be no need under the new system, as some authors have suggested,41 to confer via Council regulation binding effect on Commission non-prohibition decisions adopted pursuant to Article 81 (3), since such an effect already constitutes an established principle of community law. The same principle would apply also to Commission prohibition decisions based on Article 81 (I).42 It cannot be excluded, however—given the possibility of changing market conditions—that a Commission decision declaring that an agreement or practice satisfies the conditions of Article 81 (3), might no longer reflect reality at some point subsequent to the adoption of the decision. That is the reason why, pursuant to Article 8 (1) of Regulation 17/62, exemption decisions may currently be issued only for a specified period of time. Therefore, to allow national courts to re-examine market conditions after a Commission non-prohibition decision based on Article 81 (3) has been adopted, the amended version of Regulation 17/62 should provide that such decisions can be issued for only a specified period. In addition, or alternatively, the same Regulation could provide that if market conditions change dramatically within a short period from the adoption of the Commission non-prohibition decision, national courts may declare the agreement incompatible with Article 81 (1), provided they consult the Commission and provided the Commission agrees with their view. 2.2. Effect on national courts of NCAs' decisions based on Article 81 Under the rules proposed above, NCAs' prohibition or non-prohibition decisions based on Article 81 should be binding for a specified period on other NCAs or on the Commission, absent exceptional circumstances. On the other hand, the effects on national courts of NCA prohibition decisions based on Article 81 (1) or of positive decisions based on Article 81 (3) are another matter. Since the case law of the Court of Justice concerning the binding effect of Community acts on national courts would not apply to NCAs' decisions, these decisions might be binding on national courts only if that was expressly established by Council Regulation.43 However, by preventing national courts from 40

Fotofrost, cit., para. 15. See Ehlermann (2000, 47). 42 See, also, Opinion of Advocate General Van Gerven in Banks, cit., at p. 1264, para. 60, who expressly mentions Fotofrost to support this conclusion. 43 That would probably be possible, according to Ehlermann (2000, 52), on the basis of Article 83 EC Treaty (ex-Article 87). 41

452

Working Paper X

having jurisdiction over decisions adopted by a national administrative body, such a Regulation would certainly raise serious problems of constitutional law, that would be difficult if not impossible to overcome in many Member States. An enforcement system whereby NCAs' decisions adopted pursuant to Article 81 (1) or 81 (3) have no binding effects, is therefore the most likely scenario. Such decisions—once no longer open to appeal, or confirmed on appeal—would probably have the same effect that the Commission's comfort letters currently have.

3. Risk of conflicts between national courts The effects described above of Commission proceedings and of Commission decisions on national courts, the rules on Us pendens and related actions laid down by Articles 21 and 22 of the Brussels Convention,44 and Article 234 of the EC Treaty, should avoid the risk of conflicting decisions by different national courts applying Article 81 (1) and (3) to the same agreement or practice. So, if a case involving an action for damages between the same parties arising from the same alleged infringement of Article 81 is brought in the courts of different Member States pursuant to Article 21 of the Brussels Convention any court seized after the first must stay its proceedings until the latter has established its jurisdiction. Once such jurisdiction has been established, any court seized after the first must decline jurisdiction. On the other hand, if a case involving the same alleged infringement of Article 81 is brought by different parties against the same defendant in courts of different Member States (so-called related actions),45 and these actions are all pending at first instance, Article 22 of the Brussels Convention confers on any court other than the court first seized the discretion to stay its proceedings and, under certain circumstances, to decline jurisdiction. Although these rules do not always prevent contradictoryfindingsrelating to the same agreement or practice by national courts of different Member States, they should reduce the likelihood of this outcome. In any event, a national court that had doubts concerning the compatibility with Article 81 of an agreement that had already been declared to satisfy the conditions of Article 81 (3) by an NCA of another Member State or by a national court of another Member State—although in an action between different parties—would likely make an Article 234 EC reference to the Court of Justice. 44 Convention on jurisdiction and the enforcement of judgments in civil and commercial matters, O J C 189/1 1990. 45 Article 22 (3) of the Brussels Convention considers 'related' actions 'so closely connected that it is expedient to hear and determine them together to avoid the risk of irreconcilable judgments resulting from separate proceedings.'

2 - Coherence

453

4. Summary of the proposed rules To sum up, according to the rules proposed above, Commission decisions based on Article 81 should be binding, for a specified period, not only on NCAs but also on national courts under the new enforcement system. On the contrary, an NCA's prohibition or non-prohibition decision based on Article 81 would be binding on other NCAs, but not on national courts. These rules would be based, on the one hand on already existing principles of EC law, such as the principle that Community acts are binding on national courts, and on the other hand on a few specific procedural safeguards, such as the obligation on NCAs to inform the Advisory Committee of any intention to adopt a decision based on Article 81, and the Commission's power to withdraw NCAs'jurisdiction to apply Article 81 in all cases where there is a potential conflict between the latter's planned decisions and the view of the Commission or of another NCA.

IV. Risks of forum shopping Another concern raised by the White Paper is that the proposed rules might encourage forum shopping. Forum shopping can be defined as the complainants' freedom to file a complaint with a competition authority which is likely to take the action most favourable to the complainant's interests, or to start an action before a 'friendly' national court. This concern is essentially a corollary of the concern examined above, and is therefore also based on the assumption that (i) the reform will lead to a more decentralised application of Article 81 by NCAs and that (ii) no clear and binding rules to allocate cases between competition authorities and to prevent conflicting decisions based on divergent interpretations of Article 81 (3) will be established.46 It is clear, in fact, that if under the new rules more than one NCA may have jurisdiction to open proceedings in the same case, and if certain NCAs are allowed to interpret Article 81 (3) more strictly than others, companies alleging an infringement of Article 81(1) would prefer to file their complaint with the former rather than the latter. 1. Forum shopping before NCAs As has already been explained, it is doubtful that the proposed reform will lead to a more decentralised application of EC law by NCAs. In any event, even if 46

See Ehlermann (2000, 50-51).

454

Working Paper X

such decentralisation was really achieved, the rules proposed above to prevent the risk of inconsistent application of EC law by NCAs, should be sufficient to deter complainants from forum shopping. Moreover, the fact that NCAs' decisions adopted pursuant to Article 81 lack Community-wide binding effect could also constitute a deterrent to forum shopping. In fact, companies alleging an infringement of Article 81 would probably have a less strong incentive to lodge a complaint with the NCA of a Member State potentially favourable to their interests, if later they would not be able to rely on that NCA's decision before their own national courts.

2. Forum shopping and national courts Since the risk of forum shopping arises in the absence of rules on the allocation of cases, such a risk would be unlikely to arise with regard to actions before national courts brought on the basis Article 81, since national courts' jurisdiction within the EU is clearly determined by the Brussels Convention, and by national law, within the territory of a Member State.47

V. Conclusions The concerns about inconsistent applications of EC competition law and of forum shopping raised by the Commission's reform proposed in the White Paper have probably been overestimated since they are based on the assumption that such reform will lead to a decentralised application of EC law. However, it is doubtful, for a number of reasons, that this will be the case. In any event, simple procedural rules, such as the possibility that the Commission could take over a case from a NCA according to a mechanism similar to that of Article 9 (3) of Regulation 17/62, and an information mechanism that allows discussion within the Advisory Committee of all cases where Article 81 is applied by NCAs, would probably prevent any serious risk of inconsistent application of EC law by NCAs. Although these rules are simple and flexible, there will need to be effective provisions for their implementation. In this respect, it is worrying that, although the White Paper stated that 'the Commission would [. . .] have to be informed if an authority planned to withdraw the benefit of a block exemption', the recent amendment of Article 7 (2) of Regulation 19/65,48 (allowing NCAs to withdraw the benefit of a block 47

See Schuz (1986, 375). Article 7 (2) has been added in Regulation 19/65 by Regulation 1215/99. A similar provision is also contained in Article 7 of the recently adopted block exemption regulation on vertical restraints (Regulation 2790/1999, OJ L 336/21 1999). 48

2 - Coherence

455

exemption under certain circumstances), does not provide any duty on the NCAs to provide information.49

49 The Draft Guidelines state at para. 68 that the consultation mechanism provided for in the Notice on co-operation between national competition authorities and the Commission would avert the risk of conflicting decisions and duplication of procedures. That Notice provides at para. 49 that 'as regards cases which they deal with under Community law, it is desirable that national authorities should systematically inform the Commission of any proceedings they initiate'. The problem remains, however, that the provision of such information is entirely voluntary and it does not allow the Commission to make remedies if it does not agree with a national competition authority's decision to withdraw the block exemption, unless the parties concerned decide to notify the agreement for an individual exemption. Only in such a case could the Commission open proceedings according to Article 9 (3) of Regulation 17/62, depriving the national competition authority of the power to apply Article 81 (1) and probably also of the power to withdraw the block exemption.

XI James S. Venit1 The Decentralised Application of Article 81: Italian Banks, Cohabitation, Private Enforcement and other issues Raised by the White Paper

I. Introduction This paper examines three sets of issues that are central to the proposed White Paper reforms.2 The first set of questions centres around whether private enforcement is a better and more efficient alternative to regulatory action by the national authorities (and the Commission) outside the areas of the Commission's newly defined enforcement priorities, e.g. mergers and other structural arrangements, cartels and abuses of dominant positions. The issues to be examined here are: First, whether the decentralised enforcement of Article 81 by national competition authorities (hereafter referred to as 'cohabitation'), as distinguished from private enforcement by national courts, is an inevitable and necessary consequence of the Commission's proposal to reorder its enforcement agenda and reform the application of Community competition law. Second, whether, with the possible exception of cartels, reliance on private enforcement of Community law is preferable to its enforcement by the national authorities. Third, whether, in a 'decentralised' post-White Paper world in which the Commission has abandoned its monopoly over Article 81 (3), the application of national competition law no longer raises the same conflictive Walt Wilhelm3 type of issues as it did in the centralised pre-White Paper world. The second set of questions revolves around the Court of Justice's judgment in Bagnasco v Banca Popolare di Novara et. al. (hereafter Italian Banks),4 a

judgment on an Article 234 (ex Article 177) reference which appears to have significantly limited the scope for the application of Article 81 to national agreements. These questions are: First, whether Italian Banks precludes the application of Community competition law to agreements that effect consumers in only one Member State, 1 The author would like to express his gratitude to his colleague Andrea De Matteis for his invaluable assistance with this paper. 2 European Commission: White Paper on modernisation of the rules implementing articles 85 and 86 of the EC Treaty, OJ C 132/1 of 12 May 1999. 3 Case 14/68 Walt Wilhelm v Bundeskartellamt [1969] ECR 1. 4 Joined Cases C-215/96 and C-216/96 Carlo Bagnasco and others v Banca Popolare di Novara s.c.a., Cassa di Risparmio di Genova e Imperia SpA (Carige) [1999] ECR 1-135.

458

Working Paper XI

thereby eliminating the one area—primarily domestic agreements that nevertheless arguably affect trade between Member States—where the decentralised application of Article 81 might have raised fewer problems of coherence. I.e. the application of Article 81 by the national authorities would give rise to fewer risks when applied to agreements primarily affecting national markets that may require and legitimise decentralised treatment. Second, if Italian Banks precludes the application of Community competition law to 'national' agreements, does the White Paper become an invitation to national competition authorities to apply Community competition law to agreements which most perforce have effects outside their jurisdiction, thereby maximising the risks of incoherence? Third, does Italian Banks shift decentralisation to the level of national competition law? If so, does that mean that the national authorities will not play an active role in the decentralised enforcement of Article 81, but may find themselves burdened with an increasing number of notifications under national laws based on Article 81 and Regulation 17/62? The third set of questions focuses on whether the uniform application of the competition rules at the national level is in all cases desirable and considers the degree of diversity that can be tolerated in a mature system of competition law with reference to the experience in the United States.

II. The proposed reforms—background The reforms proposed in the White Paper are the product of a re-ordering of the Commission's enforcement priorities. This re-ordering properly puts structural concerns (merger control expanded to include production joint ventures) and enforcement of the prohibitions on cartels and abuses of dominant positions at the forefront of the Commission's enforcement agenda. Vertical agreements and less pernicious horizontal arrangements are to be the subject of Commission guidelines, but, as a result of the Commission's abandonment of its monopoly over the granting of exemptions, will be subject to decentralised control by the national authorities and national courts, although this 'control' 1 will fall short of granting formal exemptions under Article 81 (3) pursuant to a system requiring prior notification but, rather, will be limited to assessment of the 'exemptibility' of agreements.5 5

The publication of the White Paper initially gave rise to some confusion as to whether the national authorities would be able to grant exemptions pursuant to a procedural mechanism involving or requiring prior notification as under Regulation 17. In informal comments on the White Paper, Commission officials have indicated that this is not the intention. Thus, at the level of the national authorities, application of Article 81 (3) would, if this approach is followed, mean that the national competition authorities (and courts) will be able to assess the applicability of Article 81 (1) and 81 (3) at the same time without, however, requiring notification and, conceivably, without adopting formal

2 - Coherence

459

The demarcation of competence proposed by the White Paper: The Commission has proposed to surrender its monopoly over the power to grant exemptions under Article 81 (3). As a result, Article 81 (3) will be directly enforceable in national courts. The consequence is that not notified restrictive agreements will no longer be automatically unenforceable and companies will have an Article 81(3) defence when sued for damages for the infringement of Articles 81 in the absence of notification. On the other hand, decentralised application of Article 81 (3) will foster more frequent resort to national courts, given the latter's power to render judgments and award damages in competition cases. The absence of a prior review mechanism centralised in the Commission and the Commission's discretion over whether to consider complaints will encourage private enforcement. In its proposed form, the market sharing arrangement described in the White Paper gives the national competition authorities and courts competence over certain arrangements that could have cross-border implications, such as cooperation arrangements, alliances, cartels, licensing, distribution and joint research and development co-operation that stops short of joint production. The debate about coherence and forum shopping that has followed the publication of the White Paper might well have been avoided if national authorities had not been given the ability to apply Article 81 in its entirety, or if their ability to do so were limited to agreements and practices that affect only the consumers in their Member State. Whilst this restraint on their competence may not have been attractive or politically acceptable to the Member States, such a result would clearly be preferable if simplicity and avoidance of conflict are first order priorities. This apercu provides the background for the first issue discussed below.

III. Cohabitation with the national competition authorities concerning the application of Article 81 (3) 1. Is extension to the national competition authorities of powers under Article 81 an inevitable consequence of the reordering of the Commission's enforcement priorities? The answer to the threshold question—whether the Commission could have accomplished its goals of reprioritising its agenda without sharing its powers under Article 81 (3) with the national authorities—is no doubt 'yes' from a legal and institutional perspective, but apparently 'no' from a political perspective. decisions to exempt. It is evident that the same considerations that have prompted the Commission's reforms argue against the adoption of procedures at the national level that would require or make possible the prior notification of restrictive agreements.

460

Working Paper XI

Legally, the Commission could have accomplished its goals by (i) securing the amendment of Regulation 17/62 to eliminate the requirement for prior notification; and/or (ii) consistent with the Court's jurisprudence,6 applying Article 81 (1) in a less inclusive fashion, without formally abandoning the Treaty-mandated bifurcation between Articles 81 (1) and 81 (3).7 The first approach, which by itself would have been sufficient, would have eliminated the procedural obstacles to the unified application of Article 81. The second approach, which would not be strictly necessary in light of thefirst,would, without the need for amendment of the Treaty, have enabled the Commission to eliminate a large number of cases by disapplying Article 81 (1), thus leaving the application of Article 81 (3) to very rare and statistically insignificant number of cases. The history of the White Paper suggests that, once the Commission embarked on a course that required Council action—the amendment of Regulation 17/62 to eliminate the requirement for prior notification—it became necessary to expand the powers of the national authorities. Thus, it appears that the price that will be exacted by the Member States for amending Regulation 17'62 is the extension to the national competition authorities of the power to apply Article 81 in is entirety. 2. Is cohabitation with the national competition authorities of powers under Article 81 a desirable consequence of reform? This section treats the second and third questions posed in the introduction under a single heading. 6 See, inter alia, Case 258/78 L. C. Nungesser KG and Kurt Eisele v Commission [ 1982] ECR 2015; Case 262/81 Coditel SA, Compagnie General pour la Diffusion de la Television, and Others v Cine-Vog Films SA and Others [1982] ECR 3381; Case 27/87 Louis Erauw-Jacquery SPRL v La Hesbignonne SC[1988] ECR 1919; Case 42/84 Remia BVand Others v Commission [1985] ECR 2545; Case 161/84 Pronuptia de Paris GmbHv Pronuptia de Paris IrmgardShillgalis [1986] ECR 353; and Case C-250/92 Gottrup-Klim Grovvareforening and Others v Dansk Landbrugs Grovvareselskab AmbA (DLG) [1994] ECR 1-5641. 7 This view is clearly not shared by those who take a strict view of Article 81 (1) and who view the overly broad application of the latter in the Commission's prior administrative practice as being mandated by the Article 81 (1)/81 (3) split. The White Paper in essence seeks to avoid this issue by focusing on (and eliminating) the procedural bifurcation of Article 81 which is the product of the system of prior notification enshrined in Regulation 17/62, thereby avoiding the substantive legal debate concerning the scope of Article 81 (1) by enabling the Commission, the Courts and the national authorities to consider the application of both Articles 81 (1) and 81 (3) at the same time in a single procedure.

2 - Coherence

461

In speaking of decentralisation it is essential to distinguish between the roles of the national courts and of the national competition authorities. The less inclusive application of Article 81(1) and/or the abandonment of a system of prior notification, coupled with the re-prioritisation of the Commission's enforcement agenda, could be expected to lead to more vigorous private enforcement of Article 81 in the national courts as a result of the need to fill the vacuum and the national court's greater power to do so, resulting in its turn from the elimination of the need to defer to the Commission for the application of Article 81 (3). Indeed, increased private enforcement should be viewed as a positive development which would, in effect, provide an outlet for market forces that was arguably not always available under the current system because of distraction and resource strain caused by the system of prior notification. Reform need not, however, inevitably lead to the increased application of Article 81 by the national competition authorities. Indeed, outside the area of enforcement priorities such as cartels, such a development can be seen as, on balance, disadvantageous, for at least three reasons. First, the abandonment of the Commission's monopoly over Article 81 (3) creates the significant possibility for greater incoherence, by giving fifteen national authorities the possibility to apply Article 81 (1) very broadly and/or grant exemptions without ensuring any common policy or analytic framework. Indeed, as we shall see from the discussion of Italian Banks, it appears that the only 'national' agreements to which the national authorities may be able to apply Article 81 will be agreements with primarily multi-state effects. Whereas this sort of risk existed before as concerns the application of Article 81 (1) by the national courts,8 cohabitation as envisaged in the White Paper proposals 8

Under Regulation 17/62 the risk concerning the application of Article 81 (1) by the national authorities could be eliminated by the expedient of notification, although the Commission had recently taken the (questionable) position that it would not intervene in cases in which notifications were filed to prevent the continuation of procedures that had been launched by national competition authorities. See paras. 57 and 62 of the Commission Notice on cooperation between national competition authorities and the Commission in handling cases within the scope of Articles 85 or 86 of the EC Treaty (OJ C 313/3 of 15 October 1997) below. The Commission, for example, decided not to initiate a procedure under Article 9 (2) of Regulation 17/62 with respect to an exclusive electricity supply agreement between RWE and the municipality of Nordhorn, which was notified to the Commission while infringement proceedings under Article 85 (1) (now Article 81 (1) EC) were pending before the German Bundeskartellamt (Federal Cartel Office.) Subsequently, the Federal Cartel Office prohibited the implementation of certain provisions of the agreement pursuant to Article 85(1) in conjunction with paragraph 47 of the GWB (Gesetz gegen Wettbewerbsbeschrankungen—law against restrictions on competition). RWE and the municipality of Nordhorn lodged an appeal against the decision before the Berlin Court of Appeal. The Court of Appeal made a reference to the Court of Justice for a preliminary ruling, asking, inter alia, whether Article 9 (3) of Regulation 17/62 is unlawful on the ground that Article 85 (1) and Article 85 (3) form an indivisible whole and the national authorities are not granted the power to apply Article 85 (3). (Case C-34/97 RWEandStadt Nordhorn v Bundeskartellamt, OJ C

462

Working Paper XI

will spread the problem to the national competition authorities and to Article 81 (3). This has already given birth to proposals, described in the White Paper, designed to enable the Commission to monitor and control the problems of inconsistency that are likely to result. Second, even if in the strictly legal sense there is no impact on the relationship between the application of national and Community competition laws as defined in Walt Wilhelm, the abandonment of the Commission's exclusive power to exempt magnifies the possibility for incoherence between the Commission and the national authorities. This effects is induced by placing the latter on equal footing with the former as concerns the application of Article 81, at least until such time as the Commission seeks to assert its primacy. This change may seriously weaken the hierarchical relationship created by the combination of Regulation 17/62 (which gave the Commission exclusive jurisdiction over Article 81 once a notification was filed) and Walt Wilhelm (which restricted the possibility for the application of national competition law in conflict with the Commission's application of Community law.)9 As a result, 94/7 of 22 Mar 1997.) RWE also brought before the Court of First Instance an action for failure to act, under Article 175 (now Article 232 EC,) against the Commission. (Case T-131/96 RWE v Commission OJ C 318/19 of 26 October 1996.) Eventually, the Court of Justice concluded that there was no need to adjudicate on the action. (Order of the Court of First Instance of 15 January 1999 in Case T-131/96 RWE\ Commission OJ C 86/18 of 27 March 1999; Removal from the register of Case C-34/97, OJ C 397/28 of 19 February 1998.) Paragraph 57 of the Commission Notice on cooperation between national competition authorities and the Commission in handling cases within the scope of Articles 85 or 86 of the EC Treaty reads: 'The Commission recognises, of course, that a firm requesting exemption is entitled to obtain from it a decision on the substance of its request (see point 38). However, if the Commission takes the view that such notification is chiefly aimed at suspending the national proceedings, given its exclusive powers to grant exemptions it considers itself justified in not examining it as a matter of priority.' Paragraph 62 reads: 'The Commission will not itself initiate proceedings in the same case before the proceedings pending before the national authority have been completed; in accordance with Article 9(3) of Regulation No 17, such action would have the effect of taking the matter out of the hands of the national authority. The Commission will do this only in quite exceptional circumstances—in a situation where, against all expectations, the national authority is liable to find that there has been no infringement of Articles 85 or 86 or of its national competition law, or where the national proceedings are unduly long drawn-out.' 9 Since the Member States have now adopted national competition laws patterned on Articles 81 and 82, there is, in theory at least, less significant substantive difference between national competition law and Community competition law, although some differences remain at the national level. Rather, the primary difference is institutional: national competition law cannot override or negate the results of Community law in areas where Community laws still applies and the Commission has acted to apply it. However, subject arguably to the obligation of loyal cooperation, Community law as applied by the national authorities has equal value to Community law as applied by the Commission until the national application is overturned by the Court of Justice, pursuant to Article 225 {ex Article 169) proceedings, on a case-by-case basis. The uncertainty that arises—and it is an important one—is whether the Member States'

2 - Coherence

463

cohabitation risks eliminating or at least potentially diluting the supremacy of Community competition law as applied by the Commission.10 Moreover, if the national competition authorities can apply both Articles 81(1) and 81 (3) (and can apply either in a manner that may differ from that applied by the Commission), the supremacy of Community law over national competition law, as denned in Walt Wilhelm, is greatly reduced in practice, even if in theory it continues to exist.1' Third, cohabitation creates the new possibility of conflict between the national authorities and the Commission in respect of the application of Article 81 (3) without providing a mechanism for resolving that conflict short of informal consultation and, should that fail, resorting to the Court of Justice. As a result, cohabitation opens up a new, hitherto non-existent field for potential and quite substantial legal uncertainty and procedural complication and delay. Whereas in the pre-reform world the Commission was the sole arbiter of Article 81 (3), its concerns over the consequences of cohabitation have led it to propose a set of new reporting procedures designed to permit it to monitor— and, if necessary intervene in—the national authorities' application of Articles 81(1) and 81 (3). From the perspective of the private consumers of competition law, it would arguably be preferable if the whole set of issues and the procedurally and substantively complicated situations to which they will give rise, were avoided in the first place. This could be done by limiting the national authorities to the enforcement of national, as opposed to Community, competition law, and thus restricting the scope for conflict to the Community's judicial system under

obligations under Article 10 EC (ex Article 5) apply in a situation where the Member States have amended secondary legislation to eliminate the Commission's monopoly over the application of Article 81. Arguably Articles 84 (ex Article 88) and 85 (ex Article 89) may not provide a basis for the Commission's reliance on the duty of loyal cooperation contained in Article 10, as they apply only to cases where secondary legislation had not yet been adopted—as opposed to cases in which it has been modified. Whatever secondary legislation is adopted will no doubt impact on the answer to this question and it is one of the points that has not been left unmentioned in the White Paper. 10 As noted above, cohabitation may not eliminate the hierarchical relationship between the Commission and the national competition authorities altogether to the extent that the Commission is able to rely on the Member States' Treaty obligation of loyal cooperation (Article 10 EC, ex Article 5) to limit the scope for the latter to act in a manner inconsistent with action taken or contemplated by the Commission. It is this obligation that presumably provides the legal basis for the Commission's ability to act pursuant to Article 225 (ex Article 169) as proposed in the White Paper. As noted above, it is not altogether clear whether this obligation will remain in force following the modification of Regulation 17/62. 1 ' Since the Member States have now adopted national laws largely or in many cases completely mirroring Articles 81 and 82, the substantive differences between national and Community competition law have been reduced or eliminated. What remains are procedural differences and the crispness of the delineation of the hierarchical relationship between the Commission and the Member States.

464

Working Paper XI

Article 234 (ex Article 177) without involving the Commission and Article 226 (ex Article 169) proceedings. Whilst it could be argued that the problems identified above are also inherent in increased private enforcement, and would be worsened by the judiciary's lack of experience in applying economic law, there is a strong response to this line of argument. First, enforcement of Article 81 in its entirety by the national courts in the context of private litigation is swifter than enforcement by the national competition authorities. This is because judicial enforcement condenses the application of the law into a single proceeding. Enforcement by the national authorities requires two proceedings—those of the national authorities and, thereafter, subsequent review by the courts. By any normal standard of time and efficiency, the duplication involved by enforcement through the national authorities does not appear advantageous. Moreover, the White Paper envisions another layer of possible judicial review—pursuant to Article 225 upon the initiative of the Commission—in those situations in which the Commission objects to the manner in which the national authorities have applied Article 81. Second, the argument frequently made in this connection—that the national authorities are better trained and thus better positioned than the courts to apply Article 81—is not convincing. For all their skills, training and experience, enforcement agencies are very dependent on applicants and third parties for their understanding of the industries with which they are dealing, and the quality of their work varies proportionately with the quality of the input from applicants and third parties.12 This input will also be available to national courts in the context of private enforcement actions, and, indeed, will be positively stimulated by the adversarial process that requires it. A competition authority with an agenda of its own can more readily choose to ignore the submissions of applicants and/or third parties than a court, which has no institutional mandate other than to resolve the case before it on the merits. Against this background, it is arguable that decision-making quality is not likely to suffer through private enforcement. Indeed, there is an argument that, given courts' impartiality and access to expert testimony, the quality of enforcement will be higher, assuming that the private bar and consultants are equal to their responsibilities, than in the case of enforcement by national authorities. This is not to say that private judicial enforcement would not benefit from the 12

If experience at the level of the Commission is a fair guide, private parties very often play the key role in educating a competition authority as to intricacies and economics of the cases before it. This can be confirmed by a comparison of the analytic quality of Commission decisions in which there have been third-party interveners and those in which there have not. It is submitted that such a review would show that the Commission's analysis has been far richer—and normally better focused on the relevant issues, whatever their resolution—when third parties have intervened and helped identify the issues.

2 - Coherence

465

creation of specialised competition courts, since that would clearly be the case.13

IV. The significance of Italian Banks for the White Paper Prior to Italian Banks, the Commission and the European Courts had taken a broad view of the applicability of Article 81 and had found that even agreements whose participants were from the same Member State could affect trade between Member States.14 In particular, agreements that have traditionally come within the ambit of Article 81 because of their effect on trade between Member States included: (i) agreements between undertakings established in different Member States;15 (ii) agreements between undertakings established in only one Member State, where those agreements concerned intra-Community imports or exports16 or activities in several Member States;17 and—the most relevant category for the present analysis—(iii) so-called 'national agreements', i.e. agreements covering the whole or the greater part of the national territory between parties domiciled in that territory. As concerns the latter, the Court has repeatedly held that 'national agreements' have 'by [their] very nature, the effect of reinforcing compartimentalisation of national markets, thereby holding up the economic interpenetration which the Treaty is intended to bring about.'18 13

In addition, the public availability of courts' rulings will increase the transparency of the competition enforcement system. 14 Joined Cases 56 and 58/64 Consten & Grundig v Commission [1966] ECR 299, p. 341 and Case 22/78 Hugin [1979] ECR 1869, p. 1898. 15 Commission Decision of 13 December 1985 Sole distribution agreements for whisky and gin (Case IV/30.570, O.J. No L369/19, 31 December 1985) and Commission Decision of 15 December 1986 BoussoislInterpane (Case 1V/31.302, O.J. No L50/30, 19 February 1987.) 16 Case 19/77 Miller v Commission [1977] ECR 131, p. 148; Case 172/80 Zuchner v Bayerische Vereinsbank [1981] ECR 2021, p. 2032, and Case C-279/87 Tipp-Ex v Commission [1990] ECR 1-261. Even 'any agreement whose object or effect is to restrict competition by fixing minimum prices for an intermediate product is capable of affecting intra-Community trade, even if there is no trade in that intermediate product between the Member States, where the product constitutes the raw material for another product marketed elsewhere in the Community' (Case 123/83 BNICv Clair [1985] ECR 391, para. 425.) 17 Commission Decision of 8 December 1983 Carbon Gas Technologie (Case IV/29.955, O.J. No L376/17, 31 December 1983) and Joined Cases 43 and 63/82 VBVB and VBBB [1984] ECR 19, p. 66. 18 Case 8/72 Vereniging van Cementhandelaren [1972] ECR 977, p. 991, paragraph 29; Case 73/74 Papiers Peints de Belgique, [1975] ECR 1491, p. 1514, paragraph 26; Case 42/84 Remia and Others v Commission [1985] ECR 2545, paragraph 22; Case 322/81 Michelin v Commission [1983] ECR 3461, p. 3522, Case 246/86 Belasco and Others v Commission [1989] ECR 2117, p. 2191, Case T-29/92 Vereniging van Samenwerkende Prijsregelende Organisaties in de Bouwnijverheid and others v Commission of the European Communities [1995] ECR 11-289.

466

Working Paper XI

Such a broad approach was facilitated by the wording of Article 81 EC which does not require proof of actual effect on trade in order for its provisions to become operative.19

1. Does Italian Banks preclude the application of Community law to 'national' agreements? In January 1999, the Court of Justice called the existing broad approach to the 'appreciable affect on trade between Member States' requirement into question. It ruled that agreements between the leading Italian banks that in essence impacted only on Italian consumers did not have an affect on trade between Member States and, therefore, that Community law was not applicable to them.20 This judgment was given on a reference for a preliminary ruling concerning the compatibility with Articles 81 and 82 of the EC Treaty of the Uniform Banking Conditions (UBC), which are imposed by the Italian Banking Association (ABI) on current account credit facilities and for provision of general guarantees. The referring judge had asked, inter alia, whether the UBC laid down by the ABI for its members in relation to contracts for the opening of current-account credit facilities, and the provision of general guarantee covering the facilities, are liable to affect trade between the Member States and have as their object and effect the prevention, restriction or distortion of competition within the common market. Before the reference, the UBC had been notified to the Commission. The latter, without expressing a view as to the existence of any restriction of competition and in what may have been an attempt at decentralisation on its part, decided to examine only 3 of the 26 agreements notified. In its view, Article 81 was not applicable to the remainder of the agreements, including those for the opening of current-account credit facilities and the provision of general guarantees, because, in the Commission's view, they did not appear capable of affecting, entirely or appreciably, trade between Member States. In this regards the Commission noted that: (i) the banking services in question were limited to national territory and involved economic activities which, under contractual provisions or by rea19

A s the Court of Justice has often emphasized, 'in prohibiting agreements which may affect trade between Member States and which have as their object or effect the restriction of competition Article 85 (1) of the Treaty (now Article 81 (1) EC) does not require proof that such agreements have in fact appreciably affected such trade, which would moreover be difficult in the majority of cases to establish for legal purposes, but merely requires that it be established that such agreements are capable of having that effect' (Case 19/77 Miller v. Commission [1978] E C R 131, paragraph 15.) See also Case 226/84 British Leylandv Commission [1986] ECR 3263, paragraph 20. 20 Joined Cases C-215/96 and C-216/96 Carlo Bagnasco and others v Banca Popolare diNovaras.ca., Cassa di Risparmio di Genova e Imperia SpA (Carige) [1999] ECR 1-135.

2 - Coherence

467

son of their very nature, must be carried on only within Italian territory or have a very limited influence on trade between Member States, and (ii) the participation of subsidiaries or branches of non-Italian financial establishments was limited. The Commission referred the other agreements to the Italian competition authority for the banking sector, the Banca d'ltalia, which scrutinised the agreements under Italian competition law. In Italian Banks, the Court of Justice accepted the Commission's findings— which, it is worth noting, had not been challenged before the Court—that: (i) the banking services in question involved economic activities that have a very limited impact on trade between Member States, and (ii) the participation of the subsidiaries or branches of non-Italian financial establishments was limited. The Court noted that, as the Commission had made clear in its reply to a question put to it by the Court, 'potential recourse to contracts for credit facilities and contracts for the provision of general guarantees by the main customers of foreign banks, that is to say large undertakings and foreign economic operators, is not great and, in any event, is not a factor of decisive importance in the choice made by foreign banks as to whether or not to establish themselves in Italy, insofar as contracts of the kind at issue in the main proceedings are only rarely used by customers of that kind'.21 In addition, the Court found that 'there [was] nothing else in the documents before the Court to justify the conclusion, with a sufficient degree of probability, that the reservations entertained by customers wishing to conclude a current-account credit facility contract regarding their choice of bank by reason of the existence of standard bank conditions relating to the provision of general guarantees is of such a kind as to have an appreciable effect on intraCommunity trade'.22 21

Italian Banks, cited above, paragraph 51. Italian Banks, cited above, paragraph 52. It should be n o t e d that Advocate General Colomer concluded that the U B C were incompatible with Article 81 o n the basis that the agreements in question did have an effect on trade between M e m b e r States. The Advocate General took the view that competition between banks from different Member States is crucial to ensure the proper functioning of the single banking market, with the result that all banking services, without exception, should therefore be subject to scrutiny under European competition rules. More specifically, the Advocate General strongly criticized in his opinion the Commission's practice of determining whether intra-Community trade is affected or not on the basis of the nature of the banking services and distinguishing between banking services which give rise to cross-border banking transactions (international payments, acceptance and cashing of foreign credit instruments, and foreign currency transactions), and banking services of an 'internal' nature (ATMs, safes and automatic bank drafts.) According to the Advocate General, such a distinction can no longer be made in light of (i) current technological developments ('home banking'), which greatly facilitate the provision of banking services outside the national market; (ii) the liberalisation of 22

468

Working Paper XI

The Court's judgment in Italian Banks leaves open the applicability of Article 81 to agreements between undertakings established in different Member States and to agreements between undertakings established in only one Member State, where those agreements concern intra-Community imports or exports, or activities in several Member States. However, the Court's judgment suggests that, possibly in respect of these, but clearly in respect of purely/primarily 'national' agreements, the 'appreciability' requirement under the 'effect on trade' test in Article 81 may be considerably less inclusive than was hitherto thought to be, although it is unclear where the line will be drawn. In any event, Italian Banks suggests that the option of limiting the application of Article 81 by the national authorities and courts to purely domestic agreements that primarily affect only consumers in a single Member State may no longer be open. This now appears to be one category of agreements that falls outside the scope of Article 81. Rather, it appears that, as a matter of law, the national authorities and courts will be required to apply only domestic competition law to such purely (or primarily) 'national' agreements that do not affect trade between Member States. The result is that no conflict will be possible between the Commission and the national authorities as concerns the application of either Community or national competition law to such agreements.23 To the extent that the Member States have adopted national laws based on Article 81, the application of national law rather than Article 81 may not have any significant substantive consequences. Moreover, if the Member States' duties of loyal co-operation under Article 10 EC (ex Article 5) are in effect an embodiment of the ruling in Walt Wilhelm, there may not be significant institutional differences either. Though it should be borne in mind that Walt Wilhelm does not apply in areas where only national law applies. In this sense, Italian Banks capital movements in the Community and worldwide, favored by the globalisation of the world economy; and (iii) the completion of the single banking market through implementation of the Second Banking Directive and the impending introduction of the single currency. (Opinion of Advocate General Ruiz-Jarabo Colomer in Joined Cases C-215/96 and C-216/96 Carlo Bagnasco and others v Banca Popolare di Novara s.c.a., Cassa di Risparmio di Genova e Imperia SpA (Carige) [1999] ECR 1-135.) 23 Conflict between national and Community law will be precluded since Walt Wilhelm will not apply given the non-applicability of Community law. It should be noted that the possibility for such conflict might remain as concerns transnational 'networks' of such national agreements. However, the full extent to which Italian Banks will reduce the scope for the application of Article 81 remains to be seen. In a related case, the Commission has, however, already followed the Italian Banks case in its decision of September 1999 regarding the Dutch Banking Association's rules on acceptance of giro systems for domestic business payments, even though this case dealt with a policy issue—the status of inter-bank payment systems—which was of considerable policy importance to the Commission and had provoked its intense study. Commission Decision of 8 September 1999 Nederlandse Vereniging van Banken (1991 GSA agreement) et al. (Cases IV/34.010, IV/33.793, IV/34.234 and IV/34.888, O.J. No L 271/28, 21 October 1999.)

2 - Coherence

469

can be said to expand the exclusive competence of the national competition authorities.

2. Is Italian Banks an invitation to the national authorities to apply Article 81 to agreements that have effects in more than one Member State? Perversely, Italian Banks leaves open the possibility, subject to territorial limitations in the exercise of their authority, that the national authorities may seek to apply Community law—and may so do inconsistently—to arrangements that have effects in more than one Member State, including presumably networks of national agreements. This is, of course, precisely the area where cohabitation is likely to create the greatest problems, to the extent that it may produce inconsistent approaches to the application of Article 81 to the same arrangement by different national authorities. Those who have been involved in transatlantic airline alliances, in respect of which something akin to cohabitation also applies because of the lacuna in Regulation 3975/87 (which does not apply to agreements affecting air transport between the Community and third countries),24 can attest to the potential for conflict, delay and uncertainty that can result from this sort of situation. 3. Is there a conflict between cohabitation envisioned in the White

Paper and Italian Banks! Italian Banks and, indeed, the position adopted by the Commission in its approach to the agreements in that case, can be best understood as one 24

See Commission notice concerning the alliance between British Airways and American Airlines, O.J. No C239/10, 30 July 1998; Commission notice concerning the alliance agreement between British Airways and American Airlines, O.J. No C289/4, 2 October 1996; Commission notice concerning the alliance between Lufthansa, SAS and United Airlines O.J. No C239/5, 30 July 1998; Commission notice concerning the alliance agreements between KLM Royal Dutch Airlines and Northwest Airlines, Inc., O.J. No Cl 17/8, 15 April 1997; Commission notice concerning the alliance agreements between Delta, Sabena, Swissair and Austrian Airlines, O.J. No C289/6,2 October 1996; Commission notice concerning the Lufthansa, United Airlines and Scandinavian Airlines System alliances, OJ C 289/8,2 October 1996. Legally the Commission does not have primary jurisdiction over transatlantic alliances since they are not covered by Regulation 3975/87. As a result, the Commission has had to rely on the Member States' duty of loyal cooperation, in tandem with Articles 84 and 85 EC (ex 88 and 89) once it has initiated proceedings, to seek to limit the scope for substantive conflict with the Member States application of Article 81 to these agreements. Arguably the Commission's Treaty-based ability to do so will remain in place following the White Paper reforms. However, consideration needs to be given to the question of whether secondary legislation is required to ensure this result.

470

Working Paper XI

possible form of decentralisation—i.e., decentralisation without cohabitation. As such, the approach in Italian Banks cannot be said to be inconsistent with the goals of the White Paper, although it achieves them by another means. That is, Italian Banks constitutes a parallel means of achieving decentralisation—by reducing the scope for the application of Community law by both the Commission and the national competition authorities. Thus, both Italian Banks and the White Paper result from the same recognition that the Commission is not able to ensure the uniform application of Community law in all areas and that it is therefore necessary to delegate more cases to the national authorities and courts. Nevertheless, the White Paper does not appear to have considered the implications of the position adopted by the Commission or the Court in Italian Banks. The latter has not been the subject of extensive commentary in the various submissions and articles which have been prepared in response to the White Paper. The consequences of Italian Banks appear paradoxical. First, with the exception of hard core cartels—which regulators would no doubt like to subject to as much administrative and legal inconvenience as possible (notwithstanding potential problems concerning double jeopardy)25—it would make sense to give the national courts and the national competition authorities Article 81 jurisdiction over only those agreements that impact on consumers within their territorial jurisdiction. This would impose a territorial limit on the possible conflicting national approaches to Community law. However, Italian Banks limits the ability of the national authorities to apply Article 81 by holding that neither national authorities nor the courts may apply Article 81 to primarily national agreements. By reducing the jurisdictional scope of Article 81, Italian Banks limits the impact of cohabitation as envisioned in the White Paper by excluding primarily national agreements from the scope of Article 81. The key question is whether in the process it will do mischief by in effect leaving the national authorities, subject to territorial restraints on their exercise of jurisdiction under Article 81, free rein to apply Article 81 to agreements that primarily affect trade between Member States and may therefore also be subject to the competing attentions of other Member State authorities. Second, as a consequence of Italian Banks, in Member States that have adopted procedural rules similar to those in Regulation 17/62, primarily 25

The European Court of Justice, despite holding that the same agreement may, in principle, be the subject of parallel proceedings, one set before the Community authorities under Article 81 EC and the other before the national authorities under domestic law, has made it clear that, in principle, the national competition authorities may take action regarding situations likely to be the subject of a Commission decision. However, if the ultimate aim of the Treaty is to be respected, parallel application of the national system can be allowed only in so far as it does not prejudice the uniform application throughout the common market of the Community rules on cartels and of the full effect of the measures adopted in implementation of those rules. {Walt Wilhelm, cited above, paragraph 4.)

2 — Coherence

471

domestic agreements will be subject to a prior notification requirement to receive an exemption under national law. To the extent that the national authorities have prior notification rules, they may find that they face a significant number of notifications that may be increased as a result of Italian Banks. This, of course, raises the question whether the national authorities should not, for the same reasons that have led the Commission to do so at the Community level, seek to eliminate their rules requiring prior notification of restrictive agreements.

III. Is the uniform application of EC competition law necessary or desirable? 1. Background—the coherence measures proposed by the White Paper The White Paper proposes a series of measures to ensure coherence and uniformity in the decentralised application of Article 81: (i) Article 226 EC (ex Article 169) actions against the Member States (presumably primarily, but arguably not exclusively, the national competition authorities); (ii) Article 234 EC (ex Article 177) references to the ECJ from the national courts; (iii) A system of file swapping between the Member States and between the Member States and the Commission; (iv) Transparency measures so that the Commission is kept informed, in advance, of what the national authorities and national courts are up to; and (v) Amicus curiae interventions by the Commission in Article 234 EC (ex Article 177) court proceedings. These measures all seem reasonable on their face. However, the measures aimed at controlling the national authorities, in particular the possibility of Article 226 actions brought by the Commission against the latter, would seriously complicate the lives of companies subject to the attention of both the national authorities and the Commission, in a manner with no precedent under the ancien regime.26 It is doubtful whether there are any countervailing benefits to compensate for the potential inconvenience. 26

References for a preliminary ruling can also be criticized on the grounds that they require too much time, almost two years at the Court of Justice's level, plus, possibly, several years at the national courts' level, given that only courts of last instance are under a legal obligation to make a reference to the Court of Justice. The reform that has been recently proposed to modify the present preliminary ruling mechanism may not be sufficient to address this concern.

472

Working Paper XI

2. The need for resolution of different court interpretations It is probably heretical to suggest that there could be some value in permitting different approaches in different jurisdictions, at least for a limited period required to test them. However, in certain limited cases there may be something to be said for such an approach, to the extent that the existence of different rules in different national markets with different specificities can serve either in the short term as a benchmarking device, or can be justified in the long term by different local conditions. An approach tolerating a limited degree of diversity would arguably be most defensible in the area of distribution, where different national specificities are likely to be decisive. Though to the extent that the approach in Italian Banks governs, this may be an area where there is no scope for the application of Community law at all. This situation has already occurred under national law in some Member States to the extent that different national rules exist respecting certain (normally distribution) practices. For example, there are some Member States that do not permit a non-discrimination rule concerning credit cards. As a result in, for example, the UK and Sweden, shop owners and others are allowed to charge a commission or higher price for credit card transactions. The beneficial (or pernicious) effects of such a rule can be tested by benchmarking against Member States in which the rule on non-discrimination is legal and can be enforced.27 The US experience: certiorari. In the United States, the Supreme Court has the power to take jurisdiction in order to resolve splits between the circuits on antitrust rules and often does so, but, many times, it does not. The latter can lead to conflicts as between different Federal circuit courts. However, according to the results of a study published in 1998 by Arthur D. Hellman, of the University of Pittsburgh School of Law, these conflicts rarely lead to forum shopping or cause harm to multi-circuit actors.28 In his study, Professor Hellman analysed 201 conflicts that were denied review by the Supreme Court in the 1988,1989, and 1990 Terms, and concluded that 62 had been put to rest by legislative or judicial action. Another 63 conflicts had died a natural death: either the underlying issue has disappeared or there was no longer any evidence of intercircuit disagreement. Among the still existing conflicts, there were only 50 that manifested some potential for affecting outcomes. However, as concerns these cases, Professor Hellman concluded that lawyers rarely file suit in one circuit rather than another in order to take advantage of favourable circuit law. The disincentives tofilingsuit in any place 27 It is possible that interdiction of the non discrimination rule may work well in some Member States—e.g. the UK—where there is a strong consumer lobby and less well in others. 28 Hellman, Arthur D. (1998) 'Light on a Darkling Plain: Intercircuit Conflicts in the Perspective of Time and Experience,' Supreme Court Review 1998, 247.

2 - Coherence

473

but the client's home forum (and that of counsel) are numerous and strong for five reasons: (i) The point on which the circuits differed would only rarely determine whether a lawsuit is winnable; (ii) Lawyers generally view federal decisions as constituting a single body of law; (iii) Experienced lawyers give less weight to individual precedents in antitrust cases; (iv) Federal appellate decisions are only one source of the legal rules that lawyers consider in litigation and counselling (for antitrust lawyers, Department of Justice and Federal Trade Commission guidelines may set the bounds of permissible activity); and (v) Legal rules are only one of the elements that lawyers consider in counselling and litigation. Thus, according to Professor Hellman, 'unresolved intercircuit conflicts do not constitute a problem of serious magnitude in the federal judicial system.'29

V. Conclusions The White Paper has a something of an air of political compromise. Arguably, if permitted by the Member States to do so, the Commission could have accomplished all its aims (and avoided the potential for conflict between itself and the national authorities) by a combination of one or more of the following: (i) Relaxing its approach to Article 81 (1); (ii) Abolishing the system of prior notification; (iii) Deputising the Member States to assist it in cartel or other investigations or complaints better handled at the Member State level; (iv) Leaving to (specialised) national courts the litigation over the enforceability of contracts and private enforcement actions. However, the price exacted for the co-operation of the Member States in amending Regulation 17/62 appears to be a form of cohabitation requiring the Commission to share its Article 81 (3) powers with the Member States. Since cohabitation is arguably not necessary to bring about the desired reforms, it can be regarded as an unfortunate source of potential, but unnecessary, conflict. In particular, giving the Member States the competing ability to apply Article 81 in its entirety and/or to exempt agreements pursuant to Article 81 (3) does not appear to be in and of itself desirable. Reliance on private enforcement through the judicial process poses fewer problems both in terms of procedural delay and coherence and is arguably a more efficient means to the same end. 29

Hellman, cit., at 249.

474

Working Paper XI

Second, cohabitation raises the question of whether there is (or will be) a legal basis for the Commission to assert its supremacy over the application of Article 81. In the absence of secondary legislation conferring such powers on the Commission, it may be argued that the Member States are no longer subject to a duty of loyal co-operation, pursuant to Article 10 EC. Inter alia Article 10 arguably imposes obligations when read together with Article 84 and 85. The latter may not be applicable where the Member States have modified secondary legislation to eliminate the Commission's monopoly over Article 81 (3). Third, there is now a risk that, as a result of Italian Banks, the White Paper may be transformed into an invitation for the Member States to begin applying Article 81 to agreements and practices that primarily affect competitors and consumers in other Member States. It can be argued that the some possibility for conflict exists through the application of national competition law, which (by and large, if not perfectly) now substantively resembles Community law, and which in areas of exclusive national competence—an area that has now been expanded by the Italian Banks judgment—is not subject to the Commission's application of Article 81 under the Walt Wilhelm doctrine. However, the jurisdictional scope of the national authorities to apply national law would clearly be limited—to agreement affecting only consumers in their own territory—and outside that area, the relationship between Community and national law is, given Walt Wilhelm, subject to a hierarchical ordering, as a matter of Community law—designed to eliminate conflict. As noted above, it is not clear whether such clear legal principles will apply in the post White Paper world absent express secondary legislation. The potential for conflict between the application of Community law at the national and Commission levels is more easily supported in the case of private party litigation—since such cases are resolved in the courts and produce rulings that are limited to the case at hand. Conflict is, however, more likely where one or more a regulatory bodies, with potentially varying policy agendas, are involved. It is difficult to see why this aspect of the White Paper proposal—and the countermeasures proposed by the Commission in an attempt to maintain or reassert its authority—are desirable or necessary. Decentralisation could be achieved in an equally satisfactory manner by combining private enforcement of Community law in the courts with the enforcement by the Member States of national competition laws patterned on Article 81 at the purely local level.

PANEL THREE COURTS AND JUDGES

1 PANEL DISCUSSION

PARTICIPANTS

Carl Baudenbacher Simon Bishop Jochen Burrichter John D. Cooke David Edward Claus-Dieter Ehlermann Eleanor Fox Dan Goyder Jochem Groning Frederic Jenny

Santiago Martinez Lage Emil Paulis Alberto Pera James Rill Alexander Schaub Mario Siragusa Anne Spiritus-Dassesse Sarah S. Vance Diane P. Wood

3 - Courts and Judges

All

Panel Three: Courts and Judges • CLAUS-DIETER EHLERMANN—During this session we will address the problems linked to the judiciary—problems that we have already spoke about quite a bit during the previous sessions. I am mindful of the fact that we left the arbitration and civil procedure issues aside, though I suppose they will come forth during this debate. Our first speaker today is Professor Carl Baudenbacher, who is here today in his capacity of judge at the EFTA Court. • CARL BAUDENBACHER—As a judge and a law professor I deal with competition law issues only from time to time, among my many other assignments. Therefore, in the presence of so many distinguished competition specialists, I feel like a lover among long-married people: sometimes enthusiastic, but every now and then with a bad conscience. The competition rules of the EEA Agreement, in particular Articles 53 and 54, mirror Articles 81 and 82 of the EC Treaty. These provisions are enforced by two agencies: the Commission of the European Communities and the EFTA Surveillance Authority. As far as the division of enforcement competencies is concerned, there is a one-stop shop that, as Jacques Bourgeois mentioned, would also be a good idea for the future enforcement system in the EU. 'One-stop shop' means that either the EFTA Surveillance Authority or the Commission handles cases affecting trade between the EFTA States and the EU Member States, and the EFTA remembers gratefully that Claus-Dieter Ehlermann has been instrumental in setting up this system. EFTA national competition authorities do not yet have the right to apply Articles 53 and 54 EEA, but the national courts in the EFTA States do apply those provisions. I should mention here that there is a dualistic tradition in international law in the Nordic Countries. However, in this case, this tradition does not create any problems because the EEA agreement has been incorporated into the national law. By the way, the EFTA Court has ruled in a case involving an Icelandic woman, with the beautiful name of Erla Maria Sveinbjdrnsdottir that the EEA Agreement has itself created a new legal order, to be distinguished from the legal orders of the contracting parties. With respect to the impact of the reform proposed in the White Paper on the European Economic Area: the EFTA States basically support the Commission's proposal. It is also clear that, from the point of view of homogeneity in enforcement, the reform should be extended to cover the European Economic Area. The EFTA States are aware that this implies that they will have to enable their national competition authorities to apply the EEA competition rules. The EFTA States agree that in a decentralised competition law enforcement system, Commission block exemption regulations, notices and guidelines will continue to be of importance and, furthermore, they ask the Commission to consider the need for harmonising national procedural laws. ESA has even pointed out that clear rules on the allocation of cases are needed.

478

Panel Discussion

With respect to the question whether national courts will be able to do the job, I will limit myself to four remarks. The first is that antitrust is not the only legal area posing complex problems. The second is that the arguments put forward against entrusting national courts with the application of provisions that presuppose a certain economic knowledge are not new; I remind you of the law on unfair competition, just as an example. The third is that, with all due respect, the competition law community (or certain parts of it) is in danger of doing too much navel gazing. This might have to do with the fact that there is a lot of money involved in this area of law enforcement, but also with the fact that Europe competition law is some sort of 'secret science' as it is not taught at most of our universities. My fourth remark is that, when reading all the objections that were made against the Commission's reform proposal, one might get the impression that the criticism is based on a tacit assumption that the protagonists that handled EC competition cases so far are almost impeccable. I would just like to remind you a famous statement by Justice Robert H. Jackson, namely that reversal by a higher court is not proof that justice is thereby served better, since there is no doubt that if a 'Super' Supreme Court also existed, a substantial portion of the Supreme Court decisions reversing State court decisions would also be reversed in turn. Then there is the famous sentence: 'We are not final because we are infallible, but we are infallible only because we are final'. I recall in this context that, for instance, the Bundesfinanzhof (the Supreme Court dealing with financial matters in Germany) refused, for a certain time, to request preliminary rulings from the ECJ because the judges had the feeling that the ECJ did not handle such cases in an appropriate way. So, everything is relative. One will of course agree that there are problems with the enforcement of EC competition rules in national courts, given their shortage of resources, the actual knowledge of competition matters, education and experience, but all these problems seem to be manageable. Nevertheless, we should not hold too many illusions: we will not be able to have it both ways, and a certain amount of forum shopping will be unavoidable, since national judges will compete with each other. As far as the role of the European Courts is concerned, I want to make just one comment. The White Paper essentially relies on the preliminary ruling procedure. Yet this cannot be compared to direct actions, since the facts of the case basically remain established by the national court. At this point, the focus has to be on the ongoing reform of the judicial architecture of the European Union. My feeling is that the White Paper, and the discussion around it, have not taken the importance of this reform into account to a sufficient extent. You could say that, in this respect, the White Paper is trying to shoot a moving target. Regarding the goal of reaching intensified control of trespassing agreements, the EFTA States suggest—and this is an interesting proposal—that the provisions in US antitrust law concerning information-gathering powers of the competition authorities should be further explored.

3 - Courts and Judges

479

From the viewpoint of EFTA States, the hot potato in the reform proposal is the following: at present, the ESA has certain co-operation rights with respect to cases over which the Commission has jurisdiction under the EEA Agreement, and fears that, with decentralisation, the national competition authorities of the Member States will not respect those co-operation rights. This problem could be solved by obliging the national competition authorities of the Member States to co-operate on the same terms as under the present arrangement with the Commission, but EFTA's preferred solution would be that the Commission take the relevant cases out of the jurisdiction of national competition authorities. • JOCHEN BURRICHTER—I will try to concentrate on three items. First, I would like to say a few words about the role of private enforcement under the new system. Second, I shall go into some of the details concerning the application of Article 81 (3) under the new system. Finally, I will come to the issue of forum shopping. It has been said that one should not rely too much on private enforcement in the new system, and I think that is true. In my understanding, the new system should not place private enforcement at the forefront, but private enforcement should nevertheless be enhanced and made easier. This is one of the pillars of the future system, though of course not the only one, as enforcement by the national competition authorities is still at the forefront. I think, however, that the Commission's reform ideas will in effect be able to enhance private enforcement, which is nowadays becoming more and more important. In relation to this, I share the doubts that Professor Immenga expressed on behalf of Professor Mestmacker, namely that at present it is quite difficult (at least for a third party) to enforce antitrust law against the members of a cartel. But I see in practice that there is a move in the business community towards changing this state of affairs. For example, in the past in Germany nobody would have thought of suing against cartels, but more recently, after the vitamin cartel case there are many more companies actually considering doing so (third parties, that is, customers). I also know that interested parties urge the Commission to declare its objections to granting an exemption more quickly, and to give them access to the file. This might be an idea for the Commission to consider; namely that in cartel cases the files of the Commission should be opened to third parties interested in suing the cartel members before national courts. I will move on to some of the aspects concerning the direct application of Article 81 (3) in national courts. I think there is no doubt that the courts are able to apply Article 81 (3), and we have already covered many of the issues involved here. I therefore only want to mention something related to the policy aspect: in fact, as I said yesterday, general EU policy aspects should have no place in the application of 81 (3), but the courts should of course also take into consideration the general provisions of the Treaty when interpreting the

480

Panel Discussion

conditions for granting an exemption. I think that is something that a Court could do in civil law proceedings. We have already discussed about the discretion previously granted by European Courts to the Commission in the application of 81 (3). I want to take this discussion further by referring to one of the objections as to the applicability of 81 (3) by national courts, namely that adversarial procedures cannot produce fair results and 'find the truth'. This is a surprising argument, if taking into account that legislators in most Member States leave other important legal matters—for example, corporate law, or labor law which also implies public interest considerations—to be decided in adversarial procedures. So, I think that this argument is not justified. If we look at the cases decided by the courts, these are more and more disputes between the parties to an agreement about its validity and, certainly, also claims by third parties about alleged infringements of competition rules. In the latter kind of cases, it can be expected that the plaintiffs will submit facts to back their case, or that fact-finding should not be more difficult than in other legal areas, or more difficult than when the cases are dealt with by a competition authority. In adversarial procedures, judges have a variety of possibilities to explore the facts, and to point to legal aspects so that the parties have the opportunity to put forward the right arguments. The Commission can also strengthen private enforcement by assisting the third parties, for example, in finding the facts to back their claims against competition law infringements, including by opening thefilesof a fine case, for instance. I should mention that one of the possibilities for correcting wrong decisions of the courts could be that of allowing the competition authorities to issue administrative decisions. I mentioned yesterday that I would even advocate the issuing of Commission positive decision in cases where a national court wrongly came to the conclusion that Article 81 (3) was not applicable. This was to clarify yesterday's controversy with Mr Boge. I wanted to remark that this is actually not extraordinary under German law, because even the Federal Cartel Office in Germany comes to the conclusion that, for example, the rules prohibiting of cartels are not applicable. In such cases, even if the civil court's decision is definitive, the Federal Cartel Office has the possibility of issuing a decision prohibiting the cartel if it comes to a different conclusion to the court on the case. Another argument invoked against the applicability of Article 81 (3) in national courts is the fact that, under the current system, exemptions can only be granted for a limited period of time, and this is not compatible with the nature of civil court decisions. Here I think one has to consider that, under the new enforcement system, exemptions will not be granted in the operative part of the court decision, but rather the possibility of an exemption is an issue that the courts have to look at implicitly when granting an award for a claim or refusing to grant it. So, issuing a decision with effects limited to a certain period of time should not be a problem. However, in clear cases where the court comes to the conclusion that an exemption can only be granted for a certain period of time, and the claim goes beyond that period, the Court can refuse to grant, for

3 - Courts and Judges

481

example, the statement of validity beyond that period of time, a solution that makes it possible to limit the exemption in time. I would also make some short remarks about arbitration procedures, because that' is another aspect that arose in our discussion. I think it is important to understand that arbitration courts must observe competition rules, otherwise they risk that their decision be challenged by the loosing party during the enforcement proceedings. Then the risk arises that the arbitration decision cannot be enforced, and that the arbitrators have to return their fees. I would think that under the new system it would be sufficient for arbitration courts to be enabled to ask the Commission for the same information that Commission would have to provide to the ordinary courts in case of doubt, but I do not think it is necessary to grant the arbitration courts the possibility of requesting preliminary rulings to the European Court of Justice under Article 234 EC Treaty. There is no time left to cover the issue of forum shopping, but my feeling is that this fear is exaggerated and in most cases there is no danger of forum shopping—but I can elaborate on that later during the discussion.

• JOHN COOKE—I believe that what the Commission proposes to do in the White Paper is both legally feasible and practically possible. I do not share the doubts about the compatibility of the reform proposal with the EC Treaty that were mentioned yesterday. I also think it is possible to make the reform work, provided that the new Regulation—replacing Regulation 17/62—will very clearly define what the role of the Commission will be in the new enforcement system, both in relation to the national competition authorities and in relation to national courts. If you consider in particular that in the foreseeable future we are going to have a Union of possibly twenty or twenty-five Member States, with an equivalent number of national competition authorities, it is absolutely crucial that the relationship between the Commission and both the national courts and the national competition authorities is clearly and simply defined, so that everybody understands exactly who is doing what, who is entitled to do what, and when they are entitled to do it. The White Paper makes me suspect a tendency on the part of the Commission to have it both ways. If the Commission believes in decentralisation, then let's have decentralization, and not some obscure compromise in which there are exceptional provisions to override, to withdraw, to hold back, to second-guess what happens at national level. Nor do I share the doubts that have been expressed about the ability of national judges to apply Article 81 (3). It is the vanity of all experts to believe that their own area of specialisation is so arcane that it cannot be put into the hands of people outside that particular expertise. When you think about it, national judges are applying other legal provisions that are every bit as delicate, complex and framed within public policy. I cannot speak for all the judicial systems in the EU, but if you take the common law jurisdictions, a judge sitting in

482

Panel Discussion

a commercial court is often asked to authorise the reduction of the capital of a public company, where very arcane and sophisticated accounting concepts come into play and both the interest of the public and public policy in the area have to be taken into account. For many years in Ireland and the UK, in the field of intellectual property, there used to be applications to extend the duration of patents where the judge was asked not only to understand a complex patented scientific formula, but also to look at the remuneration of the patentee over a certain period and decide whether he had been adequately remunerated. In addition, you had the administration arguing for public interest, and the current policy of the patent office in thatfield.So, there is nothing particularly special from an ordinary judge's point of view about competition law when compared with other legal areas. It seems to me that the real question is how do you ensure that the national judges are adequately informed about Community competition policy, and how do you ensure that the Commission is adequately and promptly informed of important decisions taken or pending to be taken in national courts, particularly in the area of application of Article 81 (3). The White Paper talks about the Commission intervening in national court proceedings as amicus curiae. I think this wrong: the role of the Commission at national level should be as close to nil as possible. When you think about it, if you are going to have multiple enforcers, then the Commission is already represented through the national competition authorities in the Member States. I therefore think the answer to these questions is relatively simple: the new Regulation should require each Member State to ensure that its rules of court procedure contain a provision whereby every party to litigation first raising a claim or a defense based on Articles 81, 82, or 81 (3) is obliged to serve notice of that claim, or defense—whatever it may be—on the national competition authority. The national competition authority should then have a limited period of time in which to decide, in consultation or after consultation with the European Commission, whether it wishes to intervene in that case. This is why I think that the concept of amicus curiae in this area is inappropriate, because if the issue that is raised is sufficiently important to warrant the intervention of the Commission through the national competition authority, then it seems clear that the Commission and the national competition authority will have taken a view on the point at issue. In those circumstances, I believe it is more appropriate and more helpful for the national judge if the Commission or the national competition authorities intervene in support of one side or another. The vague idea of amicus curiae is inappropriate: I think it is wrong for the Commission to intervene, and even on constitutional grounds it is a mistake for the Commission to risk compromising its privileged position under the Protocol on the Privileges and Immunities of the European Communities by becoming involved in litigation at the national level. All this is unnecessary if the network of national competition authorities acts on the Commission's behalf and in consultation with it. The national competition authorities would, in any event, frequently be involved in litigation in their own courts in relation

3 - Courts and Judges

483

to the application of national competition law, and thus they will know the judges and the rules of procedure, they will have their own legal teams already in place. It would therefore be a much more efficient employment of resources to use the national competition authorities for any such intervention on behalf of the Commission. And, in any event, if we already have a problem of resources in Brussels in the Competition Directorate, I cannot see how these resources are going to be spared for an effective and prompt intervention at national level by the Commission. I wanted to make onefinalpoint, which I mentioned in my paper, and here I think it is worth reading a passage from the White Paper, because I am hoping to learn from the participants at this conference what it means. It is paragraph 102 (2) of the White Paper, which reads: 'When a national authority has adopted a positive decision which is either no longer open to appeal or which has been confirmed on appeal, or a court has delivered a positive judgment (for example rejection of a complaint on the ground that a restrictive practice satisfies the tests of Article 81 (3) which is either no longer open to appeal or has been confirmed on appeal) the Commission can always intervene to prohibit the agreement, subject only to the principle of res judicata that applies to the dispute between the parties themselves, which has been decided once and for all by the national court.' That does not seem to me to make sense. In fact, I think it is dangerous nonsense. If the matter has been decided once and for all by a national judge, how can the Commission overturn that decision? More importantly, if the principle of res judicata applies, it seems that the very parties whom the Commission believes have infringed the Community law will get away with it, whereas third parties are apparently going to be bound by the subsequent Commission reversal of the original outcome. This also seems to be dangerous from the point of view of the rule of law: I do not think that national judges will take kindly to a situation in which they are left in the position that, however final their decision in a case, an administrative authority may reverse the effect of what they have decided. And all this is quite apart from the fact that, once national courts are applying Community law, Article 6 of the European Convention on Human Rights comes into play. • DAVID EDWARD—First of all, I am happy to say that there is no disagreement between the two European Courts. I entirely agree with John Cooke, and would repeat what Mr Kist said yesterday: do not jeopardise this enterprise by inadequate preparation and do not underestimate the seriousness of the problems that were pointed out yesterday and will be raised today. It is absolutely futile to proceed with the reform on the assumption that it is just the selfinterested lawyers who create problems. If you do not face up to these problems now, they will come back to haunt you later. Second, I endorse entirely what John Cooke said about using the potential of the national competition authorities. The great 'secret' of the Community legal system, is that the Treaty uses national bodies as enforcers of Community law,

484

Panel Discussion

and there is absolutely no reason why national competition authorities should not be used as enforcers of Community competition law. This is absolutely in line with the Community principle of subsidiarity, which was in the Treaty from the very beginning: subsidiarity was not invented in Maastricht, it was invented back in 1957 with the use of the national courts as primary enforcers of Community law. Third, taking further a comment that John Cooke made before, it is perfectly true that the national courts will be dealing with disputes between party A and party B about their contract, but one of the most pernicious elements in competition law is the 'network' of identical contracts between party A and parties B, C, D, and E. From a legal point of view, each is a separate contract, and each is justiciable separately. From the economic point of view, competition is distorted by the network of agreements as a whole. So, do not say we do not need to bother about the Brussels Convention because the dispute is just between party A and party B. If the contract between parties A and B was subject to a court decision in one country, what is the effect of that court's decision in another country and with respect to a contract concluded on identical terms between party A and party C? And in particular, what happens when parallel proceedings between one party and several different parties are started in the courts of different Member States, but all concern the same terms of identical contracts in the 'network'? The same question arises in arbitration: arbitration proceedings may involve two parties, but the effect of the arbitral decision might have a knock-on effect on identical contracts with other parties. Fourth, I agree with John Cooke that Article 81 (3) is justiciable. Analogous balancing problems occur in other areas of Community law as well, and the need to balance different Treaty provisions is ever increasing. Under the Treaty of Rome, the main provisions to be considered were those on the free circulation of goods and competition, but social developments and successive Treaty amendments have brought about several new legal and policy considerations to be balanced against each other (fundamental rights, trade marks and designations of origin, environment, consumer protection, biodiversity, drugs and terrorism, and so on). In the old days, the judge was operating a sort of on-off switch: he looked for the norm to apply, and he applied it. Nowadays the judge is operating a synthesizer, that has a number of pulses coming through, so he must assess the strength of these different pulses or considerations, adjust them, and find a solution whereby several considerations come through together, and not simply one norm of positive law. This exercise is new for many judges, and some are afraid of it, particularly in the national courts. Some judges simply resist doing it, as you can see from the evidence of Mr Justice Ferry and Mr Justice Laddy before the House of Lords. They just do not like the idea that this is what they have to do. So, do not underestimate the potential for judicial resistance to what you are expecting them to do under the new enforcement system. Having said that, unlike political institutions, the courts are not in a position to choose what they want to do. In Europe at least, they cannot just say 'This

3 - Courts and Judges

485

case is unattractive, or uninteresting, and we are not going to deal with it'. The courts have to take the case and, what is more, they also have to decide it. The consequence of this is that if you devolve enforcement powers to national courts, you will get decisions, and many of them might be unattractive from your point of view. They will be decisions from overloaded national judges, and not necessarily those in London, Brussels, or Paris, but also those in Rochester, Yorkshire, Santiago de Compostela, and any town in Finland. These will be the judges who decide Community competition law issues and, please, distinguish between specialised courts dealing with the judicial review of administrative decisions, and ordinary courts that do not deal with competition law aspects only, but which mix competition law aspects with other civil law considerations in the same case. So, the real question is not whether the judges can do it, but what will be the quality of the decisions that you will get from them. As I see it, this quality will depend on the manpower of the national courts, the training of the judges, and the availability of 'back-up' (for example, libraries, access to information, and the help that the judge can get generally). In many countries, the judge depends extensively on the lawyers to present the court with the evidence. This means that the Bar must be brought into this training and information operation as well. But above all, as John Cooke said, there must be clarity as to what is expected of the national judges: clarity as to the burden of proof, clarity as to what has to be taken into account, and what does not. Do not underestimate the magnitude of the need for training and guidance. Equally importantly, do not underestimate the burden the reform will place on the European Court of Justice. At the moment, both the Court of First Instance and the European Court of Justice are overloaded. The European Court of Justice in particular is seriously overloaded with requests for preliminary rulings, so do not underestimate the difficulties with using the Community Courts as a modality to ensure coherence in the enforcement of Community competition law, unless substantial additional resources are provided to them. My next comment goes back to another point made by John Cooke: please respect the separation of powers. The Commission cannot have it both ways: if you want the judges to do it, the judges will do it, but the Commission cannot start interfering on an administrative basis with what the judges will do. With respect to procedural issues: do not be misled by parallels with the US system. The US is procedurally homogeneous, Europe is not. We must take account of the fact that procedures in the national courts differ considerably from one Member State to another, and—as someone said yesterday—it is often the procedure that decides the substance of the case. My final point is one that I made very briefly at the end of my paper: if the national courts' decisions are subject to appeal on the merits of the case, and therefore under the new system we will have full judicial review of national court decisions in individual competition law cases, then you must also consider whether it is tolerable to continue with Commission administrative decisions that are subject only to limited judicial review. Is it tolerable that a small

486

Panel Discussion

contract—between small private parties, if you like-—be open to complete review on the merits of the case and, if necessary, also open to a reference to the European Court of Justice, while major contracts concerning important market players are subject to administrative decisions and judicial review on legality aspects only? People have now been making this point for years and years and years. Even at a very simple level, what will be the status of the hearing officer and his report under the new system? Will the report be made available to the parties? And so on. We should therefore consider the knock-on effect of judicialising part of the Community's competition law enforcement while continuing with a limited administrative procedure in the Commission's own backyard. • DAN GOYDER—May I begin by expressing all my support for what Judge Edward just said about not mixing up judicial and administrative functions. I also said something along those lines yesterday, and I think this is a very important aspect of the reform. You cannot have it both ways: you can have either an administrative or a judicial enforcement system, and you must be very careful that the administrative side does not—as it would-—take the precedence. I have great sympathy for the judges facing complex economic issues and I think that, as a starting point, we must not be greedy. Judges in every jurisdiction are very valuable assets: they have limited time and energy and, and they cannot spend all of both dealing with competition cases. Yet competition cases are very 'greedy'. I mentioned in my paper the Premier League case, which was, in my opinion, an appalling example of a waste of resources, and of course, the [previous U.K.] Restrictive Practices Act must take some of the blame for that. But let us just stop at this point: we need judges, the judges should be involved, and they can cope with the problems, but they need a great deal of help, and that help can be provided in a number of ways. I am going to mention two such ways in particular. The first was touched upon by Judge Cooke, and I think it is very important: the new Regulation should deal with the relationship between national competition authorities and national courts. I would like to see an enforcement system whereby not only the parties are obliged to file their arguments with the national competition authorities—as was suggested by Judge Cooke—but also the national courts have the right to ask for all relevant documents in the possession of the national competition authorities, and the latter would be obliged to help the courts, for example, with market definition—particularly when a lot of research has already been done on the markets at issue, and given that market definition is a very time-consuming occupation—and other matters relevant to the case. In some cases there might be no relevant records already available at the national competition authorities' offices, yet in others there will be quite a considerable amount of valuable information. It is therefore very important that the work carried out by the national competition authorities from a public interest perspective be brought into the cases before courts: this will not only

3 - Courts and Judges

487

save time, but very often will also ensure that important issues, which might otherwise be overlooked, are brought to light. The judges already face huge problems in establishing the facts in many of competition law cases. These problems will increase when the courts can apply Article 81 (3) because such cases involve a tremendous amount of factual evidence, especially when the parties are large and have many resources. Therefore, the greater the help that courts can get from the national competition authorities, who in turn will bring in—as was suggested—resources from Brussels, the better. The second kind of support consists of the fact that both judges and the parties should be confident that the hierarchy of appeal is logical and correct. As already mentioned, at present there is an obvious discrepancy between cases heard by the Commission and those dealt with by the national courts, a situation that is clearly unsatisfactory. I would therefore make a strong plea that Article 234 EC Treaty be amended. It seems to me—and this has also been publicly acknowledged by the President of the European Court of Justice—that a system in which it takes nearly two years to obtain an opinion, or simply an answer to a preliminary question, from the European Court of Justice, is not justice. Justice delayed is justice denied. Frankly, a system in which it takes more than a year to obtain such opinions or preliminary rulings is not something we can continue. Considering the rationale on which the Court of First Instance was established, I would like to join Jeremy Lever Q.C. and others in arguing that it is time to give this court Article 234 jurisdiction in competition law cases—and maybe in other fields of Community law as well, for example trade marks—and that in the future the Court of First Instance should sit not only in Luxembourg but on a regional basis throughout the Community in order to deal with appeals in such cases from national courts. These are reforms that seem to me to be essential, because without a proper judicial architecture you cannot ask national judges to do something that is so difficult. Incidentally, while the idea of having judges specialised in competition law is great, you should nevertheless be warned that national judges often have more than one specialisation. When I mentioned to a friend in London that a particular UK judge would make a wonderful competition law judge, he replied: 'Oh no, you will rarely see him in London. He is one of the only judges speaking Welsh, and is needed to deal with murder and other serious crime in North Wales!' To summarise, I think we have to start by assuming that judges can deal with competition law cases, but we must give them every structural support possible and intensive documentation if we are not to see their efforts wasted, and we must realize that judges have many other things to do too. • JOCHEM GRONING—Yesterday we heard much about certain 'German concerns' about the reform proposal. I would just like to let you know that the German judges—me included—do not play any active part in this discussion, which, as far as I can see, involves the Commission, the Bundeskartellamt, the German Monopolkommission and German academics.

488

Panel Discussion

What could I contribute, as a national judge, to this Workshop? First, there is not much left for me to say anew. Second, my position as a national judge is similar to wandering in a valley and trying to find out what it looks like from above the hill. Third, I belong to a Member State whose antitrust authorities hold the reputation of being tough and effective, and their administrative actions are under full judicial review by administrative courts. As I heard during yesterday's discussion, this is not the case everywhere in the Community. So, I might easily underestimate the problems that the Commission's reform proposal might raise. With respect to changing the interpretation of Article 81: normally we change the interpretation of rules because we find out that our former interpretation was wrong. The case is different with the reform of EC competition law enforcement: the interpretation of Article 81 is not going to be changed because it has been wrong, but because it is no longer practical. This approach might be rare, but not without precedent: I remember we were astonished a couple of years ago by the decision of the European Court of Justice in the famous Keck and Mithouard case,1 where the Court cancelled an interpretation of Article 28 that had been upheld for more than twenty years (since the Dassonville and Cassis de Dijon cases).2 Still, I have my personal doubts about whether it is possible to change the interpretation of Article 81 so fundamentally. I tend to believe that the essence of the exemptions under paragraph 3 implies that they are given for a limited period of time, and as such they are not exactly suited for civil law proceedings. I understand, however, that the Commission opted for Article 81 (3) to be applied in national courts because it fears that this provision might not always be applied correctly by the national competition authorities. The Commission should, however, lay out clear rules destined to avoid parallel competencies; this also relates to the issue of forum shopping, which does not necessarily have to be picked up again. What will be the scenario when civil law judges come to apply Article 81 (3)? For me personally this is a fascinating task—I have been dealing with antitrust cases for a couple of years now, and with merger cases in particular, whose structure does not seem to be so different from that of cases involving paragraph 3, but to trying to convince all my colleagues throughout the Community that this is a very interesting task is similar to trying to convince them that the taste of lemon is sweet. It seems to me that the task of applying paragraph 3 does not pose so much legal difficulties as factual ones. Such difficulties arise from the big mass of facts that need to be brought together, and from the economic analysis required in order to put all the mosaic of facts together and get a complete image of the economic and competitive conditions in the case. As such cases are normally rather complex, it will take a certain time—a couple of years, maybe longer—for national judges to get used to them. I think that com1

Case C-268/91 Keck and Mithouard [1993] ECR 1-6097. Case 8/74 Procureur du Roi v Dassonville [1974] ECR 837, and Case 120/78 ReweZentral v Bundesmonopolverwaltung fiir Brannwein (Cassis de Dijon) [1979] ECR 649. 2

3 - Courts and Judges

489

petition as an institution will survive this. What concerns me more, however, is that the application of paragraph 3 relies upon what the litigants bring before the court. As was said before, powerful undertakings might not be willing to disclose the whole economic and competitive background of the case, while their small opponents might not be able to do so. So, this is the point at which judgments could go wrong. Now, the Commission wants to help us as amicus curiae. If that is realistic, it would be great. But I think it will take a lot of the Commission's resources to be on the run in every single national and regional market in fifteen to twenty Member States, so this kind of help seems only theoretically possible. As to other kinds of support from the Commission, such as support in the interpretation of Article 81: as an independent judge, I do not like much the idea of being told by the Commission how to interpret the law correctly. I prefer to do it on my own. Of course, I find what Judge Wood suggested in her paper very interesting, namely that the Commission should just point out the competition policy questions and the application of competition law. We should indeed be provided with information on developments in Community competition policy, but the Commission should not intervene by the judge himself. Afinalpoint: the danger of abusing the possibility of requesting preliminary rulings from the European Court of Justice will be high. In my eyes, factual problems are the object of many such requests, rather than legal interpretation ones. The judges, and especially German judges, love to ask the European Court of Justice, for example, what percentage of the customers must be misled or wronged by misleading advertisement. They have brought three such cases before the Court, and I hope they understand by now that the Court does not want to answer such questions. • SANTIAGO MARTINEZ LAGE—The White Paper aims to reduce the Commission's burden in the enforcement of Article 81 in two ways: first, by entrusting national competition authorities with applying Article 81 as a whole and, second, by opening the application of Article 81 to private enforcement before national courts. Claus Ehlermann observed—and I agree with him— that in reality the main element of novelty in the White Paper is the less prominent role attributed in the new enforcement system to competition authorities, at both the EU and national levels. In my opinion, judicial enforcement raises quite different problems from those of administrative enforcement, and, while the application of Article 81 by national competition authorities is the real decentralisation, judicial enforcement has nothing to do with it. Even more important, in my opinion, is the fact that the judiciary cannot replace the enforcement functions of administrative authorities. Judicial enforcement is not an alternative to administrative enforcement, and I will seek to argue this point of view. There are three main differences between administrative and judicial enforcement. The first is that each of the two types of enforcement implies adopting

490

Panel Discussion

quite a different perspective. When competition authorities deal with a case, they are mainly concerned with the competition policy aspects, such as the structure of the market and the conditions for competition. Therefore the interests of third parties (clients, competitors or suppliers) are always taken into consideration. By contrast, at least in my experience, when competition issues arise in court proceedings only purely private interests are considered. The second difference is closely related to the first. When administrative authorities intervene, either following notification or in infringement procedures, the parties to the agreement stand close together to defend it against third parties— usually competitors, suppliers, or clients—whereas when Article 81 cases are dealt with in court, it is usually the parties to the agreement that confront each other. The third difference is that administrative authorities and judges have different possibilities for intervention. Administrative authorities may, and usually do, suggest modifications of an agreement so as to make it compatible with the competition rules, or they might require certain commitments from the parties to the agreement, whereas judges can only accept or reject claims. Therefore, whether or not judicial enforcement of Community competition rules is desirable, it is not an alternative to administrative enforcement. Some critics of the White Paper have argued that we are actually trying to get closer to the US antitrust enforcement model without having the legal instruments that the US system employs, such as class actions, treble damages, criminal sanctions, and, last but probably not least, contingency fees for lawyers, which are so uncommon in the European jurisdictions. Since I am Spanish, allow me to tell you about the Spanish jurisprudence related to national competition rules that probably worries the European Commission. In 1985 the Spanish Supreme Court ruled that civil courts cannot apply national competition law in the case of an abuse of dominant position unless our competition authority verified whether the abuse of dominant position existed. This judgment was followed in 1993 by another quite famous decision of the Supreme Court, according to which the same condition applies to Community competition rules (namely to Article 82 cases). In relation to the latter, in December 1993 the Supreme Court ruled that, whenever the principal claim was based on a Community competition rule (and the Community competition rules were not only incidentally raised in the case), the civil court should stay the proceedings until a decision was taken by either the European Commission or the Spanish competition authority. This was confirmed last November by another Supreme Court judgment, according to which, in both matters of national law and Community competition law, agreements need to be verified by the competition authority before civil courts can decide upon the civil law sanctions applying to the contract (nullity, etc). I know that the Spanish jurisprudence I referred to should be considered as contradicting the European Court of Justice's decision in the BRT v Sabam case,3 according to which Articles 81 (1) and Article 82 have direct effect. Case 127/73 BRTv Sabam [1974] ECR 51.

3 - Courts and Judges

491

However, I wanted to underline that the same condition was first applied to national competition rules, which is very important in my opinion: it shows that both national and Community competition rules were treated in the same way. At any rate, the Spanish experience demonstrates the difficulties that ordinary courts have in dealing with the complex economic analysis involved in the application of Articles 81 (1) and 82, let alone 81 (3). So, I will conclude by sharing Judge Edward's words: do not underestimate the problems of enforcement in national courts, or in the Spanish courts at least.

• ANNE SPIRITUS-DASSESSE—I would like to comment on three particular issues: the specialisation of courts, the way in which decisions are taken by the Commission as compared to the way in which they are taken by national courts, and preliminary rulings. On specialisation: if we were to opt for specialised courts, then 'specialisation' should be understood in a broader sense, namely as 'courts with a specialisation in business cases'. The parties come before the court only if they have a personal interest. This means that competition law issues arise in court in two types of cases. One is where the parties ask the court to validate their agreement; this is where the issue of applying Article 81 (1), and eventually 81 (3), arises directly. The other type of case is where the parties come before the court for other business matters (patents, trade marks, distribution agreements, parallel imports) and the question of the validity of the agreement arises in the context of this litigation. In practice, this second type of case is the most frequent. It is in this perspective that we should look at the application of Article 81 (3) by national courts. It would be very difficult to reserve the application of this provision to specialised courts because national judges have the competence to decide upon an agreement infringing Article 81 (1), and the question of whether the agreement qualifies for an exemption under 81 (3) or not would implicitly arise in some of the cases before them. There is a beautiful example in this sense in the Belgian jurisdiction: when the state is the defendant of a case, it cannot be considered as a business entity according to Belgian law. This means that the Belgian Commercial Courts have no jurisdiction over commercial cases in which the state is the defendant. For example, the famous cases concerning the legality under the provisions of the EC Treaty of the double control of the sanitary control of the quality of meat—i.e. control in the Member State of origin and in the Member State of destination—where the Belgian state was the defendant had to go to before an ordinary court, whereas cases where the state is the plaintiff and a firm is the defendant go before the Commercial Court. Therefore, I think—and here I am in agreement with Laurence Idot—that it is very difficult to limit the interpretation and application of EC competition rules to specialised competition courts. This would imply that, in practice, civil courts would very often have to refer the EC competition law questions arising in the cases before them to specialised courts. Yet in cases concerning patents,

492

Panel Discussion

trademarks, or sale agreements—to give just a few examples—time is very often of the essence. The parties need a quick decision. The courts have to deliver it, and cannot just deny their jurisdiction over the case. As we say, in such cases we have to decide fast, and cannot ask a specialised competition court: 'please, tell us if 81 (1) or 81 (3) apply, and only then we decide about the patent, trade mark, or any other issue that is the main subject of the litigation'. This is also the reason why I disagree with what Barry Hawk said yesterday, in the sense that the application of Articles 81 and 82 EC Treaty is about competition only. For a judge, competition rules are just a part of the Treaty. In order to achieve the purposes of EC competition rules, sometimes judges also need to apply the old Article 5 of the Treaty, or combine Article 81 with the provisions on states aids, and so on. This brings me to the issue of forum shopping. I believe that forum shopping cannot be avoided. As we all know, in competition cases the substance may be decided on the basis of the EC rules, but the effects of prohibiting decisions are determined under the national law, and with respect to the latter there are big differences among the Member States. For example, an English judge can award substantial damages, whereas a Belgian judge can award only very small ones. So, even if rules on jurisdiction would send the parties before a Belgian court to settle the substance of the case, the parties can always decide afterwards to sue for damages in another Member State, where the law is more generous on this aspect. The decisions of the European Court of Justice always refer to national law in so far as the effects of decisions and the procedural aspects of their implementation are concerned. I think we should not exaggerate the dangers of forum shopping because there are positive as well as negative aspects to it. Besides, the parties must consider a variety of aspects before deciding which forum is finally more advantageous, so that in the end it is not at all so clear whether they will derive such a big advantage from choosing among jurisdictions. As to the different ways of reaching a decision within the administration and within the judiciary: Judge Edward has pointed this out—and I think he is right.—the courts base their decisions only on the evidence put before them by the lawyers. In Belgium, at least, the courts can also ask third parties to provide evidence if there are reasons to believe such evidence exists, or can ask the third parties to stand before the court as witnesses, or call expert witnesses. It is in the nature of court proceedings that the parties are in contradiction, so that they are bound to bring forward plenty of evidence to support their claims. Moreover, the parties will always bring new evidence to contradict that previously presented, so that the court finally needs to undertake a lengthy and complex analysis of the facts presented. My last considerations are about requests for preliminary rulings. At present, such rulings can be requested to the ECJ. It would be the most clamorous mistake to deprive the national courts of first instance, as well as the national courts of appeal, from the possibility to request such preliminary rulings, as some would want to do. If you look at the jurisprudence of the European Court

3 - Courts and Judges

493

of Justice, the most complex and interesting legal questions were raised before the national courts of first instance. In addition, as I mentioned before, national courts are often confronted with crisis situations and they cannot delay reaching a decision. This also means that, as national courts cannot always wait for their national supreme court to make a request for a preliminary ruling to the European Court of Justice, the decisions taken in such circumstance might sometimes be wrong. In such cases a definitive damage is done, even if afterwards the European Court of Justice comes to a different conclusion on the case. Bankruptcy cases are the best illustration in this sense: such cases are timesensitive, and although an erroneous bankruptcy decision can be overturned on appeal, the reversal on appeal of the erroneous bankruptcy decision cannot restore the position of the company that was declared bankrupt. The benefits of the possibilities of appeal are purely academic, because an erroneous bankruptcy decision causes an irreparable damage: a firm never survives after having been declared bankrupt. • CLAUS-DIETER EHLERMANN—I will pass the floor now to Judge Sarah Vance. We are very privileged and grateful to her and Judge Diane Wood for being here to share with us their US experiences and impressions about this European debate. If somebody can bring forth a considered judgment about our European idiosyncrasies when it comes to antitrust enforcement, it should be the American judges. • SARAH S. VANCE—I will confine my remarks to my views on the ability of the national judges to apply Article 81 (3), and to what I think is necessary for them to do this effectively. Some of my remarks will be based on what I would want to know if I were a national judge in Europe called upon to apply Article 81 (3). You must know, however, that my perspective is colored by the US traditions of a strong and independent federal judiciary. In the United States, federal judges serve life tenures. They also are generalists, meaning that they hear any kind of case, from competition cases to intellectual property cases, criminal cases or interstate traffic accidents, as long as people from different States and enough money are involved. So, it should not surprise anyone that I think that national judges are quite capable of deciding on competition issues and of applying Article 81 (3). It has been pointed out by a number of speakers here that national judges already apply EC law in other areas, that they apply Article 82, and are experienced in balancing divergent interests, so they should be reasonably capable of performing the economic analysis involved in the application of Article 81 (3). Having said this, I think that several factors are necessary in order to assist the national courts in applying EC competition rules effectively. First of all, education and training on the fundamentals of EC competition law should be made available to national judges, particularly if there are some judiciaries that are less sophisticated than others on competition issues. In the US, the Federal

494

Panel Discussion

Judicial Centre provides federal judges with educational seminars by prominent academics, and there are numerous Bar associations offering training programmes and materials for continuing this legal education. I think that, if the Commission invested some resources in the training and education of national judges, it would be time and money well spent. Judges also need immediate access to Commission decisions and decisions by the courts of other Member States based on Article 81 (3). In the US, we have instantaneous Internet access to the decisions of State courts, federal courts, and the US Supreme Court, and this is very beneficial in keeping us up to date on the law. In addition, if national courts are expected to make use of economic analysis, they must have access to the expertise of economists. There are a number of ways to accomplish this, from appointing experts to sit with judges, to enforcement agencies offering their input. In the US we usually rely on the parties to hire their own experts, but we are authorised—under the Federal Rules of Evidence—to appoint our own experts if we feel that the testimony of the experts brought by the parties is partisan and not helpful. The cost of the independent expert can be assessed to the parties. The expert will make his opinion known to the parties in advance of the trial, and he will be subject to crossexamination by the parties. Judge Kimba Wood, for example, who used to be an antitrust lawyer, has appointed an economic expert in New York v Kraft General Foods,4 and relied very heavily on his testimony. Moreover, for national courts to apply Article 81 (3) in a coherent fashion, they have to be able to answer a series of questions in a reasonably predictable way. First of all, the judge has to know if the case belongs in his/her court, or whether it should be somewhere else. If the rules on allocation of cases are not clear, then they ought to be clarified. In the US, we have provisions according to which the judges can transfer a case to a more appropriate forum, or can consolidate related cases in the same forum. These types of instruments facilitate co-operation among judges, and make the delivery of justice more efficient. Second, judges must also be able to determine the law that applies. In this regard, the issue of the primacy of EC law must be clearly established, and judges must know if they are bound by Commission decisions. In the US, if the government obtains a civil or a criminal judgment against a defendant to the effect that he/she has violated the anti-trust laws, this judgment is primafaciae evidence of a violation in any other court proceedings between that defendant and third parties. Third, the court has to know who has to prove what in order to prevail under Article 81 (3). Here, either the Commission or the Luxembourg Courts should clarify the approach under Article 81 (3): how much of the competition analysis is done under paragraph 1, and how much is done under paragraph 3? Who has the burden of proof? What is the test for market definition and market power? What factors are relevant to assess pro-competitive and anticompetitive effects? Are only competition issues relevant under Article 81 (3)? 4

New York v Kraft General Foods, Inc., et al., 926 F. Supp. 321 (S.D.N. Y 1995).

3 - Courts and Judges

495

My reading of Article 81 (3) is that it is a competition provision and not an allpurpose catch-all that will consider any policy matter that may strike the judge's fancy. Of course, if there are other statutes, other laws that are relevant—as Judge Wood pointed out yesterday—that should be taken into account, but Article 81 (3) sounds to me like an anti-trust provision, and it should be applied as such. Fourth, the court has to know what remedy to award to the prevailing party. In the US, federal courts impose the same set of remedies for antitrust violations. I suggest that harmonisation of remedies in the national courts would lend coherence to EC competition enforcement and minimise forum shopping. My experience in the US is that, in competition cases, a judge will rarely have a precedent that is on all fours with the case in question ready to pull out of a hat. Competition cases are, indeed, fact-intensive. Yet if judges know the general approach to use and they look at the right factors, this means that a certain level of consistency has already been ensured, even if reasonable minds may differ on the result. Finally, involving national courts in the application of Article 81 (3) will undoubtedly produce some inconsistent and some wrong decisions. I do not think this should be unnecessarily troubling as long as there is an effective mechanism for the Community interest to be asserted and established. In this regard, in my experience as a US trial judge, it is an enormous self-discipline to know that I can be reversed on appeal. Yet in Europe there is no possibility of appeal from a national court to Community courts. If there is concern about parochialism in national courts, an option would be to enable the Commission to appeal national court decisions to the Community Courts. If that is not possible, I think that the preliminary ruling mechanism needs to be strengthened, so that it can be faster and more responsive to the national courts. Here I agree that lower national courts should be able to ask for preliminary rulings, because it is there that the questions arise in the first place.

• DIANE P. WOOD—Before all, I should say that I was impressed when listening this morning to the other judges around the table by our common understanding about what it means to be a judge and what it means to entrust certain enforcement tasks to the judiciary. I will quickly review the arguments in my paper, and, if time permits, I will also make a couple of other observations. The first aspect I discussed in my paper refers to ascertaining what the law is. Here, as a number of the participants around this table already pointed out, there are a number of sources to be considered: there are, of course, the Commission's decisions, and there are the decisions of the two European Courts. I think that the weight given to the Commission's official pronouncements will increase if the Commission continues to exert jurisdiction over the more important competition cases, or those cases that seem to need its attention. I mentioned in my paper the Chevron doctrine in the US, according to which, in essence, the judges should give the benefit of the doubt to the antitrust

496

Panel Discussion

agency instead of approaching a question from a fresh perspective, as long as the agency has stayed within the outer boundaries of its discretion. It seems to me that this doctrine is similar to your suggestion that the Commission should continue to have a leadership role in the enforcement of EC competition law. It is not that there are no outer limits to such a role and, as we all know, the Commission has sometimes stepped beyond those limits, but this usually does not happen. And if there are reasons why the Commission has interpreted the Treaty competition provisions in a certain way, those reasons should normally be respected by the courts. A short remark about balancing tests: if you want to see a line of decisions vigorously criticising balancing tests, I recommend those of my colleague at the United States Court of Appeals for the Seventh Circuit Frank Easterbrook, who might be known to some of you from his antitrust writings. I actually share many of Frank's views about this, because there are two kinds of balancing tests. One kind are those saying: 'Here are x factors, think about them all and come up with a result.' With such tests it is hard to predict the outcome of a case, because the factors can be balanced in several different ways. Such balancing tests are difficult for judges to apply, and they do not bring about much predictability in enforcement for the parties and the public as a whole. The other kind are the so-called 'weighted balancing tests', where the judge is not just given five random factors, but is also told that, when these factors are in conflict, number one is more important than number two, and number two is more important than number three, for example. I strongly recommend adoption of the second kind of balancing test with respect to rule of reason cases. While the rule of reason might itself be the ultimate balancing test in antitrust, in the US (or at least at the United States Court of Appeals for the Seventh Circuit) we made it clear that market power is a prerequisite for worrying about any rule of reason case, since those are in fact the cases where you are looking for market effects. I think this has brought a great deal more definition into what we are doing. Let me move now to the issue of co-ordination among national courts. There are so many possible ways to help the national courts communicate among each other and with the various European institutions about the content of the law they are about to apply, that I do not think we need to review them all, so let me just make on or two remarks on this aspect. In the US, the federal antitrust authorities have the right to intervene during the court proceedings, with party status, whenever the cases pending before the court are considered to raise public interest issues. As I mentioned in my paper, federal statutes require that the Attorney General be notified whenever a matter pending before the court could involve the interests of the United States. In reality, this procedure is not much used in antitrust cases, yet another procedure is much more used, and this is the following. At the time when I was working at the Antitrust Division of the Department of Justice, one of my responsibilities—aside from the international work—was to handle the appellate work of the Division. This included a variety of issues,

3 - Courts and Judges

497

from criminal appeals on sentencing guidelines to the modified final judgment in the AT&T case, to give just some examples. As Appellate Deputy I had to hear parties trying to persuade the Antitrust Division to file an amicus brief in an antitrust case pending before a federal court which appeared to raise matters involving our antitrust enforcement policy in general. Such parties came to us quite often, and sometimes I would be persuaded that the case was important. Then I would go to the Solicitor General and try to obtain his permission to file an amicus brief, and sometimes I got it. This was a way to allow people to come in and talk to us, the federal authority. Sometimes we would inform the federal court that we did not want to have the status of a party in the case—you might not always want to jump into the deep end of the pool—but simply wanted to inform the court of the government's interest in the case. We could do that, or we could do something more involving a greater time and energy commitment on behalf of the federal antitrust authorities. On the subject of coherence, a couple of issues were raised during our discussion here. One of them refers to the effects of the decisions taken by the national courts in other Member States and in Brussels. On this issue, I think that courts decide cases between two parties, and if a case involving competition issues was decided between parties A and B, it seems to me that this should be the end of the matter between A and B, even if A and B move from one jurisdiction to another, unless there is some profound legitimacy problem with the original decision. Assuming there is nothing fundamentally wrong with the original decision, that should be the end of it. The decision might be right or it may be wrong but, at least in the US system, the principles of finality are very important, and sometimes they override the objectives of getting it right. The last point I wanted to make is about arbitration. Here I think we have to keep in mind two important points. One is that arbitration works fine if you are talking about consensual arrangements, or arrangements that people make in the open: vertical agreements, joint ventures, and agreements in general. I do not think, though, that arbitration makes any sense at all if you are dealing with a cartel case, where people are sneaking around in back rooms doing something that is plainly illegal. What are they going to do, arbitrate the case? Or is a third party going to say,' Hey, we would like an arbitration?' I do not think so. We might want to think about which antitrust cases lend themselves to arbitration. Second, and perhaps much more importantly, is that even if you figured out how to get arbitration into the court system—for instance, an arbitral award might be put before a national court for recognition and enforcement—the scope of review of arbitral awards is still very narrow. In fact if the courts just sat down and re-did the case, this would destroy the arbitration system. Moreover, only the most fundamental arbitration errors—under the New York Convention or other national arbitration laws—can be reviewed in court, and this poses a real problem for competition law enforcement, to the extent that we have a public interest to ensure that this body of law is properly administered.

498

Panel Discussion

• CLAUS-DIETER EHLERMANN—One of the greatest obstacles to any sort of co-operation among the courts of different Member States will be the language barrier. For the same reason, the network of competition authorities is already difficult to achieve. Yet in this respect the judges are clearly even more disfavored, and this goes way beyond competition law enforcement. It is a more general problem. • JAMES VENIT—I will try to be provocative, because I think we have the right mix of participants here to raise these questions. When we discussed the need to reform the EC antitrust enforcement system at the Workshop of four years ago, one of the differences between the EC and the US antitrust enforcement systems that emerged from the discussion and struck me most was that the US system relies to a greater extent on private enforcement and third party actions. So, it occurred to some of us that part of the reform of the Community competition law enforcement system and, in a way, also a form of decentralisation, would be the strengthening of private enforcement. When one looked at the problem, there were certain incentives to private enforcement in the US that did' not exist in Europe. Most notably, these are treble damages, positive discovery rules, maybe class action suits, maybe contingency fees also. I mean, one thing you have in Europe in most jurisdictions is cost, which is another feature of the US system. And then, as one thought further about that, one realised that those measures were needed to provide an incentive for third party enforcement action. How would you go about adopting that? One possibility would be to harmonise national rules through some sort of Community instrument. The one question that I have is: how would national judges react to that? The second possibility would be to create a system of federal courts. There are some attractions to that, particularly to the extent that the European Court of Justice and the Court of First Instance are severely overloaded. Creating an additional judicial body to absorb some of that workload might be attractive. So, I guess the question is whether strengthening private enforcement would be a desirable thing? If it were, should it be accomplished at the national level, or at federal level? And if it is not, why not? • JAMES RILL—I would like to comment of the issue of specialised courts. Here I draw on my experience shortly after having left Law School, in the early 1960s, when a report by the Administrative Conference proposed the establishment of a specialised Trade Court in the US. Both the Congress and the Bar rejected this proposal, retaining that, first, 'trade is too broad to be specialised', and second, it was feared that specialised Trade Court judges appointed for life could stultify, rather than provide for theflexibilityneeded in antitrust deliberations. This kind of skepticism with respect to specialised courts persists to our day. It would be interesting to learn the views of Judge Vance or Judge Wood on the efficiency of the federal circuit, which seems to be quite specialised in

3 - Courts and Judges

.

499

handling patent and some other intellectual property disputes, and which one finds cannot be 'vacuum-packed' so as to not have spill-over effects into antitrust. I think that judges dealing with antitrust matters should have experience with a broader range of judicial matters, especially in the field of commercial and contract law. Antitrust cannot be 'vacuum-packed'; this area of law enforcement is influenced by the reality of the time, and I think that trying to 'vacuum-pack' judges into trade law with no experience in other commercial activities is a mistake. At least that is what the US experience suggests. At the same time, I think the suggestion that the Commission should be able to intervene in national court proceedings as amicus curiae is good. As indicated by Judge Wood, in the US the experience with filing amicus briefs has proved very helpful. I should mention that the amicus brief instrument can be very important: the US government expressed through it its position on predatory pricing in the Cargill case, and on summary judgment in the Matsushita case, for example. I think these briefs were quite influential on the Supreme Court, as was the government's position both in the Court of Appeals and in the Supreme Court on the Insurance Antitrust Litigation case, both as to the insurance exemption and as to extra-territorial jurisdiction. I think amicus briefs are not so labor intensive as to drain resources from the Commission, and the influence that the Commission might have within the deliberation in national courts seems to me to be potentially salutary. • GIULIANO MARENCO—I would like to deal telegraphically with several points. The first relates to the broad interpretation given to Article 81 (1), an aspect that has been commented on by several participants. I would like to say that, in my view, in this regard the present system is perverse in several ways. Above all, the civil servant who must look at thousands of pages of contracts in order to decide whether an agreement can be exempted or not will be frustrated if at the end of this examination he does not find a restrictive clause so that he could oblige the undertakings to modify it. Yet even more perverse is the effect on the undertakings, which prefer to ask for a negative clearance, rather than an exemption. They would claim their agreement restricts competition, so that it be considered as falling under the scope of Article 81(1) and benefit from an exemption under Article 81 (3). Finally, I cannot help thinking that even the European Court of Justice, in realizing the awkwardness of the enforcement system under Regulation 17/62, has in a way broadened the scope of Article 81 (1), in the sense of including some of Article 81 (3) into Article 81 (1) in order to make things work better in real life. My second point relates to the temporary nature of exemptions under Article 81 (3). If I understood well, Judge Groning said that the content of Article 81 (3) intrinsically points to the temporary nature of the exemptions. I do not see that in Article 81 (3), but I see it in the new enforcement system proposed by the White Paper. Under the new system there will be no indefinite exemptions at all, because all exemption decisions will apply to the past or the

500

Panel Discussion

present. So, there will be nothing more temporary than the application of Article 81 (3). As to the four conditions under which Article 81 (3) applies, I understand— but I would like the American participants around the table to confirm this or not—this provision as intending to codify the American rule of reason. I see nothing beyond competition criteria in Article 81 (3), and if some Commission decisions (like Ford- Volkswagen5) included remarks going beyond competition, one should forgive the Commission for these remarks which were probably out of place. I think one should confess one's own weaknesses. What is important is that the European Court of Justice has established that an agreement needs to comply with the four conditions in Article 81 (3), and only those four conditions, in order to be given an exemption, and that is what counts, nothing else has importance. Judge Cooke has drawn our attention to paragraph 102 in the White paper, saying that the Commission reserves the possibility to prohibit an agreement despite the fact that the agreement has already been cleared by a national court. I think it is difficult to discuss this in the abstract, let me give an example: a firm sues another before a national judge for the non-execution of a contract. The defendant claims that the agreement is void and inapplicable because it is contrary to Article 81. The judge arrives at the conclusion that the agreement is all right, and pronounces against the party that has not executed it. Does this prevent the Commission, at a subsequent time, from issuing an order against the parties because the agreement is contrary to Article 81? As much as I understand and share the concerns voiced around this table in the sense that the Commission should not attempt to bind the national judges because there is a principle of separation of powers, at the same time I think that, if we give concurrent jurisdiction to administrative authorities and courts, then the administrative authority should also be free to decide. This is not against the court's pronouncement, because if you look at the operative part of what the judge has said and the operative part of the Commission's decision, there is no conflict between them. Another point: it has been said that the allocation of cases between the Commission and the national competition authorities should follow very clear criteria. Here I would have one question: are clear criteria for the allocation of cases a necessity for the undertakings, or for the national competition authorities? In my understanding, only the national competition authorities have an interest in clear allocation criteria, since I do not think that the undertakings have to know in advance which administration will treat their problem under the Community competition rules. A final point: I notice with sadness (or I have the impression) that the European Court of Justice would like to get rid of competition cases, and to leave only the Court of First Instance to deal with them in so far as preliminary 5

Ford-Volkswagen, Commission Decision 93/49/EEC of 23.12.1992, OJ L 20/14 1993.

3 - Courts and Judges

501

rulings are concerned. If I am wrong, I would be happy. Though I would like to remind you that in the US, where antitrust rules are not of a constitutional nature but were adopted at the legislative level, the Supreme Court deals with antitrust cases every year, and it does not miss an occasion to say that the Sherman Act is the economic constitutional chart of the US, thus giving antitrust law a prominent position within the legal system. •

CLAUS-DIETER EHLERMANN—What is within Article 81 (3)? One of the

greatest contributions that the Commission could make to the debate on this aspect is to spell out its position on the interpretation of 81 (3). I think this should be done by the Commission, not by DG COMP, and not by the European Court of Justice—which in the end will decide—because what you just said about the Ford- Volkswagen decision is of the utmost importance. • FREDERIC JENNY—I heard Judge Vance saying before 'To me, it looks like Article 81 (3) is part of normal competition law, it is about efficiency.' I asked some of the practitioners around this table how many Commission decisions they thought might have been different if Article 81 (3) were strictly about efficiency analysis. Some answered 'Not many, really.' Others said 'Well, maybe the Commission would have opened some more cases, but it did not do so because of those 'other policy goals' hidden in Article 81 (3).' To me it seems that Article 81 (3) could, at least in part, be read as pointing towards the straight efficiency analysis one would expect in competition law enforcement. If this is right, then Article 81 (3) is not too far from Article 10 of the French competition law. In fact, Articles 7 and 10 of the French competition law are basically the equivalent of Article 81 (1) and (3) under the EC Treaty, and these provisions are enforced by the Conseil de la Concurrence as well as by courts. And in France no one has ever argued that judges, who know 'nothing' about economics, have considerably erred in applying Articles 7 and 10.1 was very interested by what Judge Spiritus-Dassesse said about specialised courts, because, by contrast, I believe that court specialisation, or the concentration of cases in the hands of a reduced number of judges, is a useful thing. Judge Spiritus-Dassesse said that specialisation is hard to work with at the first instance level, because competition law issues arise in all kinds of different contexts. I recognise that. I just wanted to mention that the Court of Appeal of Paris, which handles appeals against the decisions of the Conseil de la Concurrence but also regular appeals against decisions by lower courts in the Paris area, decided to concentrate the competition cases in the hands of about ten of its judges—although there is no legal provision requiring for such specialisation—and in practice this has proved to be a helpful step. The judges acquired a better understanding of the economic issues, and the relationship between the Conseil de la Concurrence and the Court of Appeal has improved. So, I would strongly recommend the establishment of specialised courts, at least at the appellate level. Or, if this cannot be done, I would recommend finding a

502

Panel Discussion

way to concentrate competition cases in the hands of few judges in the civil courts, so that a competition culture and economic understanding of competition issues can develop. In so far as the courts offirstinstance are concerned, the French law provides for the possibility of referring cases involving competition law issues to the Conseil de la Concurrence for an opinion. This referral system has two advantages. One of these derives from the fact that civil judges might not always know what factual evidence is really relevant for the economic analysis required in competition cases. In such instances, the judge can request an expert opinion from the Conseil de la Concurrence, which has investigative power and understands the issue. The judge can then decide how to use this expert opinion, in other words, he can decide whether to follow or not the opinion given by the competition agency. The second advantage is, of course, one for the parties. Namely, the parties can petition the court to send their case to the Conseil de la Concurrence for an opinion, asking the Conseil to use its investigative means to establish the facts in full, or to find evidence on a particular aspect of the case. This referral system, however, is not without flaws, since it is up to the courts to decide whether to request the opinion of the Conseil or not. An alternative would be to oblige the courts receiving a petition in this sense from a party to send the case to the Conseil for an opinion, although in the end the courts would still have the possibility to decide whether to take into account the expert agency's opinion or not. In any case, so far quite a number of references of the kind were made to the Conseil, and I think the system has proved rather efficient in so far as both the relationship between the judges and the national competition authority and the role played by the latter in enforcement are concerned. And I think that this experience is not too far from the Spanish one, which is different procedurally, but quite interesting as well. I believe that this is something to consider in terms of modalities to facilitate the work of the judges in the new enforcement system. • ULRICH IMMENGA—I would like first to follow up on the issue of court specialisation, which, as far as I can derive from this debate, is quite controversial. We also have some experience with specialised courts in Germany, and I think this experience has been positive. Just to clarify one point: in Germany there are about twelve or fifteen courts specialised in competition cases at the first instance level. Therefore, if competition issues arise in a case before a first instance civil court, the case will be referred to a specialised court and will be dealt with there. More precisely, the court does not simply refer to the specialised court the competition question while preserving the case, and this reference system has no dilatory effect with respect to the delivery of the decision. As a second point, I just wanted to emphasise an issue that was pointed out by the judges around this table: in Germany, and as far as I understand, in other jurisdictions as well, procedural rules applying to court proceedings oblige the

3 - Courts and Judges

503

judges to decide on the basis of the facts and arguments brought up by the parties, and that means by their lawyers. Third, I think it is necessary to introduce some of Professor Mestmacker's considerations proposals concerning the burden of proof and the legal consequences of the application of Article 81 as a whole in private courts. I quote: 'The Commission expects Article 81 to be applied as a whole in three kinds of proceedings: contractual liability proceedings (disputes between parties to an agreement), non-contractual liability proceedings (disputes between a third party and one or more parties to an agreement), and applications for injunctions (White Paper para. 99).' And then: 'In dealing with private enforcement, the narrow limits of Community law must be observed. The only Community law sanction is Article 81 (2), providing for invalidation of restrictive agreements, All other sanctions and remedies are a matter of civil law of the Member States. There are nevertheless important indirect effects of Article 81 (2) onto tort and unfair competition law. For private enforcement to be effective there is obviously a trade-off between legal certainty of the parties to an agreement and effective protection of legitimate third party interests. The Commission appears to be more concerned with creating legal certainty for parties to an agreement than with the protection of potential victims. The implications of the initial tolerance for all restrictive agreements and the absence of a presumption that Article 81 (3) does not apply are among the most important reasons why private enforcement is bound to be ineffective.'

I find the last point very important, since such 'initial tolerance' would have implications in so far as the primacy of Community law over national law is concerned. The 'globally satisfactory legal certainty' proposed by the White Paper applies to the parties to the agreement only, yet it has important indirect effects for the stability of the agreement and its defense against third parties. The Commission's reform proposal implies that the issue of the validity of an agreement will become relevant in all tort and unfair competition cases, since, according to the Commission, there will be no presumption under Community law in the sense that restrictive agreements are null and void. This also implies that, in all such instances, the general rules of evidence apply. In this respect, the position of German courts is representative: where the validity of an agreement is challenged, the plaintiff bears the burden of proving the facts establishing the invalidity of the agreement. The German courts refuse to modify this rule of evidence even if an EC Regulation or the case law of the European Court of Justice requires it. Yet in competition cases it is difficult for the plaintiff to obtain and prove the facts relevant to the application of Article 81 (3). In cartel cases, for instance, the plaintiff would have to prove facts that by their very nature are known to the cartel members only. To sum up, the Commission's confidence in the enforcement of Community competition rules in national court proceedings stands in striking contrast with its neglect of the conditions determining the effectiveness of this remedy. It is

504

Panel Discussion

also at odds with the key role played by individual rights in the enforcement of directly applicable Treaty provisions, and it is hardly compatible with the Commission's own position with respect to the direct applicability of Article 81 as a whole. • JOCHEM GRONING—I would like to reply to Giuliano Marenco: I am well aware that it is not written down in Article 81 (3) that exemptions are limited in time. For me there is a simple reason for assuming that: temporary limits are in the very nature of exemptions. If a contract infringes Article 81 (1) and it is given an exemption under Article 81 (3), this exemption cannot be but temporary, because one does not know how competition or the conditions of the market will develop over time. I think that the Commission knows this, and this is why it is trying to change the paradigm completely. I admire this, it is mentally so radical, and usually we are not so radical in our daily legal practice. I just entertain some doubts as to whether it is to the benefit of competition or to the benefit of the parties to obtain judgments that are good only for the past and the present. The litigants prefer to have judgments covering the future, and this is what they had so far with the system of Commission exemptions. I think that the concept of exemptions is workable, and could in the future be applied by the national competition authorities. Nevertheless, as I already mentioned several times, I do not think the concept of exemptions really fits with the procedural principles of civil litigation. • DAN GOYDER—In the UK an important step has already been taken towards the integration of public and private enforcement in such a way as to help judges. For the first time, under Section 58 of the [1998] Competition Act, the findings of fact by the Office of Fair Trading are binding on courts, and that includes third party actions for damages. That, of course, is subject to one qualification, namely that the findings of fact of the OFT are not overturned on appeal by the Competition Commission. This is an important first step, which of course does not provide all the necessary answers, but at least it does give the judges some important help. • ELEANOR FOX—I wanted to say something about the issue of 'private rights versus the public interest'. Diane Wood and James Rill also referred to this before. I wanted to add something about laying the foundation right: when law provisions have direct effect, this creates rights for the individuals in terms of enforcement, which is often confounded with the idea that the individual obtains a right on its own. The new enforcement system will place more emphasis on complaints, and we all know that complaints are very often made by competitors who do not like competition. And so, it is very important to recognise that private rights might sometimes conflict with the public interest in competition cases, that competitors might sometimes be complaining about

3 - Courts and Judges

505

competition itself, and that if effect is given to the individual rights without thinking further about whether those rights do or do not correspond with the interests of consumers, distorted case law might develop. In the Community jurisprudence, a violation of Article 28 EC Treaty (the old Article 30) whereby private individuals are deprived of their right to sell their goods across the border corresponds with the public interest, in this instance the free movement of goods. Yet when an individual sues because a competitor has an exclusive contract, this might conflict with the public interest. I think it is very important to recognise that in competition cases it is often very difficult to distinguish between private and public interests, and therefore the courts should not just assume that because a private party sues, public interest is at stake. • JOHN COOKE—Giuliano Marenco gave us before a concrete example in the application of the hypothesis at paragraph. 102 of the White Paper, and then addressed a question to me. I should try to answer his question, because if the position of the Commission is as outlined by Giuliano Marenco, then we may have a potentially serious constitutional problem. Take the following example: company A and company B concluded afive-yearcontract. At the end of the first year, company B refuses to perform. Company A sues for a court declaration that company B is obliged to perform, as well as for damages for the nonperformance. Company B defends itself arguing that the contract is in violation of Article 81, and therefore is void. Company A replies: 'If that is the case, which I deny, then the contract is 'saved' by Article 81 (3).' In the scheme of the White Paper, in such a case either the national competition authority (in my scenario) or the Commission (in the White Paper's scenario) would be notified of the existence of that litigation and given the opportunity to either submit observations or intervene. Whether the Commission declines that invitation, or whether it intervenes and its arguments are rejected by the court, it seems to me that the Commission should at any rate be bound by the court's decision. If the contract is upheld by the court and damages are awarded on the grounds that 81 (3) applies, then it seems to me that if the Commission could nevertheless later issue a 'cease and desist' order with respect to the performance of the contract the rights of both parties will have been damaged in a way that is inimical to the rule of law. Namely, the plaintiff is deprived of the performance of the contract, to which he was entitled for the balance of the five years, and the defendant has paid damages that he never would have had to pay. That, in my view, is the scenario envisaged in the White Paper, and this scenario raises important constitutional problems. • DAVID EDWARD—In relation to a question addressed to me earlier on by Emil Paulis: it is very important to distinguish, for judicial purposes, between Commission acts that have legal effect, and Commission soft law pronouncements that, as far as courts are concerned, may influence the approach in the application of the law, but cannot be binding. Moreover, according to the

506

Panel Discussion

Treaty, even Commission decisions are binding only on those to whom they are addressed. Therefore, an individual Commission decision, however significant it may be, does not bind a court in another issue. In addition, we should make a distinction between concurrent jurisdiction, which is what Giuliano Marenco was talking about, and the notion of administrative reversal of a judicial decision, which is what John Cooke was talking about. Please, bear in mind that, according to Article 81 (2)—a provision that we have mentioned here very little—an agreement or a practice that infringes Article 81 (1) is void—in French nulet non avenu—as if it never existed. Now, if a court decides that an agreement is valid, and the Commission takes later an administrative decision in the sense that the agreement it is not compatible with Article 81 (1), what is then the significance of Article 81 (2)? Has the agreement become nul et non avenu, or what has happened to it? The Commission must be clear on this. All I ask of you is to bear the distinction in mind and be clear about what your answer is. Second, on the issue of forum shopping: I entirely agree with Judge SpiritusDassesse, we will never be able to avoid forum shopping. Yet forum shopping is only one aspect of the conflict of laws, and there are other aspects to be considered too. One is competence or jurisdiction of the Court, and that is what affects forum shopping and, broadly speaking, is already dealt with by the Brussels Convention. But a second case of conflict of law is when two courts are competent, but one has to give priority to another. The Brussels Convention deals with that simply where you have the same issues between the same parties, and the Court has said in the recent Drouot case6 that 'same parties' means same parties, and not parties with a like interest. My illustration here is that of a standard contract entered into by a manufacturer with a number of suppliers. It is clear that the contract between A and B is not the same as the contract between A and C: it may have the same subject, but it is not between the same parties. You have to consider whether you need new rules, not on jurisdiction, but on Us alibi pendens—or on suspending or yielding the case to another court already seized with the same question, and then distinguish between conflicts of competence, and conflicts with respect to the effects of court decisions. For instance, what are the effects of one court's judgment with respect to the judgment of another court? This issue is completely different from the one of conflicts of jurisdiction. It goes back to the question of what is the effect of a Commission decision, which according to the Treaty is only binding between the parties. Is the decision of the Spanish Court on a dispute between the manufacturer and the distributor in Spain binding on a dispute in France between the manufacturer and the distributor in France? And to what extent is the decision of any of these Courts binding on the Commission? These are legal questions to consider. Last point: Giuliano Marenco said he has the impression that the European Court of Justice would like to transfer jurisdiction and competition cases to the 6

Case C-351/96 Drouot [1998] ECR 1-3075.

3 - Courts and Judges

507

Court of First Instance. The European Court of Justice has no position on that, and certainly my personal position is that it is up to the political institutions of the Community to decide who shall do what. But please make the rules of allocation of competence clear, and do not create conflicts of competence. We already have difficulties of the kind in state aid cases. When challenged by a company, a Commission decision on state aid goes to the Court of First Instance, whereas when challenged by a Member State the same Commission decision comes to the European Court of Justice. Please do not add to the length of time taken to deal with cases a conflict of jurisdiction between the European Court of Justice and the Court of First Instance or some other specialised court. Finally, Dan Goyder said that two years is too long to wait for an ECJ decision. Let me give you a precise example in this sense. Last week, as Rapporteur, I agreed in the Fifth Chamber on two judgments, and in the plenary three more. I have four other drafts with the lecteurs d'arrets. I have already signed off ten judgments this year. Of the five judgments agreed last week, three will be pronounced on the 4th of July, and two on the 14th of September, and that is because we do not have enough translators. That is, four months of delay simply because there are not enough translators. Now, you can transfer everything you like to the Court of First Instance, but unless you have the resources to make the system work, you will not get the results in the time you want. • JOCHEN BURRICHTER—I wanted to get back to the risk of forum shopping. I think that the danger of forum shopping is exaggerated. Let us take the case of a dispute about the validity of an agreement. Normally commercial agreements provide for a clear forum clause, so that there is no choice for the parties to go to different courts. In other cases the agreements contain an arbitration clause, so that again, there is no problem of forum shopping. And even if there is no express clause about the competent forum, the rules normally say—though I admit I have not done a thorough comparative study—that the competent forum is either at the place of performance of the contract or at the seat of the defendant. In this latter case, admittedly, there may be a slight possibility of forum shopping. In cases where third parties claim damages, for example, the general rules of tort indicate that the competent forum is at the place where the harm has been done or where the defendant has acted, so we have only two possibilities. Overall, I think that there are very limited, well defined cases, where forum shopping is possible, so that the risk should not be exaggerated. I would next return to the issue of court specialisation, and add a few remarks to what Professor Immenga has already said. Until January 1999, Germany had a system of distribution of court competencies whereby, if in the course of the court procedures, a competition law issue was detected, the court had to stay the proceedings and refer the competition law question to a specialised court. This reference system had exactly the dilatory effects that

508

Panel Discussion

some of you described. However, it has been abolished, and now Germany has a system whereby the specialised courts are seized with jurisdiction even if the competition issue arises on the background of another kind of litigation. Of course, objections based on competition law provisions might still arise in the course of proceedings before a civil court. In such cases there might in fact be a delay of the process. I must admit that lawyers might sometimes make use of the possibility of delaying the trial, or might submit competition-related arguments just to bypass the court dealing with the case. Yet these are exceptional cases, and this is not an argument against the establishment of specialised courts. Third, I would like to return to the issue of whether national procedural rules should or should not be harmonised. Yet we all know that the harmonisation of procedural rules will take years, maybe decades. I think, however, that if the 'disharmony' of procedural rules were to disadvantage the business community in one European jurisdiction as compared to another, then the business community itself will probably ask for the improvement of the rules. This means that a sort of competition between procedural systems might even help at an initial stage. • JACQUES BOURGEOIS—First of all, I would like to make a comment about the famous paragraph 102 in the White Paper, clarified by Giuliano Marenco and challenged by John Cooke. I was wondering what the purpose of this particular remedy was: is it to redress erroneous decisions by the national courts and avoid that they become legal precedents, or is it really to prevent the parties to an illegal agreement from 'getting away with it'? If the purpose is simply to redress erroneous court decisions, then I think that the decision should remain standing with respect to the parties, and for the rest, a remedy of the recours dans I'interet de la loi type could be used. Second, I would like to refer to the proposals made by Jochen Burrichter and somehow endorsed by Dan Goyder with respect to the Commission assisting parties in private litigation, and in particular assisting them with fact-finding. I have some difficulties with endorsing this proposal. The fact-finding activities of the Commission are performed for a public interest purpose: the Commission investigates the facts in order to issue decisions that are binding on the parties, including sanctions or penalties. Why should the result of such factfinding be put at the disposal of private parties in a private litigation? I would not like to add a detailed comment on what Santiago Martinez Lage said about the lack of incentives for private enforcement in the European jurisdictions when compared to the US system. Considering what Eleanor Fox also said about private rights, I think this is really a subject for the next Workshop. The only point I wanted to make in relation to this issue actually refers to the Chevron doctrine developed in the US and recommended to us by Judge Diane Wood. This doctrine implies that courts should have deference for the interpretation given to the rules by agencies specialised in a particular sector. (I am sim-

3 - Courts and Judges

509

plifying it, I know.) On the occasion of an article written on this issue by John Jackson et al. I contested that this would be appropriate in the EU. European courts are there precisely to review the interpretation given to the rules by the specialised agencies. In other words, European courts are, so to speak, guarantees against experts' arrogance (if I may use that word). I am not speaking about the experts of DG COMP, but about those experts who know everything about a given sector and actually contributed to the coming into being of the rules, and who would claim that they have written the rules, so they know best. This is why I think we should certainly not advocate the Chevron doctrine for Europe. • SIMON BISHOP—Dan Goyder referred in his paper to the Premiere League case and expert economist witnesses baffling the judges. As one of those expert witnesses, I would like to offer an economist's view on making economic presentations before the court. I think that, if we are genuinely moving towards a more economic effect-based approach in enforcement, then the economic experts' submissions will get more and more sophisticated over time. We already saw this happen at the level of Commission decisions. Ten years ago economic arguments were relatively simplistic. Today, econometric evidence is presented at all hearings. Unfortunately, economics involves complex issues: game theory, all these arcane economics terms, and, horror of horrors, numbers, statistics, and econometrics. I know how much lawyers 'love' numbers and statistics, but these are the tools of economic analysis. If you want an economic effects-based approach you have to face up to these kinds of arguments. Certainly, the economists must present this evidence to the court in a clear and concise way. Yet, in some cases, when you have to go so far down into the basics of economics—the only analogy I can find now is with giving expert testimony in German to a non-German speaking judge and having to explain the difference between der, die and das—your evidence will end up by having no weight. So, it seems to me you have really only two choices. One is 'do-nothing', implying that we will have an economic effects-based approach in name only, because decisions would not really be based at all on economic analysis—a better name for this approach in enforcement would be the 'lottery approach'. The other option is, obviously, 'do something about it'. And in this context it is interesting to hear about the US experience with the training of judges on economic issues and the possibility that US judges have to appoint their your own economist experts. Just to be clear: this is not an attack on courts. Some of these issues also apply to the competition authorities themselves, and I think it is laudable that the UK Office of Fair Trading and the Dutch competition authority have actually instituted courses of economic training for their staff—not just for the economists, but for the lawyers as well. The Commission will also set up a training course towards the end of this year.

510

Panel Discussion

Let me end by trying to take a leaf out of Mario Siragusa's book and be controversial by referring to the words of a US judge, which go something like this: 'The lawyer who does not know economics is apt to become an enemy of the people.' • CARL BAUDENBACHER—First, I would like to go back to the arbitration issue. I agree with Judge Wood that no one involved in a hard core cartel would conclude an arbitration clause. Yet, being from Switzerland, which is an important arbitration place by tradition, I can tell you that many important agreements have been concluded at a Swiss venue in order to benefit from the application of the Swiss law and thereby avoid EC competition law, whose standards these agreements clearly violated. However, in 1992 the Swiss Supreme Court ruled that an arbitration tribunal is basically obliged to take into account the competition provisions of the EC Treaty if, according to rules on conflict of laws, EC law applies, and I think that is remarkable. Obviously, Swiss arbitration courts are not able to refer a case to the Luxembourg courts, even if the Luxembourg courts were disposed to accept them. That leads me to the situation of EC-based arbitration tribunals. Here we have the case law of the European Court of Justice in Nordsee,1 which was addressed by David Edward in his paper. He seems to say that it is for the legislator to look for a different situation if the legislator thinks that it would be appropriate. On the same line of argument, Mr Burrichter has said that, in his view, it would be sufficient that the Commission be asked to clarify the situation of arbitration within the new enforcement system. The case law of the European Court of Justice is only case law, so it could be changed. It is not for a judge of the EFTA Court to give advice here, yet from a functional point of view, it fair to say that there is very little difference between a state court and an arbitration tribunal, at least in certain respects. My second point refers to the language problem. Claus Ehlermann has rightly said that co-operation between national judges will be very difficult because of language barriers. I would add it is not only co-operation that will be blocked, but also the courts of different Member States will rarely read each other's judgments. The EFTA Court has its own experience that is relevant on this point: we are the first court in Europe that adopted a lingua franca, namely, in direct actions we only use English, while in preliminary ruling proceedings we use English and the language of the referring court. This system saves a lot of our resources, and it is a modern solution. There is, however, a dark side to it, since we observed that our judgments are read in the English speaking countries mainly. British judges, for instance, tend to take our judgments as persuasive authority, or even as acte clair, whereas judges from other Member States, even if they read English, would just not be willing to take our jurisprudence into account. This is why I consider it would be worthwhile to make an investment in Case 102/81 Nordsee v Reederei Mond [1982] ECR 1095.

3 - Courts and Judges

511

translation computers. We should not try to get rid of the whole translation system before we have explored all the possibilities of modern technology in this field. I can tell you that I was astonished when I paid a visit to a small administrative court in Switzerland, and saw these Dragon translation computers there. I do not know whofinancedthem, but they were there. The judges told me that their use was far from being perfect but the system could be improved. • CLAUS-DIETER EHLERMANN—The issue of arbitration should be thought about as a separate issue, perhaps even considered from a Community legislative point of view, although I do not know whether Article 83 allows to deal with it. Another issue is whether something can be done specifically for the competition field, or whether it is politically wise to do that. However, in the commercial world arbitration is probably as important, if not more important than state jurisdictions. It also happens that arbitrators are well paid and therefore deliver quality judgments, and this could have a positive impact in terms of the interpretation of Articles 81 and 82 • DIANE WOOD—Actually, Claus Ehlermann touched on the point that I wanted to make: we seem to have been discussing two rather different kinds of problems. One set of problems, is directly related to the sphere of competition law, such as: what is Article 81 (1), what is in Article 81 (3), can national courts apply it, etc. Then a number of other issues came up, such as arbitration, forum shopping, the issue of mechanisms to ensure consistency in the interpretation of EU law, which go well beyond the sphere of competition law. For instance, the consistency question raises in my mind the thought that perhaps the system of references for preliminary rulings under Article 234 worked well when there were only six Member States and there was a broader scope for the interpretation of EC law. I do not know whether this would still be the case ten years from now, when the EU will have up to twenty-six Member States and a much more comprehensive law system. Personally, I will watch with interest to see if the time ever arrives when the tensions created by different interpretations of Community law by national courts will lead to the replacement of the preliminary ruling system by a system in which the European Court of Justice will have some form of direct review responsibility over the decisions of national courts. This problem is by no means unique to competition law. As to arbitration, the risk of forum shopping, and so on, one must decide which of these issues needs to be dealt with during this round of reform and which should simply be left for a later stage. One way to put the brakes on the reform is to attempt to solve all the problems of the world in one moment. It seems to me that a number of concerns invoked here are not really about competition law enforcement; they are general problems of the judiciary or of the Community institutions. Finally, a quick reply to Jacques Bourgeois: I can assure you that Chevron deference is not about rubber-stamping, but about deciding where the

512

Panel Discussion

boundaries are. There are agencies and agencies, and you will see that certain agencies—which I will not name—do not get much deference, while others command a lot of respect from the courts, because they understand where the legislative boundaries are, they understand what they are supposed to be doing, and that they are acting in a legal enforcement area where reasonable people can come to different conclusions. • CLAUS-DIETER EHLERMANN—Before I give the floor to Mr Alberto Pera, I would like to mention that he is the Secretary General of the Italian Antitrust Authority, which has proved to be so successful, and a considerable part of the merit for this should go to its Secretary General. • ALBERTO PERA—I wanted to raise a few considerations about the issue of coherence, which was the subject of yesterday afternoon's debate. It seems to me that some of the remarks made today by Judge Edward and Judge Cooke put into question the notion of amicus briefs, which is one of the mechanisms proposed by the White Paper for ensuring consistency in the application of EC competition rules by the national courts of the different Member States. I also think that there are some problems with the practicability of the amicus briefs when dealing with hundreds of national courts, and not with Supreme Courts or appeal courts only. Commission guidelines were the second mechanism proposed by the White Paper for ensuring consistency in enforcement, and here as well some objections were raised from the judiciary. Somebody said that the judges agree on the utility of guidelines, provided the judiciary's autonomy of decision is respected. The third mechanism for ensuring coherence proposed by the White Paper was that of concurrent competencies and the possibility for the Commission to over-rule decisions by the national courts. Basically, three instruments for ensuring coherence in enforcement have been put into question. So, the issue is how then to ensure coherence once the application of EC competition law will be decentralised? It therefore seems to me that ensuring coherence under the new enforcement system would anyway be a problem, even if leaving aside the debate on the split between paragraphs 1 and 3 of Article 81. However, I think that the wording of Article 81 (3) is an important cause of concern from this perspective. First, the application of this provision by the judiciary has not yet been tested. Second, the scope of Article 81 (3) should be defined more precisely. Judges might give different interpretations to this provision: some may consider it as pointing only to pure competition considerations, other may consider it requires taking into account, say, employment or industrial policy considerations. Facing this doubt, some judges might decide to refer to the European Court of Justice for a preliminary ruling. Others instead might not find this necessary. This could be a problem, since in the Community we do not have the same safety mechanisms for ensuring coherence of private enforcement at the 'federal' level as in

3 - Courts and Judges

513

the US. As far as I know, in the US there are State jurisdictions with State Supreme Courts, and then there is also the federal jurisdiction, starting from first instance level, going through the appeal level, and ending with the Supreme Court. There may be differences of opinion between, say, one Circuit Court of Appeal and another, but there is always the Supreme Court to settle the differences. I therefore think that we should either put all the burden of private enforcement on the Community courts, or open the possibility that the decisions of national courts be appealed to the European Court of Justice, at least in certain cases. In conclusion, I would suggest that 'the train of the reform' should be slowed down—as someone suggested yesterday—in order to allow time to think about these problems.

• ANNE SPIRITUS-DASSESSE—I just wanted to add a last comment on the issue of specialised courts. I think that one of the best instruments in the power of the judge in competition law cases is the direct effect of Article 81. The European Court of Justice has taken the potential of this instrument very far in deciding that national judges can even interpret themselves the provisions of Community Directives to establish whether they can be considered to have direct effect. At any rate, in so far as Article 81 (1) is concerned, there is no problem, it has been decided once and for always by the Court that this provision has direct effect. I think that the existence of specialised courts—to which cases involving competition issues should be referred—would reduce the tendency of civil courts to address on their own, the question of whether the agreement which is the subject to the dispute before them is compatible with Article 81 (1). When a judge has a crisis situation before him and has to deal quickly with the case, he would perhaps decide to go on with the case if there is not enough certainty that the agreement is invalid under Article 81 (1). I think, that if we want to take forward the European integration and promote competition on the single market, we should rather emphasise the role of the judge and give him the power to address the legal question on his own In most cases before courts the competition issues arise in defense, and in such cases eight times out of ten it is the judge to raise the Article 81 (1) question on his own. My second point goes back to something that I tried to explain yesterday in relation to Article 81 (2). As Judge Edward said, under the existent notification and prior authorisation system judges can find themselves in a situation whereby they must cancel contracts infringing Article 81 (1) with respect to the past. In other words, judges cannot do anything other than to acknowledge that the contract is invalid under Article 81 (2) for the past. As to the future, the judges can wait for a Commission decision based on Article 81 (3). In an enforcement system such as proposed by the White Paper there will be no more uncertainty in this respect: the issue of the validity of the contract under Article 81 (1) can be put on the table at any time, and it can also be solved at the end by reference to Article 81 (3). I would therefore agree with Judge Groning that

514

Panel Discussion

the new system is more in line with Article 81 (1). First, it takes into account that the economic context can change. Second, it is possible that during a litigation between parties A and B a contract may seem valid, though in another litigation, involving party C, other effects of the same contract may come forth, and the court could come to the conclusion that the same contract cannot be declared valid. My last point relates to the issue of whether an administrative authority should review a judgment or not. Here I totally agree with my colleagues from the judiciary around this table, and I would just add some arguments to support our standpoint. First of all, we must take into account the jurisprudence of the Court of Strasbourg—the Treaty of Amsterdam foresees this very clearly. The jurisprudence of the Court of Strasbourg points towards the following two aspects. First, when the Commission, as an administrative body, imposes heavy fines in competition cases, these could be seen as similar to criminal sanctions, and criminal sanctions can be applied only by the judiciary, not by the administration. Second, criminal sanctions must be open for a second degree of jurisdiction, and I think that the jurisprudence of the European courts in Luxembourg confirms this. This point is actually taken over in a number of EC Directives. For example, all EC Directives on financial services provide that the decisions taken by the national supervisory authorities must be subject to appeal to the national courts. I think these are very important reasons why judgments cannot be reviewed by administrative authorities only. • MARIO SIRAGUSA—I was somewhat surprised by Giuliano Marenco's comment in the sense that companies are not interested in clear rules on the allocation of competencies. Let us take the example Judge Edward mentioned before: a company with a selective distribution system relying on similar or identical agreements with various distributors throughout the Member States. Let us assume that in one Member State the distribution agreement is challenged by the national competition authority, which opens an investigation, while in another Member States a competitor of the local distributor is challenging the agreement in front of a court. What would be the effect of the decision of the national competition authority finding that the agreement is infringing EC competition rules? Does this decision have effect in that country only, or also in the rest of the Community? What is the result of this decision in the other Member State, where the competitor challenged the distributor in court? Finally, when would the Commission exercise its power to intervene and take up the case in Brussels—a power that the Commission, rightfully so, intends to retain? I really think we should leave aside this 'we-all-love-each-other' co-operation spirit of the White Paper. Can you imagine the consequences of such a situation as I just described? Powerful companies would run to Brussels and lobby the Commissioners to get the case. Less strong companies, however, will not be in a position to do that. I think this is unacceptable. We must have clear rules as

3 — Courts and Judges

515

to who is competent to do what, which are the cases when the Commission can intervene, what is the effect of a decision by a national competition authority in other Member States. Otherwise, I am afraid that only the most powerful will gain from the new enforcement system, while normal citizens will be the potential victims. This is not only a matter of common culture, which in the competition field is, fortunately, growing fairly rapidly in Europe. It is a matter of power as well, and having a common culture does not solve the issue of power. My second comment refers to the US Chevron doctrine, which is about courts deferring on certain issues to opinion of specialised agencies. I think there is a fairly large number of Member States where the decisions of administrative bodies are reviewed by an administrative judge, and that is exactly what happens. European administrative judges, at least on the continent, have a tradition of applying only a legality review, while being fairly reluctant to review cases on their merits—that is, unlessfindinga manifest mistake of appreciation, a failure of the investigation, or a contradiction in the reasoning, but otherwise, they are very prudent in substituting themselves to the administrative authorities. So, let us not imagine that on the continent we will have real review of administrative decisions on the merits of the case. • JOHN COOKE—I was asked why I distinguished between the notions of 'am/cus curiae' and 'intervener': perhaps here my position is slightly influenced by my common law background. I thought of the Commission's—or the national competition authorities'—right to intervene during court proceeding more as an obligation, and not so much as a possibility. I quite appreciate the utility of the amicus briefs in US antitrust litigation, but we must remember that the Commission has two 'hats': the legislative, or policy-making hat, and the enforcement hat. If the Commission's only interest in a particular private litigation is to inform the court of the policy context it wants to be taken into consideration, then an amicus brief is fine. Yet if, a la Giuliano Marenco, the interest of the Commission is an enforcement interest, that is, the Commission regards a specific agreement as being of some significance and wants to take a stand on the litigation the subject of which this agreement is, so that it comes into the case with a view on the facts, and possibly with an expert testimony to offer as to market share, reaction of other competitors in the marketplace etc., then I think it is important from the point of view of the national court that the Commission has the status of a party in the proceedings, so that the other parties in the litigation are entitled to put its intervention up to scrutiny, both on facts and on expertise. • DAVID EDWARD—This is the distinction between intervention and direct action before the European Court of Justice: the intervener intervenes to support one of the parties, while the Commission submits observations, but not as a party.

516

Panel Discussion

• EMIL PAULIS—Above all, I personally noticed a few general issues coming up in this debate. Thefirstone is that there may be some general problems in our judicial system, and in our administrative decision-making system perhaps, which are highlighted in this debate on the reform of EC competition law enforcement, but which are not specific to it. I do not think that we should make it a condition that all these problems be resolved before we can go ahead with this reform, because they may not only hold up this reform, but other reforms as well. I will also take with me after this meeting that there is a strong demand for clarity, and a strong demand on the Commission not to leave open in the new Regulation issues that would then have to be settled by the judges in Luxembourg. There is an equally strong demand for coherence. This means, on the one hand, uniform application of the law and legal certainty for all operators; and on the other hand, independence for all the decision-makers in the enforcement system, and particularly for national courts. There is of course a tension here: the more independence you leave with respect to competencies, the more this independence may conflict with the aim of coherent application of the law and legal certainty. These are conflicting claims, and we will probably have to find a middle way. I would also like to add, in relation to the previous point, that there is—as Judge Edward rightly said—a distinction to be made between the Commission's decisions and its 'soft law'—guidelines, etc.—even if the Commission's decisions are of a declaratory nature. The Commission's decisions have binding effect on their addressees, but this effect has, of course, to be implemented. The Commission has no power to enforce its decisions, so that their effects must be taken into account by the national courts. The latter must, of course, respect the Commission's decisions, otherwise they would put into question a Community act that can only be reviewed by the European Court of Justice. The European Court of Justice itself developed a jurisprudence concerning the effects of Commission decisions, based precisely on considerations of uniform application of the Community law and legal certainty. I also think that the argument of separation of powers may be fallacious. Jacques Bourgeois basically said that because administrative decisions can be reviewed by the judges they can never really have any effect on the judges. This is not the point here, because it is not up to the national courts to review the Commission's decisions, but the European Court of Justice. The issue of separation of powers arises in each legal system, but it does not play in the same way between different systems, as is the case here. However, we would probably have to continue the debate on this. A few remarks on the interpretation of Article 81 (3): after what has been said here, my personal opinion—though not everybody from the Commission might agree with me—is that the scope of Article 81 (3) must be first clarified by the Commission. The Commission should come up with a set of guidelines addressing specifically the interpretation and application of Article 81 (3), and then the European Court of Justice could of course decide whether we were right or not in our interpretation. I think though that this would be a start at least, rather than to wait for the ECJ to elaborate on everything anew.

3 - Courts and Judges

517

A short remark on the issue of 'other objectives' in Article 81 (3): these 'other objectives' are laid down elsewhere in the Treaty and do not refer specifically to Article 81 (3). General objectives, such as the protection of environment, can also for instance arise under Article 82. Take an agreement serving to promote environmental protection which involves a whole industry, and which could lead to a kind of collective dominance. Does Article 82 apply or not? So, this issue is not specific to Article 81 (3) only, but one that we have to solve in a general sense. On the issue of the burden of proof: I would like to correct a misunderstanding arising in Professor Mestmacker's paper. He apparently starts from the assumption that, by introducing a legal exemption system, we would introduce a kind of general presumption of legality under 81 (3), or we would remove what would be a general presumption of illegality under 81 (1). Neither of these is true. There is no presumption of illegality under Article 81 (1), one has to prove in casu whether this provision applies. Article 81 (3) is a rule of exemption, but by turning this rule of exemption into a directly applicable legal exemption one does not introduce a general presumption of compatibility. The legal exemption applies only if it is proven that the conditions of 81 (3) are fulfilled. I would also like to comment on the issue of the burden of proof as put by Judge Edward in his paper and during our discussion. He said that in the future enforcement system the burden of proof will depend on what approach we take in the interpretation of 81, in other words, if we would go for a unitary approach or a bifurcated approach of Article 81 (1) and Article 81 (3). I think we cannot re-write the Treaty. There are two rules: there is the rule of prohibition and there is the rule of the exemption. However, the legal exemption system will bring about an element of novelty in this respect in the sense that it will create a link between Article 81 (1) and (3) in a direct effect context. This means that if the two rules stay separate—and they have to stay separate—their balancing will require tofirstidentify the negative effects of an agreement and then identify its positive effects. These two separate rules need to be balanced, and therefore we should, in my view, regulate the issue of burden of proof. In my view, we should specify that for Article 81 (1) the burden of proof is, of course, on the party invoking this provision, and for Article 81 (3) the burden of proof is on the party who invokes the exemption. A very personal remark on Article 234: there are a number of problems with the system of references for preliminary rulings, in particular the European Court of Justice's workload. Again, this is a general problem, it is not specific to the area of competition law only. There are, however, two things I think should be said about this problem. First of all, the problem of delays should not be exaggerated, because there is no vacuum. A national judge can adopt a preliminary measure pending the reply of the European Court of Justice. Second, and this is again a very personal view, if the European Court of Justice decided that arbitrators must observe EC competition rules, then it would seem logical to me that arbitrators should also be given the possibility to make

518

Panel Discussion

references for preliminary rulings when they have legal interpretation queries. However, this would again increase the number of references to the European Court of Justice under Article 234. • CLAUS-DIETER EHLERMANN—Just two short remarks. The first is that there are limits to legal interpretation and to the questions that arbitrators can bring before the Courts in Luxembourg. In my view, sometimes such questions are beyond what interpretation allows, even if the Courts are very generous with wording. Then again, I am referring to experience at the Appellate Body of the WTO, which is slightly different from the case of the Courts in Luxembourg. Second, I think that everybody around this table assumes that the Delimitis principle of 'deference' of national courts vis-a-vis the Commission is, and will remain in place.8 Yet, if the Commission's monopoly over the application of Article 81 (3) is abandoned, I wonder whether that deference principle, particularly from the point of view of separation of powers, remains in place. • ALEXANDER SCHAUB—Just one additional remark, to avoid a misunderstanding: Mario Siragusa got very agitated about the issue of the allocation of cases, and insisted that this was a matter of power, not of competition culture. The Commission does not consider this neither a matter of power nor one of culture, but, in the first place, one of clear rules, as we already clarified in the non-paper on the network of competition authorities. We need clear rules on the allocation of cases, though, of course, we need rules with certain flexibility. Such flexibility exists even at the present: take for instance the area of merger control, where the Commission can refer cases to national competition authorities or it can refuse to do so. I wanted to emphasise this once more: this is not a matter of power. We are exposed to the worst pressures, and I could tell you long stories about that, but we have long lists of examples making it obvious that we are not giving in to the pressure, and that we have always made our own judgment as to whether we should decide a case in Brussels or refer it to a national competition authority. There is this one little pending case, concerning an electricity merger in Germany, about which you may have heard, and there were important and powerful people who would have preferred that this case be decided in Germany rather than in Brussels. And not for reasons of power and not for reasons of culture, but for reasons of judgment on the basis of existing rules, the Commission kept this case peacefully, as in many other cases of this sort. We kept it with us, and we intend to continue on the same line once the new enforcement rules will be in place.

8

Case 234//98 Delimitis v Henninger Brdu [1991] ECR 1-935.

PANEL THREE COURTS AND JUDGES

WORKING PAPERS

Carl Baudenbacher The European Commission's Reform Proposals Seen from the Point of View of the EEA/EFTA Countries, the EFTA Surveillance Authority and the EFTA Court

I. Some observations on the European Economic Area1 The EEA Agreement that entered into force on 1 January 1994 links the participating EFTA States to the Community's internal market in the closest possible way. However, as far as institutions are concerned, the Agreement is based on a two-pillar approach under which each pillar (the European Community and EFTA) has its own institutions. The EEA Agreement is therefore the only association treaty ever concluded by the European Community that allows the associated states to have their own surveillance authority and their own court. The EFTA pillar has become rather slim after Switzerland withdrew (as a consequence of a negative referendum) and the accession of Austria, Finland and Sweden to the European Union. It now consists of Iceland, Liechtenstein, and Norway. The EEA nevertheless remains an important tool for the participating EFTA States; it constitutes an intellectual challenge for all the protagonists involved, and some think that a second structure for European integration will one day be needed, beyond the boundaries of the actual EEA/EFTA States.2 The overall goal of the EEA Agreement is the establishment of a dynamic and homogeneous European Economic Area, based on common rules and equal conditions of competition, and providing for adequate means of enforcement including at the judicial level, achieved on the basis of equality and reciprocity, and of an overall balance of benefits, rights and obligations for the Contracting Parties.3 In order to reach this goal, the EEA Agreement has basically adopted the provisions of the EC Treaty on the fundamental freedoms and competition law, as well as large parts of the Community's harmonised law. Since EEA law is therefore largely identical to EC law in substance, the EEA Agreement's main concern is to secure a homogeneous development of the law in both pillars: At the legislative level, Article 102 EEA provides for a continuous adoption of new EC legislation into the EFTA pillar of the EEA by the 1

See, concerning the EEA Agreement, Norberg et al. (1993) with a foreword by Claus-Dieter Ehlermann; Blanchet et al. (1994); Muller-GrafT, P.-C. and Selvig, E. (1999). 2 See, for instance, Kapteyn and Van Themaat (1998); HSH Prince Nikolaus von Liechtenstein (2000). 3 Recital 4 of the preamble to the EEA Agreement; see also Article 1(1) EEA.

522

Working Paper I

Joint EEA Committee. The EEA/EFTA states are involved in relevant EC legislation through co-operation rights in the decision-shaping phase, and through a right to veto the adoption of EC legislation into the EFTA pillar. Article 6 EEA obliges the EFTA Surveillance Authority (ESA) and the EFTA Court to basically follow relevant EC Court of Justice (ECJ) precedents that have been rendered prior to the date of signature of the Agreement (2 May 1992). Also, under Article 3 (2) of the 'Agreement Between the EFTA States on the Establishment of a Surveillance Authority and a Court of Justice', the ESA and the EFTA Court are bound to pay due account to relevant ECJ precedents given after the signing of the Agreement. The EFTA Court has, in its very first judgment—in the case Restamark—held that its obligation to follow and to take into due account ECJ precedents refers also to judgments of the EC Court of First Instance (CFI).4 On the other hand, the ECJ and the CFI have made it clear on a couple of occasions when basing themselves on EFTA Court precedents that homogeneity is not a one-way street.5 Although the EFTA Court had, in most cases, to deal with legal issues that had not previously been decided (or at least not fully addressed) by the EC courts, there has not been a judicial conflict in the first six years. Observers agree that the EEA Agreement has functioned well in its first six years.6 The biggest challenge for the EEA is that the partner that the EFTA countries chose to dance with, the European Community, is moving faster and faster. The EFTA countries were in 1992 probably willing to go for a slow foxtrot, but now they find themselves linked to a dancer that, at least from their perspective, is ready to rock and roll. The crucial problem arises from the fact that negotiations on amendments to the EEA Main Agreement hardly seem to be feasible any more. In this situation, there are basically two options for the EFTA pillar to resolve problems arising from the EC pillar's move forward: amendment of EEA Annexes and Protocols and dynamic interpretation of existing EEA law.

II. The current EEA competition law regime 1. Substantive law The competition rules of the EEA Agreement are the result of mirror-legislation. Articles 53 and 54 EEA copy for EEA purposes Articles 81 and 82 EC. 4

1994/95 Rep. EFTA Ct., 15. See Case C-l 3/95 Stizen [1997] ECR 1-1259; Joined Cases 34/95 and 36/95 KO v de Agostini and TV Shop i Sverige [1997] ECR 1-3843; Case C-140/97 Rechberger [1999] ECR 1-3499; Case T-l 15/94 Opel Austria v Council [\997] ECR 11-39. 6 See Sevon and Johansson (1999); Forman (1999); Editorial comments (ECLR 1999); Norberg (2000). 5

3 — Courts and Judges

523

2. Competencies In accordance with Article 55 EEA, the Commission and the ESA shall ensure the application of the principles laid down in Articles 53 and 54 EEA. Article 55 reads as follows: 1. Without prejudice to the provisions giving effect to Articles 53 and 54 as contained in Protocol 21 and Annex XIV of this Agreement, the EC Commission and the EFTA Surveillance Authority . . . shall ensure the application of the principles laid down in Articles 53 and 54. The competent surveillance authority, as provided for in Article 56, shall investigate cases of suspected infringement of these principles, on its own initiative, or on application by a State within the respective territory or by the other surveillance authority. The competent surveillance authority shall carry out these investigations in co-operation with the competent national authorities in the respective territory and in co-operation with the other surveillance authority, which shall give it its assistance in accordance with its internal rules. If it finds that there has been an infringement, it shall propose appropriate measures to bring it to an end. 2. If the infringement is not brought to an end, the competent surveillance authority shall record such infringement of the principles in a reasoned decision. The competent surveillance authority may publish its decision and authorise States within the respective territory to take the measures, the conditions and details of which it shall determine, needed to remedy the situation. It may also'request the other surveillance authority to authorise States within the respective territory to take such measures.

Article 56 EEA establishes a 'one-stop shop' system, whereby either the Commission or the ESA has jurisdiction in any individual case under Articles 53 and 54 EEA. The provision states: ' 1. Individual cases falling under Article 53 shall be decided upon by the surveillance authorities in accordance with the following provisions: (a) individual cases where only trade between EFTA States is affected shall be decided upon by the EFTA Surveillance Authority; (b) without prejudice to subparagraph (c), the EFTA Surveillance Authority decides, as provided for in the provisions set out in Article 58, Protocol 21 and the rules adopted for its implementation, Protocol 23 and Annex XIV, on cases where the turnover of the undertakings concerned in the territory of the EFTA States equals 33 per cent or more of their turnover in the territory covered by this Agreement; (c) the EC Commission decides on the other cases as well as on cases under (b) where trade between EC Member States is affected, taking into account the provisions set out in Article 58, Protocol 21, Protocol 23 and Annex XIV. 2. Individual cases falling under Article 54 shall be decided upon by the surveillance authority in the territory of which a dominant position is found to exist. The rules set out in paragraph 1 (b) and (c) shall apply only if dominance exists within the territories of both surveillance authorities. 3. Individual cases falling under subparagraph (c) of paragraph 1, whose effects on

524

Working Paper I

trade between EC Member States or on competition within the Community are not appreciable, shall be decided upon by the EFTA Surveillance Authority. The terms 'undertaking' and 'turnover' are, for the purposes of this Article, defined in Protocol 22.'

It is clear that under Article 56 EEA the Commission will rather often, but not always, have jurisdiction. It must be noticed, however, that in cases under Article 56 (1) (b) and (c), the non-competent authority is nevertheless involved through co-operation rules. Overall, EEA competition law can be described as a two-pillar system based on co-operation between the Commission and the ESA. Article 58 EEA states: 'With a view to developing and maintaining a uniform surveillance throughout the European Economic Area in the field of competition and to promoting a homogeneous implementation, application and interpretation of the provisions of this Agreement to this end, the competent authorities shall co-operate in accordance with the provisions set out in Protocols 23 and 24.'

3. Enforcement 3.1. Surveillance authorities As far as enforcement is concerned, the EEA Agreement's competition law regime is based on the traditional model of centralised implementation. Pursuant to Article 1 of Protocol 21 EEA, the ESA has been entrusted with powers equivalent, and functions similar, to those of the Commission for the application of the Community competition rules. Article 3 of Protocol 21 EEA contains a list of relevant acts, including Regulation 17/62. The ESA's powers to enforce the competition rules of the EEA Agreement are laid down in the 'Agreement between the EFTA States on the Establishment of a Surveillance Authority and a Court of Justice'. Protocol 4 to this Agreement contains provisions on the functions and powers of the ESA in the field of competition. Chapter II of Protocol 4 lays down provisions mirroring Regulation 17/62. 3.2. National competition authorities The national competition authorities in Iceland and Norway have not been given the competence to apply Articles 53 (1) and 54 EEA so far. Liechtenstein does not have a national competition authority yet, but the question of establishing such an agency is under consideration. 3.3. National courts The question of the direct effect of provisions of EEA law is left open in the EEA Agreement. The Nordic countries adhere by tradition to the dualistic

3 - Courts and Judges

525

approach in international law. In the case of Articles 53 and 54 EEA, this has not posed practical problems since the EEA Main Agreement has been incorporated into national laws. Furthermore, the EFTA Court ruled in Sveinbjornsdottir7 that the EEA Agreement has created a new legal order of its own that is to be distinguished from the legal orders of the EEA/EFTA States. The extent of integration is less far-reaching than under the EC Treaty, but the scope and objective of the EEA Agreement go beyond what is usual for an agreement under public international law.8 Based on this qualification, the Court ruled that state liability is an integral part of EEA law. The Court did not express itself on whether the concept of direct effect is compatible with EEA law, since no question to that effect was put before it. One will, however, remember that the recognition of a new legal order has formed the basis for the acknowledgement of direct effect in EC law.9 Furthermore, it must not be overlooked that, according to the CFI's Opel Austria judgment,10 the EEA Agreement forms an integral part of the Community legal order and its provisions may have direct effect. A former President of the EFTA Court and now Judge of the European Court of Justice, Leif Sevon, argued in a recent article that 'the application of the principles of direct effect and primacy is . . . simply a question of whether in the EFTA States individuals, to the same extent as in the Community, before national courts should be able to rely on decisions adopted with the agreement of all the Contracting Parties and fulfilling internal constitutional requirements.' n National courts in the EEA/EFTA States do apply Articles 53(1) and 54 EEA.12 According to the comments of the EEA/EFTA States on the White Paper,13 there seem to have been substantially more nullity actions than third party actions pursuant to Articles 53 and 54 EEA so far. Under Article 34 of the Agreement between the EFTA States on the establishment of a Surveillance Authority and a Court of Justice, national courts in the EEA/EFTA States may request the EFTA Court to give preliminary rulings on the interpretation of the EEA Agreement. The procedure is largely analogous to the ECJ's preliminary reference procedure under Article 177 (now Article 234) EC, the main difference being that rulings of the EFTA Court are not formally binding upon the requesting national court. However it is clear that, should the referring court disregard an opinion of the EFTA Court that finds parts of an EFTA State's law incompatible with EEA law, it would 7

1998 Rep. EFTA Ct., 95, in particular para. 59 ff. See, concerning the new legal order already created by the EEA Agreement, Baudenbacher (1997). 9 Van Gend & Loos Case 26/62 [1963] ECR-1. 10 Case Opel Austria as cited above. 1 ' Sevon and Johansson (1999) at 373, 386. 12 See, in this respect, also the EFTA Court's judgment in Opel Norge, 1998 Rep. EFTA Court, 4 ff. 13 EEA EFTA comments on the Commission's White Paper on modernisation of the rules implementing Articles 81 and 82 of the EC Treaty of 30 September 1999, para. 33. 8

526

Working Paper I

amount to a violation of the EEA Agreement by the EFTA State concerned.14 Experience shows that, in reality, there is no difference between the situation in the Community and the EEA.

III. Implications of the system proposed in the White Paper for the EEA15 The comments submitted to the Commission by the EEA/EFTA States16 and the ESA17 are basically favourable to the idea of switching from a system of ex ante control to a system of ex post control. From the viewpoint of homogeneity any new system of application of the competition law in the EC pillar of the EEA should in fact be extended to cover the EEA. The EEA/EFTA States have stated that decentralised application of Articles 53 and 54 EEA is compatible with Articles 55 and 56 and might be achieved by amending Protocol 21 to the Agreement. Amendments to the Main Part of the Agreement would thus not seem necessary.18 1. The principle 1.1. Abandoning the notification and authorisation system The EEA/EFTA States and the ESA share the Commission's view that the enforcement of competition rules needs to be strengthened, and they support the Commission's proposal that the reform should be based on a directly applicable legal exception system, an ending of the notification and authorisation system, decentralised enforcement, and intensified ex post control. The EEA/EFTA States are of the opinion that the involvement of national competition authorities will lead to a more efficient enforcement of the competition rules. It is, however, clear in this context that Iceland and Norway will have to 14 Bull (1994); further see statements of the Norwegian Chief Justice Carsten Smith and of Justice Trond Dolva on the matter. Chief Justice Smith wrote that 'when a national court has requested an opinion, that court would certainly be rather reluctant to disregard that opinion' (Case Law Harmonization, Festskrift Till Stig Stromholm, 1997, 795, 798); and Justice Dolva said that 'the reality is that it is very difficult to foresee that we should not follow the reply of the EFTA Court' (in Sundstrom and Kauppi (1999), at p. 249). 15 See, concerning the White Paper, Schaub (2000); Ehlermann (2000). 16 See supra, footnote 13. 17 Comments of the EFTA Surveillance Authority to the European Commission's White Paper on modernisation of the rules implementing Articles 81 and 82 of the EC Treaty of 30 September 1999. 18 See supra, footnote 13, Executive summary.

3 - Courts and Judges

527

enact national legislation enabling the national competition authorities to apply the EEA competition rules. The EFTA side of the EEA did not have a chance to comment on a new proposal presented by the President of the German Federal Cartel Office on 27/28 April 2000 at the Seventh St. Gallen International Competition Law Forum.19 The Federal Cartel Office fears that competition policy will on balance be weakened with the abolition of the notification system. The Office does not overlook the fact that hard-core cartels are never 'notified'. But it is of the opinion that there is a grey zone between hard-core cartels on the one side and nonsuspicious co-operation on the other, where an obligation to inform the Commission or a national competition authority should still be required. The information received would have to be published on the Internet. Noncompliance with this duty would be sanctioned withfines.In the Federal Cartel Office's view, the obligation to inform would enable the competition authorities and private undertakings to enforce the cartel prohibition. The Commission has agreed to examine the new proposal seriously. 1.2. Adopting a more economic approach The EEA/EFTA States are not limiting themselves to agreeing to the planned adoption of a more economic approach to the application of Article 81 (1) EC Treaty/Article 53 (1) EEA, but they are of the opinion that this should be a more explicit part of the reform.20 It may be pointed out in this context that the EFTA Court relied on such an approach in its ruling in the Opel Norge case. The case was, inter alia, about the compatibility of two clauses in a motor vehicle dealership agreement with Article 53 (1) EEA, the provision mirroring Article 81 (1) EC Treaty.21 With regard to the White Paper it must nevertheless be observed that, in light of the differing schools of thought in modern economics, it is not a priori clear what exactly is meant by 'a more economic' approach.

19

B6ge(2000). See supra, footnote 13, para. 13. 21 1998 Rep. EFTA Court, 4ff.—The first clause imposed conditions on dealers regarding a certain shareholder structure, in casu that a certain individual was to own at least 51 % of the shares. The second clause prevented the shareholders in the dealer company from holding ownership interests in other companies dealing in motor vehicles. In the Court's view, the object and effect of the second clause was to prevent the dealer from selling vehicles of other makes. It was therefore capable of being restrictive of competition. With respect to the first clause, the Court held that it might, depending on the circumstances, not be restrictive of competition since it was able to strengthen the personal bond between the parties of a motor vehicle dealership agreement. A possible negative impact on competition could be outweighed by the pro-competitive effects of the clause. Read in the context of the whole agreement, however, the ownership structure provision was capable of intensifying the restrictive effects of the group clause. 20

528

Working Paper I

1.3. Exception: prior authorisation requirement for partial-function production joint ventures With regard to the Commission's proposal that partial-function production joint ventures should still be subject to a prior authorisation requirement, since such operations generally require substantial investment and far-reaching integration of operations, the EEA/EFTA States agree that a prior authorisation system seems appropriate for some operations. They emphasise, however, that substantial investment 'does not, in itself, justify a prior authorisation system. . . . Operations only involving substantial investments, without far-reaching integration, should accordingly be subject to the ordinary regime of directly applicable exceptions.'22 2. Decentralised application by national competition authorities 2.1.

Co-operation between surveillance authorities and national competition authorities

The EEA/EFTA States consider 'that the special role of the surveillance authorities in the application of Community/EEA competition law must be exercised together with and in close co-operation with the national competition authorities'. They agree that, under the new system, block exemptions will still be an important tool, and that regularly updated Notices and Guidelines will continue to be of importance.23 The ESA points out that clear rules on the allocation of cases are needed.24 In fact, the White Paper leaves open too many questions in this respect. 2.2. Positive (non-infringement) decisions The EEA/EFTA States are of the opinion that the Commission should further prove the need for positive decisions. If the competent surveillance authority considers the conditions undertaken by the parties to an agreement sufficient to overcome its objections to the agreement, it would terminate the infringement procedure by adopting a 'positive decision'. The surveillance authorities would probably have to show prima facie that an infringement has existed before agreeing to such conditions, 'even if such [positive] decisions do not state that prior to the agreement there has been an infringement of Article 81 or 82 EC and/or Article 53 or 54 EEA, and the addressees of the decision would not be obliged to accept that an infringement had existed'. In the view of the EEA/EFTA States, there is a risk 'that the surveillance authorities might adopt 22 23 24

See supra, footnote 13, para. 18. See supra, footnote 13, para. 20/21. See supra, footnore 17, p. 10.

3 - Courts and Judges

529

such decisions even if they have not been able to show prima facie that there has been an infringement. The companies may accept such undertakings in order to avoid costly procedures.'25 2.3. Power of all 18 EEA States to enforce Articles 81182 EC and 53154 EEA The EEA/EFTA States point out that, in view of the importance of decentralised enforcement, national authorities of all EU Member States and EEA/EFTA States should have the power to enforce Articles 81 and 82 EC and/or Articles 53 and 54 EEA.26 3. Network of competition authorities 3.1. General With respect to the development of the necessary climate of co-operation between the authorities, the EEA/EFTA States have certain reservations as to whether the Advisory Committee is the proper forum for such practical enforcement co-operation, and they point to the initiative taken by the Nordic competition authorities to improve the enforcement co-operation between them.27 3.2. Harmonising procedural legislation Undertakings in the EEA/EFTA States ask the Commission to consider the need for harmonising national procedural law if national competition authorities are to enforce Articles 53 and 54 EEA.28 3.3. Obligation of national competition authorities to inform the surveillance authorities The EEA/EFTA States are of the view that the national competition authorities should be obliged to inform the surveillance authorities before closing a file, as well as of plans to withdraw the benefit of a block exemption. They are, however, less convinced that there is a need for the surveillance authorities to be informed of the initiation of procedures. In view of the need to safeguard full confidentiality in the early stages of hard-core cartel investigations, it might be sufficient to give the national competition authority a right to inform the 25 26 27 28

Supra, Supra, Supra, Supra,

footnote footnote footnote footnote

13, para. 24. 13, para. 26. 13, para 31. 13, para. 5.

530

Working Paper I

surveillance authorities or another national competition authority at the initiation or at any other stage of an investigation. Imposing a duty to do so could be even counterproductive if it gives the national competition authority an incentive to apply national rather than Community/EEA competition rules.29 In a similar way to that proposed with regard to the Community, national competition authorities of the EEA/EFTA States and EU Member States should, in the view of the EEA/EFTA States, be obliged to inform either the Commission or the ESA of cases to be dealt with under Articles 53 and 54 EEA. The EEA/EFTA States propose that EU States inform the Commission whereas EEA/EFTA States inform the ESA.30 3.4. Right of the Commission to withdraw a case from a national authority^1 The EEA/EFTA States take the view that the right of the Commission to take a case out of the competence of a national competition authority should be maintained. This should, however, occur only if uniform application is seriously endangered. The ESA is basically of the same opinion. In EEA cases, both surveillance authorities should have that power. However, according to the one-stop-shop principle, only one of the surveillance authorities should have the power in any individual case. Since Article 56 EEA already addresses the allocation of cases between the Commission and the ESA, that provision might also be applicable to the question at hand. 3.5. Corrective decisions The ESA has difficulties with the Commission's proposal concerning the adoption of judgments correcting decisions by national authorities, arguing that the emphasis should rather be on preventive measures such as the right to withdraw a case or to give advice to the national authority. Corrective measures should primarily be a task for the European courts.32 4. Application by national courts 4.1. Are courts able to do the job? There has been ample debate on whether national courts are able to directly apply Articles 81(3) EC/53(1) EEA, whether they have the knowledge to do so,

29 30 31 32

Supra, footnote 13, para. 36/37 See supra, footnote 13, para. 62. See supra, footnote 13, para. 38, 61; footnote 17, p. 10. Supra, footnote 17, p. 10.

3 - Courts and Judges

531

whether this can be done under the rules of civil procedure, etc. The following may be said on this: • There cannot be any reasonable doubt that judges are basically capable of doing the job. In this regard, there is no difference between competition law and other areas of the law. There are good judges everywhere and there are bad judges everywhere. • Antitrust is not the only area of the law that poses complex problems and requires knowledge of fields other than the law itself. Other areas are, for instance, tax law, patent law, (industrial) copyright law, and social security law. • The arguments put forward against entrusting courts with the application of provisions that presuppose a certain degree of economic knowledge are not new. They have been used in other parts of economic law. In the seventies and early eighties, there was a broad discussion in the German-speaking countries of Europe on whether unfair competition law should be interpreted in a more economic way.33 Unfair competition law is, rightly understood, very close to antitrust law. In fact, practices such as predatory pricing, abuse of buying power, the giving away of goods in masses, and interference with other parties' contracts may constitute unfair competition under certain conditions because they may distort competition. • One sometimes has the impression that the competition law community is in danger of doing too much navel-gazing. This might have to do with the fact that a lot of money is involved and that antitrust law is some kind of a secret science known only to a small group of highly specialised people and is traditionally not taught at universities. • Reading all the objections that are raised against entrusting courts with tasks of a certain degree of complexity, one might have the impression that all this is based on a tacit assumption that the protagonists that have handled the cases so far are almost impeccable. This is obviously not the case. Competition authorities make mistakes, and even European courts make mistakes. It is probably appropriate to recall Justice Robert H. Jackson's famous statement that 'reversal by a higher court is not proof that justice is thereby better done. There is no doubt that if there were a super-Supreme Court, a substantial proportion of our reversals of state courts would also be reversed. We are not final because we are infallible, but we are infallible only because we arefinaF.34

Having said this, one will, of course, concede that courts do not have resources comparable to those of competition authorities; that there are differences between civil courts dealing with actions for nullity of contracts or for damages and administrative courts hearing appeals against decisions of competition 33

See, for instance, the controversy between Baudenbacher (1980) and (1981), o n the one hand, and Kraft (1980), on the other. 34 Concurring opinion in Brown v Allen, 344 U.S. 443 at 540 (1953).

532

Working Paper I

authorities; that, in general, civil procedure is not comparable to the administrative procedure before a competition authority, and that many judges do not actually have the education or experience that would enable them to immediately deal with European competition law cases. That is why the proposals to give the surveillance authorities the right to intervene as amicus curiae and to provide national courts with an advocate-general should be seriously discussed. The same holds true for considerations about creating specialised courts. Education for judges is, of course, always an option. Finally, one must not overlook that national judges already have the right of negative forecasting. They may state whether there is a realistic chance to obtain an exemption from the Commission under Article 81(3) even in cases where notification has taken place.35 We should nevertheless have no illusions: some forum shopping will be unavoidable since national judges will compete with each other. Legal certainty will be reduced and uniform application of the law will, to some extent, be a problem. However, these are not reasons to turn down the whole project. First of all, we are used to some degree of forum shopping in our national jurisdictions. Second, legal certainty is a crucial point but it is probably not the most important one. As far as uniform application of the law is concerned, one sometimes has the feeling that Europeans tend to exaggerate the significance of issues such as harmonisation, coherence and homogeneity at the expense of decisions being made close to citizens and economic operators. Quite another matter is whether private enforcement will become as important as it is in the United States. In light of the differences between the law and legal culture in the United States and in Europe, the answer will be 'no'. The contribution of national courts to the overall enforcement of EC and EEA competition law will, in all probability, not be overwhelming. This means that the main responsibility will be with the national competition authorities, the surveillance authorities, and the European courts. However, the possibility that courts might become more important in the long run is not excluded. In this context it might be of interest that the EEA/EFTA States encourage the Commission to study possible obstacles to private actions in more detail.36 4.2. Obligation of national courts to inform the surveillance authority and surveillance authority's right to intervene as amicus curiae With respect to the proposed obligation of national courts to inform the surveillance authorities of proceedings involving Community and/or EEA competition rules, and the surveillance authorities' right to intervene as an amicus curiae, the EEA/EFTA States are of the opinion that these questions should be analysed in more detail. The details should, in any case, be left to the Member 35 36

See Schroter (2000). Supra, footnote 13, para. 33.

3 - Courts and Judges

533

States in accordance with their national procedural rules.37 If, however, amicus curiae solutions should be adopted, the EEA/EFTA States submit that the Commission and the ESA should have the right to make such interventions in court proceedings concerning EEA competition rules in the Community and the EEA/EFTA States. Unless this can be ensured, only the ESA should have the power to intervene in proceedings concerning EEA competition rules before the national courts of the EFTA States.38 4.3. Power of courts to ask the Commission for information of a procedural, legal, or economic nature The EEA/EFTA States do not support the rules set out in the Commission Notice on co-operation between the Commission and the national courts, according to which courts may ask the Commission for information of a procedural, legal, or economic nature. They do not see the need for such a provision, particularly if the national courts are obliged to inform the surveillance authorities and these authorities have the right to intervene as an amicus curiae. The EEA/EFTA States furthermore point out that national courts can always refer questions to the ECJ or the EFTA Court.39 4.4. Commission's right to take action in cases of disagreeable court decisions Neither the EEA/EFTA States nor the ESA have commented on the Commission's proposed right to prohibit an agreement 'subject only to the principle of res judicata that applies to the dispute between the parties themselves, which has been decided once and for all by the national court'.40 The White Paper does not elaborate on this issue. The question is whether the approach taken by the Commission is compatible with fundamental principles underlying the constitutions of the EC and EEA/EFTA Member States, such as the separation of powers. 5. The role of the European Courts 5.7. European Union Under the current system, the Community courts come into the picture in two ways: First, decisions of the Commission may be appealed to the CFI. Second,

37 38 39 40

Supra, footnote 13, para. 40. Supra, footnote 13, para. 63. Supra, footnote 13, para. 41. White Paper, para. 102(2).

534

Working Paper I

national courts have the right or the obligation to refer questions of European law to the ECJ in certain circumstances. In a decentralised system, the national competition authorities will decide the bulk of the cases. However, there is no action to the CFI for nullification of such decisions. That leads to a question about whether a decentralised model would lead to a weakening of judicial review. The White Paper relies essentially on the preliminary ruling procedure. In fact, decisions by national competition authorities may be appealed to national courts that may seize the ECJ by way of preliminary reference. The Article 234 EC procedure can, however, not be compared to the procedure in a direct action before the CFI, the essential difference being that the ECJ's answer in Article 234 cases will essentially depend on the facts presented by the national court. At this point, the focus has to be on the ongoing reform of the judicial architecture of the European Union. In this context, many commentators have proposed to strengthen the role of the CFI. All that can be said here is that the modernisation of competition policy must take this reform into due account. One has the feeling that the Commission has not done that to the necessary extent so far. 5.2. European Economic Area With regard to the EEA, Article 56 EEA gives the Commission jurisdiction in the majority of the cases. Only in rare circumstances will the EFTA Court therefore have to deal with actions for nullity against decisions of the ESA.41 One will notice in this context that there is no Court of First Instance in EFTA, so that nullity actions against decisions of the ESA are to be brought before the EFTA Court which will hear them on points of law and on points of fact. But national courts have, of course, the right to refer questions for preliminary rulings to the Court. After decentralisation, the latter option would probably be used more frequently than in the past. 6. Intensified ex post control 6.1. Complaints With regard to the goal of reaching an intensified level of ex post control, both by the surveillance authorities and at the national level, the EEA/EFTA States agree that complaints will be more important under the new regime. They support the Commission's proposals regarding measures in order to facilitate the 41 The situation is a different one in the field of State aid; see EFTA Court Scottish Salmon Growers, 1994/95 Rep. EFTA Ct., 59ff.; Husbanken I, 1998 Rep. EFTA Ct., 38; Husbanken II, 1999 Rep. EFTA Ct., 1ff.;Norway v ESA, 1999 Rep. EFTA Ct., 74 ff.

3 - Courts and Judges

535

lodging of complaints, particularly concerning the introduction of a time limit.42 The ESA, for its part, stresses that it must also be possible to rely on informal complaints. With respect to the proposed four-month deadline, the ESA takes the view that it should not include the period that is necessary in order to obtain sufficient information required to make an initial decision.43 6.2. Investigations The EEA/EFTA States agree that there is a need to ensure that investigations be carried out simultaneously and consistently, and to strengthen the guarantees for undertakings during the investigation. They also welcome attempts to increase the surveillance authorities' information-gathering powers. The EEA/EFTA States suggest comparing the proposed provisions with the information-gathering powers of the U.S. antitrust authorities. They draw the Commission's attention also to the fact that, in some of the EEA/EFTA States, national competition authorities are entitled to enter the premises of private individuals, subject to judicial approval, and that criminalfinesmay be imposed on both legal and natural persons for violation of national competition rules. They finally ask the Commission to examine whether the principle of selfincrimination should apply to undertakings and whether the European Human Rights Convention confers such rights upon legal persons.44 From the author's point of view, the investigative powers of the surveillance authorities will have to be strengthened in any case, i.e., independently of whether Regulation 17/62 is reformed in the way the Commission proposes. The reason is simply that more and more undertakings will have learned their lesson, so that it will become more and more difficult for the Commission's civil servants to gather the necessary evidence on the basis of the means provided for in the actual Regulation 17/62. On the other hand, the rights of the defence will have to be improved. 6.3. Judicial review of decisions concerning investigations As regards centralised judicial review, the EEA/EFTA States consider that the EFTA Court might be entrusted with all cases concerning investigations in an EEA/EFTA State. With regard to their argument that, since EFTA Court judgments are not subject to judicial review such a solution would not sufficiently take into account the procedural rights of the undertakings concerned,45 it suffices to say that there is only one court in the EFTA pillar of the EEA for the time being. 42

Supra, footnote 13, para. 42. Supra, footnote 17, p. 12. 44 See supra, footnote 13, para. 44-49; with regard to the investigation powers of the Norwegian competition authority see also Rogne (2000). 45 See supra, footnote 13, para. 65. 43

536

Working Paper I

6.4. Commission's power to summon persons to its premises In the view of the EEA/EFTA States, the proposal that the Commission should be entitled to summon to its premises any person likely to be able to provide information that might be helpful in its inquiries raises the question 'of whether the Commission could be empowered to do so also with respect to persons domiciled within an EEA/EFTA State in cases involving Article 53 or 54 EEA'.46 In that case, the ESA should, in the view of the EEA/EFTA States, likewise be entitled to summon persons resident within an EU Member State to its premises in cases concerning those provisions. Which surveillance authority has the competence in individual cases would follow from Article 56 EEA. The EEA/EFTA States would, however, like to consider further whether such a system would be compatible with the current system under the EEA Agreement that does not give the surveillance authorities the power to conduct investigations in each other's territory. One solution might be to apply the principles laid down in Article 8 (3) (6) of Protocol 23 EEA in such cases. According to these provisions, the competent surveillance authority may request that the other surveillance authority carry out investigations within its territory if the competent authority considers it to be necessary. The competent surveillance authority is entitled to take an active part in the investigations and all information acquired during such an investigation shall be transmitted to the surveillance authority that requested the investigations. Furthermore, the principle laid down in Article 12 of Protocol 23 EEA—namely that undertakings have the right to address and be addressed by the ESA and the Commission in an official language of an EFTA State or of the Community of their choice—should apply.47

7. Co-operation cases in particular The hot potato from a specific EEA perspective is the question of whether decentralisation should also apply to the co-operation cases under Article 56 (1) (b) and (c) EEA. The ESA and the EEA/EFTA States claim that the decentralised application of EEA competition rules could lead to a situation in which the co-operation rights of the EFTA pillar will not be respected.48 From a legal standpoint, the EC side is obliged to ensure co-operation in the cases mentioned in Article 56 (l)(b) and (c) EEA. It would not be legally possible to abolish the EFTA pillar's right to co-operate. It might even been doubted whether decentralisation is legally possible in these cases. If it should be deemed possible, the EC pillar's obligation to co-operate could, in theory, be fulfilled by way of transferring it to the national competition authorities. Technically this could be achieved through an amendment to Protocol 23. But even if it were to 46 47 48

Supra, footnote 13, para. 66. Supra, footnote 13, para. 67, Supra, footnote 13, para 57ff.;footnote 17, p. 7f.

3 - Courts and Judges

537

be deemed legal, EU national authorities (unlike the Commission) would hardly be in a position to ensure that decisions taken under the EEA Agreement can be properly enforced throughout the territory of the EEA. The so-to-speak natural reaction of national authorities could then be renationalisation. That is why the ESA submits that cases that meet the criteria for cooperation under the EEA Agreement should be withdrawn from the national authorities. For the sake of completeness, it may be mentioned that the EEA7EFTA States are of the opinion that the national competition authorities should be able to co-operate horizontally in a decentralised system, that is, across the twopillar structure.49 This is perceived as an advantage for cases involving Norway and Sweden insofar as it might increase efficiency and the speed with which a case can be processed. The ESA does not share this view.50 In fact, this hardly seems compatible with the EEA/EFTA countries being, from the EU's perspective, third countries. As for the rest, it seems that the follow-up problems of this proposal, for instance with regard to the enforcement of common decisions in the other state or with regard to conflicting decisions, have not been fully thought through.

IV. Conclusions The EFTA side of the EEA welcomes the new approach and sees the need to extend it to the EEA. Whether this extension will take place from the very beginning or whether the EU will prefer to introduce and test the new system in the EC pillar of the EEA first, remains to be seen. However, it should be possible to resolve the EEA-specific problems without amending the EEA Main Agreement. In spite of decentralisation, the Commission wishes to retain comprehensive power toward national competition authorities and even national courts in order to ensure a coherent application of the law. In this context, there is a feeling that the Commission does not distinguish, to the necessary extent, between national competition authorities and national courts.51 The image of a judge that is underlying the Commission's White Paper is hardly consistent with the status of common law judges, and probably also not with the role and the self-image of modern civil law judges.52

49 50 51 52

Supra, footnote 13, para. 60. Supra, footnote 17, p. 7 f. See also Mestmacker (1999) at pp. 528-529. See Baudenbacher (1999).

II Jochen Burrichter The Application of Article 81 (3) by National Courts: Some Remarks from the Point of View of a Practitioner

This article discusses the application of Article 81 (3) by national courts, with an emphasis on the reactions of the business community and practising lawyers to the White Paper proposals for a new system of EC antitrust enforcement ('new system'). I will concentrate on the question of whether the national courts are able to apply Article 81 (3) (Section I). I will then briefly discuss the question of forum shopping with regard to national courts (Section II). I will continue with issues of legal certainty (Section III) and coherent application of EC competition law by the national courts (Section IV). Finally, I will try to develop some ideas about how the business community and practitioners could handle the new system in practice (Section V).

I. Are the national courts up to the task of applying Article 81 (3)? 1) Doubts have been expressed about whether the national courts are able to apply the (allegedly) complicated legal notions of Article 81 (3) which also (allegedly) imply policy decisions. These doubts are surprising because the national courts already have the power to apply Article 82. The questions about whether 'market dominance' or an 'abuse' can be established are among the most complicated questions of competition law. Further, they require an economic analysis that is not less complicated than the economic analysis necessary in the assessment of the exemption conditions of Article 81 (3). Also, none of the exemption conditions in Article 81 (3) is more complicated than notions in other legal texts, namely corporate or tax law. Under corporate law, the courts—at least in Germany—are even required to calculate and fix compensations for minority shareholders. In civil law, the courts have proven to be completely prepared to interpret and apply general clauses like § 242 of the German Civil Code which submits the performance of obligations to the rules of 'Treu und Glauben' (good morals). German lawyers know that the courts have developed a whole system of criteria in order to apply this general clause. With regard to the 'policy' aspect, it has to be stated that the criteria of Article 81 (3) do not allow general policy considerations. If such considerations have played a role in the practice of the Commission in the past, it would be

540

Working Paper II

more than an advantage to return to a strict 'rule of law' in the application of Article 81 (3) by the courts. 2) It has also been argued that the national courts are not qualified and sufficiently trained for the application of Article 81 (3). As explained above, the application of Article 81 (3) does not require special legal qualities that go beyond what can be expected from the courts and judges in each Member State. If there are doubts about the qualifications of judges in the application of the legal notions of Article 81 (3) in any Member State, then the Member State in question should immediately start a re-qualification programme in its legal system, and one that is not limited to the judges dealing with questions of Article 81 (3). This does not exclude that a certain specialisation of courts and judges in the field of competition law and the concentration of cases in certain specialised courts should be furthered by the Member States in organising their courts. However, this is not a question limited to competition law but one that also applies to other specialised fields of law. If necessary, it would be a natural task of the national bars to identify quality deficiencies in their national courts that should be remedied as soon as possible. With the background of the experience of the application of EC law by the courts in new Member States in the enlargements of the community in the past, I have no doubt that the courts in the forthcoming new Member States in Middle and Eastern Europe will also cope with the challenge to apply EC law in its entirety, including competition law. 3) It has also been argued that the Court of Justice and the Court of First Instance have granted the Commission discretion in the application of Article 81 (3), a discretion that would not fit into a system of direct application of Article 81 (3) by the national courts. The granting of this discretion, however, has been the consequence of the decision of the drafters of Regulation 17/62 to make the application of Article 81 (3) subject to an administrative procedure whereby the judicial control of the decisions of the administrative body might be limited. The practice of the European Court of Justice and of the Court of First Instance, in my view, was at least partly also a consequence of the sometimes deplorable reluctance of both courts to deal with facts and to go into the difficulties of factfinding and taking evidence, although the procedural rules would allow them to do so. Therefore, the practice of the European Court of Justice and of the Court of First Instance to grant discretion to the Commission in the application of Article 81 (3) is not an argument against the direct application of Article 81 (3) by the national courts.

3 - Courts and Judges

541

4) It has also been submitted that the adversarial procedure in civil law cases is not appropriate to produce just results and to find 'the truth'. Taking into account the fact that most legislation in the Member States leaves decisions in important other legal matters to an adversarial procedure (e.g. decisions in corporate or labour law cases, which also affect public interests to a great extent), this argument is surprising. The argument apparently reveals an unjustified mistrust in the ability of the parties (and their lawyers) to explore and submit the relevant facts, and to develop the right legal arguments. The cases that will be subject to an adversarial procedure are mainly: • disputes between the parties of an agreement about its validity; and • claims against the parties of an alleged infringement of Article 81 by third parties. In these cases it can be expected that the claimants will submit all facts that back their case, that is, facts establishing the conditions of Article 81(1) and/or facts justifying (or not) an exemption under Article 81 (3). It is doubtful whether an ex officio procedure can produce more or better facts taking into account that, under most procedural rules in the Member States, the claimants can refer to witnesses or experts that the court has to hear if the facts for which they are named are relevant for the decision of the case. Furthermore, under an adversarial procedure in civil cases, good judges have a variety of possibilities by which to explore the facts and to point to legal aspects if the parties deal with them insufficiently (for example, by a request to clarify certain submissions). Within the framework of the new system proposed by the White Paper, the courts would have to inform the Commission (and under the relevant national laws probably also the NCAs) about pending proceedings involving questions about the application of Article 81. This obligation should involve the submission of copies of the briefs of the parties to the competition authorities that would thereby be enabled to comment on wrong, misleading, insufficient or inappropriate submissions. The Commission has proposed that it be given the status of an 'amicus curiae', in which case it would have a sufficiently strong procedural position to make sure that the necessary factual and legal issues of the case are not missed. The NCA and/or the Commission, by virtue of the mere possibility of commenting on the submissions of parties to a civil case, would have the possibility of avoiding the outcome that national courts take decisions without at least having been informed about the relevant competition aspects. If the Commission wants to strengthen the private enforcement of Article 81 and to assist third parties in their fact finding for the preparation of private claims against infringements of Article 81, it should consider opening the existing files in fine proceedings to such third parties. Even if a court made a 'wrong decision', the Member States normally grant appeals against such decisions and, in many cases, further appeals (revision). If

542

Working Paper II

none of the parties appeals against a 'wrong' decision, the Commission or the NCAs should be granted the option of opening a prohibition procedure, if necessary. The 'res judicata' should not bar the opening of an administrative procedure. This is because the administrative procedure is justified by the public interest and has a subject different from the civil law case. It goes without saying, however, that no fines can be imposed on the parties in a subsequent procedure either of the Commission or the NCA in cases where a national court has applied Article 81 (3). From the point of view of public interest, it would normally do no harm if the decision of the court were 'wrong' because it is too strict—that is, by denying an exemption under Article 81 (3). If the court 'wrongly' granted an 'exemption' under Article 81 (3), the competition authorities would still have the option of commencing a prohibition procedure. Therefore, I do not believe that the adversarial procedure in civil cases creates a special problem as far as the interpretation of Article 81 (1) and (3) is concerned. In cases of appeals by the parties against a decision of a NCA, the court procedure very likely will be governed by the rules of an ex officio investigation anyhow. 5) It has also been argued that the courts in civil procedures are not well prepared to apply Article 81 (3), because 'exemptions' (in the meaning of the actual procedure according to Regulation 17/62) can only be granted for a limited period of time. This argument seems to misunderstand the new system proposed by the Commission in the White Paper. The 'exemption' would not be 'granted' in the operative part of the decision of the court. The question of whether, for example, an agreement qualifies for the application of Article 81 (3) will be decided implicitly in connection with an evaluation of whether the claim is justified. Of course the court can only decide on the facts submitted to it before the last hearing prior to the decision. If facts change after the decision, the party wanting to overturn the judgment may submit the argument of a change of the facts in a forthcoming new hearing, according to the applicable rules of the national codes of civil procedure. If, in a case pending before it, a court comes to the conclusion that the conditions for the application of Article 81 (3) are actually fulfilled but will probably cease to be fulfilled in the foreseeable future, a prudent judge could mention that in the reasons for decision, thereby giving a clear warning to the parties. In exceptional cases where a clear date can be defined from which the conditions of Article 81 (3) will no longer be fulfilled, the court could even take this into consideration in the operative part of the judgment itself (for example, in a case where a statement of the validity of an agreement is requested by the claimant, the court could state that the agreement is valid until day X only).

3 - Courts and Judges

543

6) It has further been argued that a civil court would not be able to render the exemption under Article 81 (3) conditional on the fulfilment of certain undertakings, as the Commission can do under the present Regulation 17/62. Again, this does not disqualify civil courts from the application of Article 81 (3). If a court came to the conclusion that the conditions of Article 81 (3) would only be met if the parties fulfilled certain conditions, it could communicate this as a preliminary opinion to the parties prior to the final judgment. This is certainly even possible in an adversarial procedure based on facts submitted by the parties. If the parties, according to their agreement, are expressly obliged to use their best efforts to fulfil the conditions for an exemption under Article 81 (3) (which agreements normally should provide), they might even have to jointly agree to amendments to their agreement, and to subsequently submit these amendments to the court. However, this will certainly happen only in exceptional cases. In a case in which it might be an issue that Article 81 (3) is applicable if certain conditions are fulfilled, a good judge will also try to guide the parties to find an amicable solution by pointing them to the possibility of 'curing' deficiencies (for example, by amendments of their agreements). From the point-of-view of public interest, this concern is not an argument against the application of Article 81 (3) by the national courts; if no amendments of the agreement are put before it, the court will apply Article 81(1) with the consequence of invalidating the clause or agreement in question. 7) Some remarks should be added with regard to arbitration procedures that are quite common in cases involving competition issues. Some of the arbitration courts are located in 'neutral' countries that are not Member States. If an arbitration court does not apply Article 81 correctly, the losing party in the enforcement procedure before the ordinary courts normally has the possibility of invoking a violation of ordre public, thereby rendering the award unenforceable. In cases where there is an issue of possible violation of Article 81 (1), but an exemption under Article 81 (3) could not be excluded, arbitration courts generally hesitate to request that parties notify the Commission of the agreement, because this contradicts the main purpose of arbitration procedures, that is, to render fast awards. Also for this reason, arbitration courts avoid dealing with competition law issues by overlooking or denying the applicability of Article 81 (1). Under the new system, arbitration courts would similarly be allowed to apply Article 81 (1) and (3) in its entirety, thereby enabling the ordinary court to exercise its power of control under ordre public aspects.

544

Working Paper II

II. Is there a danger of court forum shopping under the new system? It has been argued that there are so many differences and disparities between the legal systems of the Member States that they could provoke 'forum shopping', especially in cases where certain courts in certain Member States are known to be more lenient or strict than others in the application of Article 81. By subscribing to the Conventions of Brussels and Lugano, the Member States have shown that they do not regard the differences and disparities of their legal systems as being so important; otherwise they would not have agreed to mutually acknowledge the decisions of the courts of the Member States. The Conventions also provide sufficient rules to avoid multiple litigation between the same parties on the same subject matter (Articles 21 and 22 of the Convention of Brussels). Therefore, it does not seem necessary to harmonise the procedural rules of the Member States as a precondition of allowing the direct application of Article 81 (3). It goes without saying that such harmonisation would very likely not take years but decades. The possibility of forum shopping in civil cases is limited, because: 1) In cases involving disputes about the validity of agreements, the parties have only very limited possibilities of choosing between competent courts. First, such agreements normally contain express provisions about the competent courts in case of dispute, as well as about the applicable law. Second, many agreements, namely commercially important agreements, will provide for an arbitration clause with no possibility of forum shopping. Third, even in cases of agreements without forum clauses, the general rules that are commonly applied in the Member States allow only very limited choice about the place of performance of the obligation of the parties concerned. 2) Cases involving complaints by third parties against the participants of an agreement which, in the view of the third party, violates Article 81, will normally be based on tort aspects. In these cases there might be possibilities for forum shopping if the alleged infringement produces damages in different places or if the defendant has acted in different places. In these cases the claimant will be interested in submitting its case to those courts that are more open to the claims, and not to those that are known to be more lenient in the application of Article 81. From the pointof-view of the public interest, this does no harm. 3) In a case where a third party has claimed that the parties to an agreement are violating Article 81, the parties of the agreement could initiate a negative clearance procedure against the third party in order to defend their positions.

3 - Courts and Judges

545

Here, again, the possibilities of forum shopping are very limited because the seat of the third party would be very likely to be the appropriate forum. 4) A party that wants to go 'forum shopping' also has to realise that the enforcement of a foreign court's decision that clearly violates community competition law is endangered if the violation of community competition law is violating the ordre public of the country where the enforcement has to take place. In Germany, the Federal Supreme Court has decided that violations of EC competition law may constitute a violation of the German 'ordre public' (BGH NJW 69/978, 980). 5) Finally, in cases of clear violations of community competition law by a national court, the Commission may initiate a violation procedure according to Article 226 of the Treaty.

III. Does the new system ensure sufficient legal certainty? It has been argued that the new system would not produce sufficient legal certainty in cases involving substantial investments and in which other business decisions cannot be taken without an ex ante administrative clearance of the permissibility of the agreement under Article 81 (1) and (3). As the aspect of legal certainty is dealt with in other papers, I will be very brief on this. 1) Because it is difficult to unravel them after a competition authority has interfered, the White Paper accepts the necessity of ex ante clearance in cases of partial-function production joint ventures that require substantial investments and far reaching integration of operations. The White Paper proposes to apply the Merger Regulation to such partial-function production joint ventures. The critics are right in pointing out that partial-function production joint ventures are only one example of agreements where substantial investments and far reaching business decisions are necessary and require prior clearance. Therefore, the application of the Merger Regulation should not be limited to partial-function production joint ventures but should instead be extended to all other joint ventures requiring substantial investments and far-reaching business decisions. Furthermore, the Commission should consider whether the ex ante clearance under the procedural rules of the Merger Regulation should be extended to all co-operative joint ventures requiring similar investments and far reaching business decisions, even if they do not constitute a 'merger' in the meaning of Article 3 of the Merger Regulation or do not have 'a community dimension' in the meaning of Article 1 of the Merger Regulation.

546

Working Paper II

2) The White Paper provides for the possibility of adopting individual 'positive' (non-prohibition) decisions in exceptional cases on the grounds of a general interest where a transaction raises a new question. It has been requested that this possibility be extended to other cases. It is true that, beyond the cases dealing with joint ventures, there might be exceptional situations in which business decisions of the parties cannot be taken without prior authorisation because they imply important organisational measures (for example, in the cases of airline alliances requiring the organisation of slots). The criterion of general interest does not cover exceptional cases of this kind. In order to avoid inflationary requests for 'positive' decisions of the Commission confirming the applicability of Article 81 (3), the introduction of fees for the mere request for a 'positive' decision should be considered, regardless of whether the Commission finally issues such a decision. 3) The Commission has proposed providing for the possibility that national courts, in the course of proceedings before them, might ask the Commission for information of a procedural, legal or economic nature. It would be logical to extend the possibility of such opinion letters to cases where the parties of an agreement address such questions to the Commission in good faith, namely where new legal or economic issues are concerned. The possibility of such business letters would also have the side-effect of informing the Commission about new developments involving new questions concerning the application of Article 81. In order to avoid the reintroduction of a system of ex ante clearance, it should be made clear that there is no right to receive such business letters and that it is up to the discretion of the Commission whether such letters are granted. To a certain extent, the new regulation should formalise the conditions for such letters and define the effects of such opinion letters. From their nature it follows that opinion letters cannot be binding on national courts or national competition authorities. Again, in order to avoid a flood of requests for opinion letters, the introduction of fees should be considered. 4) It has also been proposed that not only the Commission but also the NCAs should be entitled to deliver opinions. It should be made clear that the formalised business letters should be limited to the Commission. However, NCAs cannot be hindered from informally expressing their views on legal or factual questions concerning the application of Article 81. On these occasions, the information and consultation procedure between the Commission and the NCAs, as provided for in the White Paper, should apply also to business letters of the NCAs. 5) With regard to the predictability aspect of legal certainty, it will be decisive that the Commission issues guidelines that are sufficiently precise and reliable.

3 - Courts and Judges

547

With regard to previous Commission decisions, it has to be emphasised that their guidance value is limited due to the fact that, in its earlier decisions, the Commission extended the scope of Article 81 (1) under the freedom of action test, with the consequence that the application of Article 81 (3) had partly to correct unreasonable results of the extensive application of Article 81 (1).

IV. Does the new system sufficiently guarantee a coherent application of Article 81 by the national courts? It has been questioned whether the new system provides sufficient guarantees for a coherent application of Article 81 by the national courts. 1) The White Paper provides for: • the provision of information to the Commission by national courts in cases implying the application of Article 81; • granting the Commission the status of an amicus curiae • answers to questions by the national courts, concerning procedural, legal or economic questions. These measures should normally be sufficient to exclude the risk that a national court disregards or misunderstands community competition law due to a lack of information. Another measure to safeguard coherent decisions would be to give prohibition or 'positive' exemption decisions of the Commission a binding effect. The binding effect would be covered by the principle of primacy of community decisions. Further, it should be considered that the Commission is entitled to issue prohibition or 'positive' exemption decisions against parties to a judgment of a national court, even it is final and can no longer be appealed. The different nature of the proceedings (administrative v. civil procedure) does not exclude this possibility, although it goes without saying that such decisions should be a rare exception. 2) It has to be added that: • procedures under Article 234 (preliminary questions to the European Court of Justice); and • the possibility of infringement actions against Member States under Article 226, are additional measures to safeguard the coherent application of Article 81. In other areas of community law these provisions have produced satisfying results in the coherent application of community law by the national courts. There is no reason to believe that the provisions will not produce the same effects under the new system. This requires, however, that the Court of Justice will be

548

Working Paper II

equipped with sufficient resources. Therefore this is a major objective for the community.

V. How could the business community and advisors handle the new rules in practice? As we do not know how the Commission will 'translate' the new system of the White Paper into a draft regulation, speculation about the reactions of the business community and its advisors must limit itself to general principles. As in the past, undertakings intending to enter into an agreement that might have an impact on Article 81 will seek advice of experienced (in-house or external) lawyers. Based on precedents decided by the Commission and/or courts, and based on existing or future guidelines of the Commission (namely the guidelines concerning vertical and horizontal agreements), it is likely that the advisors will be able to render a reasoned opinion as to whether or not Article 81 (1) is applicable and whether the conditions for an exemption under Article 81 (3) are fulfilled in quite a number of cases. However, as already mentioned, the value of many precedents of the Commission is limited. Nevertheless, in quite a number of cases it is likely that advisors will reach conclusions with a reliability that is sufficient in order for the undertakings to enter into the agreements in question without further administrative ex ante clearance. Under the existing system of Regulation 17/62 there are quite a number of agreements that have been 'cleared' by the undertakings and their advisors, with the result that no notification was made, because there was no risk of disputes between the parties of the agreement or of complaints from third parties, and (if necessary) the agreements could be 'cleared' by a comfort letter. In cases where doubts remain due to a lack of guidance—either of previous decisions or by means of the guidelines—the advisors will delete provisions falling under Article 81 (1) if possible. Alternatively they would advise parties to ask either the Commission or an NCA for a 'business letter'. It would facilitate the introduction of the new system if the Commission and the NCAs—at least in the beginning—were not too restrictive and hesitant in granting such business letters in cases where the request is justified. This is all the more necessary in cases in which disputes about the validity of certain provisions of the agreement under Article 81 (2) arise after the agreement has commenced. In these cases a business letter could avoid unnecessary court proceedings. It would be advisable for the parties of an agreement to either exchange opinion letters from their respective advisors or to ask for a joint opinion letter that would become the basis of a common understanding between the parties. This would help to prevent future disputes without, of course, hindering the parties from submitting new arguments or referring to new factual developments in

3 - Courts and Judges

549

order to justify the non-performance of an obligation for reasons of invalidity under Article 81 (2). Such joint opinion letters would also be a factual means by which to protect the parties against fines in cases in which the Commission or a national competition authority later had different views concerning the applicability of Article 81 (3). If the parties have acted in good faith in this way, it would not be justified to impose fines on them for a possible violation of Article 81. In a case where the validity of certain provisions depends on the applicability of Article 81 (1) or the fulfilment of the conditions of Article 81 (3), the parties should not only draft good severability clauses in their agreement but should also provide for an express obligation to amend their agreement and to co-operate in good faith in order to render the agreement valid in accordance with EC competition law. There remain two groups of agreements that cannot be handled in the above way: 1) agreements involving considerable investments; and/or 2) agreements that are very likely to immediately attract comments or complaints from third parties. We discuss each of these two groups below: 1) The Commission proposes to submit partial-function production joint ventures to the rules of the Merger Control Regulation. As already explained, this would be a solution for particular, but certainly not all, cases involving considerable investments. Therefore, the proposed derogation for partialfunction production joint ventures should be extended as proposed. Alternatively, these would be typical cases for which a 'declaratory exemption decision' of the Commission is required. It has already been submitted that, in the absence of such an administrative ex ante clearance, the undertakings would tend to enter into more concentrated structures than necessary (that is, by forming a joint venture that would qualify for the application of the Merger Regulation, although they would not need this degree of concentration). The disadvantage of such structures would be that they are more difficult to terminate than agreements with a more co-operative structure. 2) Examples of agreements in the second category would be airline alliance cases with a number of participants and potential conflicts with third parties. These agreements might also involve far-reaching business decisions, including the distribution of slots among the companies concerned. For such business decisions, an ex ante clearance by a 'positive act' of an authority within a reasonable period of time would be necessary. In the case of disputes between the parties of an agreement, the aboverecommended joint opinion becomes relevant in the pre-court discussions of the parties. It would also play an important role in court proceedings as evidence of the common understanding of the parties. This could have the

550

Working Paper II

consequence that the party alleging that the underlying facts or legal considerations were or are wrong or have changed 'factually', bears a certain burden of proof for this argument. However, it goes without saying that the general rules of the burden of proof will not be affected, that is, that the party invoking the applicability of Article 81 (3) bears the burden of proof. Third parties affected by alleged infringements of Article 81 under the new system would have the same opportunities to make complaints to either the Commission or to the competent national authorities as they do under the existing system. As the Commission has promised to use its resources to proceed against violations of Article 81, such complaints might be more successful in the future than under the present system. Therefore, the new system might increase the number of complaints from third parties if the Commission keeps its word to concentrate on such cases.

VI. Conclusions From the above, the following conclusions can be drawn: 1) The national courts are prepared and able to apply Article 81 (3). 2) There is no considerable danger of forum shopping among the national courts. 3) The new system will produce sufficient legal certainty for not only partialfunction production joint ventures but also all other joint ventures that require substantial investments. Far-reaching business decisions are treated under the Merger Regulation or under similar procedural rules. Further, the Commission should adopt individual 'positive' exemption decisions in cases where important organisational measures have to be taken. The Commission should also consider issuing non-binding business letters in cases where new factual or legal questions have to be decided. In order to avoid inflationary requests for 'positive' exemption decisions or business letters, the introduction of fees should be considered. 4) The new system will sufficiently guarantee a coherent application of Article 81 by the national courts.

Ill Judge John D. Cooke Commission White Paper on Decentralisation of Competition Rules: The Threat to Consistency

I. Introduction At the 1996 edition of the EUI Competition Law and Policy Workshop, one of the subjects discussed was the deficiencies of the existing centralised system of enforcing Articles 81 and 82 EC Treaty (then Articles 85 and 86). As the publication following the event shows, there was a general consensus amongst participants that the accumulation of power in Brussels, resulting from a broad interpretation of Article 81(1) combined with the Commission's monopoly of the exemption power under Article 81 (3), had become unworkable. It was clear that some form of decentralisation of enforcement was necessary if the competition rules were to remain efficient and effective; but there appears to have been no consensus as to exactly how that might be achieved. Clearly, the European Commission has taken to heart the lessons of the 1996 Workshop. The Commission's White Paper proposes what appears to be a radical restructuring of the Community competition enforcement regime. Nevertheless, in spite of the detail of the White Paper, it is by no means clear that the Commission has worked out in full how the proposed decentralised regime might operate in practice. This is particularly so, it seems to me, as regards the intended relationship between the Commission and the national competition authorities of the Member States ('NCAs'), the relationship of those authorities inter se and, most importantly, the relationship between the Commission and the national courts when adjudicating on the application of the Community competition rules. The basic proposal can be stated very simply. Prior notification for exemption under Article 81 (3) will be discontinued; the entirety of Article 81, including the exemption provision of paragraph 3, will become directly applicable; and the NCAs will be empowered to apply and enforce the competition rules thereby, in effect, taking over at national level a major part of investigation, prosecution and enforcement hitherto performed centrally in Brussels. The White Paper explains fully why such a reform has now become a matter of urgent need. It also explains in detail the Commission's rationale for choosing this particular approach as the optimum solution to its current problem of limited resources in the face of an increasing volume of work. The White Paper is also candid in acknowledging that such a switch to decentralised enforcement poses an obvious risk to the uniform and consistent application of Community competition rules (paragraphs 83 & 101). Unfortunately,

552

Working Paper III

the White Paper is at its weakest when it seeks to suggest how this risk might be minimised and what types of 'mechanism' (see paragraph 103) might be put in place in order to avoid conflicts between Commission competition policy and NCA decisions and between decisions of different NCAs and national courts. In spite of these areas of uncertainty, I believe that the proposals are both legally feasible and practically achievable provided care is taken in the new Regulation to spell out in clear terms what the respective powers and responsibilities of the various agencies will be. In a Union with a possible 20 or more Member States and an equivalent number of NCAs it will be vital to both the efficient functioning of a new regime and to its public credibility that its structures be clear cut and easily understood rather than an awkward regulatory maze of exceptions, reserve powers and overlapping competencies leading to inter-agency rivalry and bureaucratic confusion.

II. The consistency problem Any change that involves substituting a multiplicity of enforcement agencies for a single body that as hitherto both formulated policy and exercised the primary enforcement power diminishes the uniformity and consistency of approach. The Commission acknowledges this risk at paragraph 83 of the White Paper: 'decentralisation must not be allowed to result in inconsistent application of Community competition law'. The real question, therefore, is not whether such a revised regime can jeopardise the consistency of Community competition law, but whether it is possible to structure the necessary measures so as to minimise the extent to which this will happen, and whether at the end of the day any remaining degree of risk is tolerable in view of the overall benefits to be gained from the new regime. Of course, it is an exaggeration to categorise the Commission's proposal as involving a radical switch from an exclusively unitary and centralised regime to a wholly decentralised one instead. The Community competition rules have always had an element of decentralisation, in the sense that Article 81 (1) and the entirety of Article 82 have, since they came into force, been directly applicable and capable of being enforced by national courts in all Member States and by those NCAs which have been so empowered (see White Paper paragraph 94). Accordingly, it is not so much the devolving of investigation and prosecution of infringements by the Commission which raises the threat to consistency as the combination of that devolution with the important policy decision to abolish the prior notification system for all but the limited category of partial-function production joint ventures (see White Paper paragraph 79) and to render Article 81 (3) directly applicable.

3 - Courts and Judges

553

III. Potential conflicts Clearly, the potential for inconsistency is both vertical and horizontal. A NCA may take a view of a particular agreement or practice which is at variance with the Commission's assessment of its effect or with the Commission's policy in the area in question (see White Paper paragraph 101). Two or more NCAs may have a simultaneous interest in a given agreement or alleged abuse and reach divergent decisions. Even in the absence of any element of bad faith or local partiality, NCAs may tend to have differences of approach to particular economic sectors for reasons of experience, tradition or national sensitivity. (In recent months we have seen how local sensibilities can quickly excite national anxieties in particular sectors, like banking in France and Portugal, or volume car manufacture in the U.K., for example). As regards the vertical perspective, part of the problem lies in the fact that there may be an inherent incompatibility between the essential policy considerations of an exemption decision under paragraph (3) and the purely territorial perspective and competence of individual NCAs. Indeed, the basic rules embodied in Articles 81 and 82 have as their fundamental rationale the fact that the agreements, practices and abuses thereby prohibited are 'incompatible with the common market'. As long ago as 1966, the European Court described the nature of the task carried out by the Commission in applying paragraph (3) as that of maintaining '... workable competition, that is to say the degree of competition necessary to ensure the observance of the basic requirements and the attainment of the objectives of the Treaty, in particular the creation of the single market achieving conditions similar to those of a domestic market.'1

In that case too and repeatedly in the jurisprudence ever since, the European Court recognised the essential policy function of the Commission in applying paragraph (3). It said that the Commission's task '. . . necessarily implies complex evaluations of economic matters. Judicial review of these evaluations must take account of their nature by confining itself to an examination of the relevance of the facts and of the legal consequences which the Commission deduces therefrom. This review must in the first place be carried out in respect of the reasons given for the decisions which must set out the facts and considerations on which the said evaluation are based.'

IV. Territoriality In other words, the very scope of the judicial review remedy is effectively dictated by the nature of the function imposed by the Treaty on the Commission, Cases 56 and 58/64 Consten and Grundig v Commission [1996] ECR 299.

554

' Working Paper HI

namely, to administer competition policy and to apply the competition rules with a view to achieving on a Community basis the conditions which obtain in a single domestic market. If the application of the exemption function in paragraph (3) is inherently a policy-based exercise, and one which is intended to be operated in the light of the common or single nature of the Community market, one could query whether it is consistent with the essential character of Article 81 that it should be administered through a fragmented administrative system based on agencies exercising a purely territorial competence. In the brief passage at paragraphs 91-98, where the enhanced role of the NCAs in applying competition rules is outlined, the White Paper appears to acknowledge that individual decisions of NCAs will be confined in their effects to the territory of the agency in question. Thus, paragraph 95 reads: 'As the logical consequence of the abandonment of the Commission's monopoly of exemption, the national authorities would have to be able to withdraw the benefit of a block exemption on their own territory if that territory or part of it constituted a separate market.'

This seems to acknowledge that a situation might well evolve in due course in which, for example, a distribution agreement between two undertakings covered by Regulation 1983/83 (or its successor Regulation 2790/99) could operate differently from one Member State to another depending on the view taken by a given NCA. If one NCA took the view that the agreement did not 'allow consumers a fair share of the resulting benefit', it would appear to be entitled to withdraw the benefit of the block exemption. The NCA in the next door Member State might take a more benign view and reject a complaint by refusing to withdraw the block exemption. Such a partitioning operation of a piece of Community legislation hardly seems compatible with the single market function of the competition rules. So far as proceedings before national competition authorities are concerned, it is probably true to say that most serious conflicts can probably be avoided by the imposition of a prompt notification obligation, so that the Commission is informed in good time of both the commencement of an investigation into an alleged infringement and of any proposed decision. Obviously, this latter notification would be particularly important in any case where the NCA proposes to apply Article 81 (3) or to withdraw a block exemption. Such a notification system should be adequate to reduce major inconsistencies provided national authorities can be trusted to operate it efficiently and provided the Commission applies adequate resources to monitoring and following up the notifications it receives, so that it intervenes in good time where it is satisfied intervention is warranted.

3 - Courts and Judges

555

V. National courts The avoidance of conflicts between Commission policy and the judgements of national courts, on the other hand, may well be more problematic. Once the necessary legal powers and obligations are conferred or imposed upon the NCAs, administrative co-operation between them and the Commission should relatively be a routine. Indeed, it is probably true to say that a routine of co-operation between the Commission and the authorities in Member States is already well established. The position with regard to the intervention of the Commission in proceedings before national courts, however, involves radically different considerations. In the first place, court rules and procedures vary considerably throughout the Member States. There are different constitutional traditions and cultures to be taken into consideration when requiring national courts to co-operate with agencies that are not themselves parties to a particular proceeding. Once prior notification for exemption has been abolished and Art. 81 (3) is made directly applicable, it is highly likely that the most difficult issues for national courts will arise in the context of adversarial litigation, where the availability of exemptions is used as a ground of defence in, for example, contract disputes. It is in this area rather than in national judicial review of NCA decisions that problems are most likely to arise. At paragraph 107, the White Paper takes what appears to be somewhat simplistic view of the potential problem. It says '. . . it is vital, first of all, that the Commission should be aware of proceedings in which Articles 85 and 86 are invoked before the courts, so that it is made aware of any problems of textual interpretation or lacunae in the legislative framework. It is therefore proposed that the regulation should require courts to supply such information.'

VI. A notification mechanism It is, no doubt, a relatively simple matter to require any court to notify the Commission of the commencement of a case in which the application of Article 81 or 82 is invoked. But at what point should this be done, and how extensive should the information given to the Commission be? Logically, one would suppose that the Commission should be notified as soon as possible, namely, once the case had been introduced in the registry of the court in question. However, in the common law system of the United Kingdom and Ireland for example this may be of little utility, because common law litigation relies entirely on the initiative of the parties and the role of the court registry is limited. Even where a writ or summons is lodged with the registry with an express

556

Working Paper HI

claim based on Article 81 or 82, the registry itself may know nothing whatsoever as to the substance of the case. Secondly, a case may be introduced as a claim for damages for breach of contract, for example, and it may not be until several weeks or even months later when a defence is lodged that any reference to a Community competition rule arises. Even where Article 81 or 82 has been expressly mentioned in written pleadings, the judge hearing the case may obtain no detailed appreciation of the nature and substance of the competition issues involved until the case is actually at hearing before him. Thus, in the common law jurisdictions at least, the court itself may not be in a position to give any useful notification to the European Commission until the litigation is at a very advanced stage. Accordingly, if the sort of notification arrangement envisaged in paragraph 107 of the White Paper is to operate efficiently in those jurisdictions, it may well be necessary to formulate a different approach, in which the notification obligation is imposed primarily upon the parties rather than the court. It is also questionable whether it is an efficient solution to require the information to be supplied to the Commission. Given its existing shortfall in resources, it seems highly improbable that the Commission will ever be able to deploy the resources required to collate and follow-up efficiently such information. If the bulk of day-to-day surveillance and enforcement is entrusted to the local NCAs, it would seem far more efficient to exploit their presence in the Member States for this purpose. The NCAs will be far more familiar with local conditions and practices; they will often, presumably, be engaged as litigants in the national courts in their own right enforcing the rules or defending their own decisions. They will know the judges and procedures involved; they will most likely have their own legal service or standing counsel. It would seem logical and more effective therefore for the Commission to use its network of NCAs as its local agents for this purpose. Thus, a simple and more effective solution to this problem would be to provide in the regulation that the court rules of procedure in the Member States require any party to litigation who invokes Article 81 and 82, whether by way of claim or defence, to give notice of that fact to the NCA of the Member State concerned and to furnish the relevant pleadings to it. The NCA would then be afforded a period in which, after consultation with the Commission, it might decide to intervene in the action; to decline all participation or to merely require to be notified in advance by the Court in the event that it was proposed to make a particular form of order at the conclusion of the case—lifting a block exemption or applying Article 81 (3) for example. Notification through the agency of the NCAs would, in my view, not only be a more efficient use of resources but would overcome a more sensitive aspect of the Commission's proposed intervention at national court level.

3 - Courts and Judges

557

VII. Commission intervention In paragraph 107 of the White Paper the Commission also proposes that it should be entitled on foot of such notification to intervene 'subject to leave of the court... in judicial proceedings that come to its notice' and to do so apparently as amicus curiae. While the concept of the intervention in litigation as amicus curiae is, if rarely encountered, nonetheless understood in the common law jurisdictions, it must be questioned whether this is a prudent or legitimate role for the Commission in the context of a decentralised regime for enforcing competition rules. In the first place, is it prudent for the Commission as the institution representing the Community (Article 282) to become involved directly in litigation between private parties in national courts? Suppose for example the Member State concerned or one of its ministries has a divergent view of the issues and intervenes on the opposite side? Is it desirable from the viewpoint of the Community constitutional order and the relationships between institutions and Member State that the Commission become involved in litigation against a Member State in a national court? Secondly, is the role of amicus curiae adequate or appropriate given the particular interests the Commission has in competition matters, where it is at once a policy maker, legislator, enforcer and adjudicator? In cases where the Commission's interest is merely to inform the court of the Commission policy in relation to the sector, intervention confined to filing of an amicus brief, as happens in U.S. antitrust trials, may well be helpful to the court and sufficient for the Commission's purpose. But the implication of the proposal in the White Paper is that the entitlement to intervene has the more far-reaching objective of promoting consistency of interpretation and of avoiding decisions running contrary to the Commission's view of the effect of an agreement or practice. In such cases the interest of the Commission goes beyond that of an amicus and its role becomes that of a protagonist for a particular result to the case. That being so, it seems to me to be necessary that any contribution by the Commission (or by the NCA on its behalf) should be as an intervening party supporting the relief contended for by one or other party or for an alternative ruling contended for by itself. Certainly, so far as the common law jurisdictions are concerned, national courts will require to know from the regulation precisely what the role and standing of the Commission or NCA is to be, because only then will the judge be able to determine how to treat any advice, information or documents which the Commission puts forward. It must be borne in mind that in many cases the very facts upon which the application of Article 81 or 82 turns will be hotly contested between the parties. If the Commission has a view of factual issues, will it be obliged to advance proof? Will its witnesses or the authors of its expert reports be open to cross-examination?

558

Working Paper III

If parties are required to notify pending cases to the Commission and the latter is entitled to intervene where it sees fit, there are the further implications for the court and the parties. In such interpartes litigation before a common law court, it would not be unusual for the parties to attempt to reach an amicable settlement. If the Commission has intervened, would its approval of any settlement be necessary before the court could embody such an agreement in its formal order? If the Commission has decided not to intervene, would the court have any obligation to notify the Commission of any proposed settlement particularly, for example, if the terms of the settlement involved an acknowledgement that Article 81 (3) did or did not apply?

VIII. Reserve power to prohibit Perhaps the most questionable of all of the Commission's proposals in relation to its intervention at national level, is that which is mentioned at paragraph 102(2). This paragraph refers to situations where an NCA decision which is considered to be incorrect by the Commission has become definitive, either because no appeal has been brought against it within the time permitted, or because it has been confirmed on appeal, or where a national court has given a positive judgement rejecting a complaint on the ground that Article 81 (3) applies. In such cases the White Paper claims that '.. . the Commission can always intervene to prohibit the agreement, subject only to the principle of resjudicata that applies to the dispute between the parties themselves, which has been decided once and for all by the national court'.

It is by no means clear what this passage is intended to convey in practical terms. But if it means that the Commission reserves to itself the prerogative of declaring illegal under Article 81 an agreement which a national court has held to be lawful, then the constitutional propriety of such an approach must be open to question. It seems to be suggested that where a national court has upheld as valid an agreement or practice between two or more undertakings, the Commission could nevertheless intervene to declare that same agreement prohibited under Article 81 (1) but not as regards the very parties to the agreement or practice itself. In other words, the very parties who have infringed Article 81 are the only ones to 'get away with it'. It is difficult to understand both the practical implication of such a proposition and its legal effect. If the validity of the agreement or practice 'has been decided once and for all by the national courts' so that the parties to it are perfectly entitled to continue performing their contract or operating the practice, to whom is the prohibition to apply and what third parties can invoke its benefits and how? Furthermore, the type of situation which is described will be one which has presumably arisen after the notification arrangement of paragraph 107 has been complied with. In other words, at some stage the Commission will have been notified of the liti-

3 - Courts and Judges

559

gation in question and will, presumably, have declined to intervene. In those circumstances, if there has been no change of circumstances which justifies the adoption of a different appraisal it must be highly doubtful whether it is either legally appropriate or practically wise for the Commission to contemplate the adoption'of a prohibition decision. In effect, the Commission cannot seek to have its cake and eat it. It cannot on the one hand opt for a decentralised system of this kind and, on the other hand, seek to reserve to itself both the right to intervene at national level to promote its view and still exercise the prerogative of overturning results of which it disapproves. In applying Articles 81 and 82 (including a directly applicable Article 81 (3), the national courts act as Community courts. Supervision of their application of Community law operates through the mechanism of Article 234 in the European Court of Justice and not through Regulation 17/62. There must, in other words, be some implication for the rule of law and the independence of the national judiciary, not to mention the rights of litigant undertakings, if a final judgement handed down in litigation in which the Commission has had the opportunity to intervene and which involves the lawful (even if mistaken) application of Article 81 can be set aside by an administrative authority which disagrees with the result. At the very least, the knowledge that such a possibility exists must act as a disincentive to national judges to take seriously their role as an important link in the Community judicial structure for the application of competition law.

IX. Inconsistency of judicial review There is one final and important observation that should be made about the White Paper proposals in the context of mechanisms to reduce inconsistency. Clearly, if the main burden of the day-to-day application of Articles 81 and 82 is henceforth to fall upon the NCAs, it is obvious that there will be a consequential increase in the involvement of national courts in litigation at national level arising out of Article 81 and 82. Quite apart from the removal of notification for exemption as an effective obstacle to litigation based on Article 81, the individual decisions of the NCAs on complaints of infringement and their decisions in relation to the grant or refusal of exemptions and the withdrawal of block exemption will arise for judicial review before the national courts. It follows that there is likely to be an increased reliance by national courts on the preliminary reference procedure of Article 234 and, indeed, the White Paper (at paragraph 104) itself recognises that the preliminary reference procedure is one of the essential mechanisms for reducing the danger of inconsistency. Unfortunately, this aspect of the proposal raises a complication difficult to resolve as matters stand. Since 1989 judicial review of Commission decisions under Articles 81 and 82, and particularly its decisions in the application of paragraph (3) of Article 81,

560

Working Paper III

have been examined on appeal initially by the Court of First Instance. Indeed, one of the main reasons for establishing the CFI in 1989 was the very need to create a jurisdiction in which the complex legal and economic issues raised by the application of these Articles could be given a more detailed examination than the ECJ was then able to offer. By and large it seems to be agreed that the CFI has discharged this function in a satisfactory manner during the last 10 years and, certainly, the conscientiousness of its approach and the diligence of its judicial scrutiny of Commission decisions have been consistently endorsed on further appeal by the ECJ. One of the consequences of the White Paper proposal to make Art. 81 (3) directly applicable and to decentralise its application in favour of the NCAs is, of course, that the judicial review of those NCA decisions by the national courts will only receive judicial examination at Community level by the ECJ under Article 234. In other words, the proposal has the effect of bypassing the CFI and removing from its jurisdiction the very cases which it was originally established to deal with. Ideally, of course, references from national courts engaged in the review of decisions by NCAs applying Articles 81 and 82 ought to be dealt with, at least initially, by the CFI. However, the preliminary reference procedure under Article 234 is the one category of cases which is precluded from being transferred to the CFI from the ECJ under Article 225 of the Treaty. The ECJ and the CFI have, however, recently formulated proposals to the current Intergovernmental Conference for reform of the existing judicial structures so as to alleviate the difficulties both courts are currently experiencing. One of these proposals is that the Treaty should be amended so as remove this prohibition in respect of Article 234 references, and to thereby open up the possibility that certain discrete categories of 234 references might be devolved to the CFI. Clearly, references from national courts arising out of judicial review of NCA decisions applying Articles 81 and 82 would be an obvious category of case ideally suited to such a transfer. It remains to be seen, however, whether the political will exists among the Member States to implement such a Treaty amendment. If such an amendment is not politically feasible, however, the undoubted potential for inconsistency in the application of Community competition rules will be exacerbated by a simultaneous weakening of judicial review at Community level. The decisions taken by the European Commission as the primary and central enforcement agency will, of course, remain subject to immediate scrutiny and review on appeal by the CFI. On the other hand, corresponding decisions when taken by national competition authorities will be judicially reviewed by national courts and the Community judicature will be involved, if at all, only through the mechanism of Article 234. Thus, decentralisation involves not only a switch of the scrutiny of a material number of infringement decisions from the CFI to the national courts, but also a change in the nature of the scrutiny at Community level, from the direct and exacting instrument of the annulment action of article 230 to the essentially advisory instrument of article 234.

3 - Courts and Judges

561

As it happens, there is a useful illustration of this issue in two cases currently pending before the Luxembourg Courts. They are the latest round in the ice cream cabinet wars between the Unilever and Mars Groups that have been going on around Europe for some years (see also Case T—7/93 Langnese-Iglo GmbHw Commission [1995] ECR 11-1533). In Case C-344/98 Masterfoods Ltdv HB Ice Cream Ltd, the Court of Justice is asked to rule on three questions referred by the Supreme Court of Ireland arising out of an appeal to that court from a 1992 judgment in which the Irish High Court had dismissed the claims of Masterfoods (the Mars Inc. Irish subsidiary,) to the effect that the exclusivity clauses in the HB Ice Cream (the Unilever subsidiary) contracts with retailers infringed Article 85 and that the latter's conduct infringed Article 86. The High Court had additionally granted a permanent injunction to HB restraining Masterfoods from inducing retailers to stock Mars products in cabinets to which the exclusivity contracts applied. In Case T-65/98 Van den Berg Foods Ltd v Commission of the European Communities, HB Ice Cream Ltd (now under its changed name) is seeking under Article 230(4) to annul a Commission decision of 11 March 1998 in which the Commission held, inter alia, that the exclusivity clauses in the HB contracts did infringe Article 85 (1) and could not be exempted under Article 85 (3), and that HB's inducements to other retailers to enter into exclusivity commitments constituted an abuse of its dominant position in violation of Article 86. This case has had to be suspended by the CFI under Article 47(3) of the Statute of the Court of Justice and Article 77a of the CFI Rules of Procedure pending delivery of the ruling on the reference in Case C-344/98. In the reference for preliminary ruling, the Irish court has asked two straightforward but obviously crucial and difficult questions going to the heart of the substance of this dispute: '2. Having regard to the legal and economic context of the cabinet agreements at issue in the market for single-wrapped items of impulse ice-cream, does the practice whereby a manufacturer and/or supplier of ice-cream provides a freezer to a retailer at no direct charge—or otherwise induces the retailer to accept the freezer-—subject to the condition that the retailer stock no ice cream is such freezer other than that supplied by the said manufacturer and/or supplier constitute an infringement of the provisions of Article 85 (1) and/or Article 86 of the EC Treaty? Are freezer exclusivity agreements protected from challenge under Articles 85 and 86 of the EC Treaty by reason of the provisions of Article 222 of the EC Treaty?' In Case T-65/98 on the other hand, the applicant has mounted a detailed attack which ranges over the full factual, economic and legal analysis of the Commission decision and raises issues of error of assessment and of law touching on Articles 85 (1), 85 (3), 86, 190 and 222 (now, of course, Articles 81 (1), 81 (3), 82, 253 and 295). Thus, even a cursory statement of the gist of the substantive issues in the cases serves to illustrate the radically different nature of the judicial tasked called for in each case and the depth of scrutiny which each involves. Indeed, the point has already been adverted to in the opinion of Advocate General Cosmas in Case C-344/98 delivered on 16 May 2000 where

562

Working Paper III

he comments on the significance of the fact that the parallel proceeding is pending in the CFI. At paragraph 53, he points out that the distinction between the annulment action and the Article 234 reference may not be of practical importance when the subject of judicial review is the validity of a Community legislative measure, but he then adds: '53. .. .The assessment of the legality of those measures is confined principally to the exercise of pure judicial review without it being necessary to examine questions of substance. Conversely, in the case of individual administrative measures such as that before the Court, the exercise of full review of the substance is capital for the effective provision of judicial protection. 54. Further, the Commission decision before the Court in this case has another special feature. It relates to the application in a particular case of the provisions of Articles 85 (1) and 86 of the EC Treaty. It presupposes, in other words, complex technical and economic assessments, which, if they are to be correct, require exhaustive review of the substance by a specialised judicial authority. In order to meet that need, inter alia, the Community legislature on constitutional matters was led to set up the Court of First Instance. By its systematic hearing of actions for the annulment of Commission decisions, which resemble the decision at issue here, that court has succeeded in deepening and strengthening judicial review of those decisions, thus contributing to the improvement of the Community system for the provision of judicial protection.'

One final point is, I think, illustrated by these two pending cases. In addition to the two substantive questions posed by the Irish court and quoted above, the reference contains a first question directed towards the procedural dilemma that the national court faces when confronted with an administrative decision of the Commission intervening during or after the judicial process at national level in the same or a related case. The question posed is two-fold, and asks in effect whether the national court is obliged to suspend proceedings until the CFI has disposed of the appeal against the Commission decision and any appeal to the ECJ, and whether the effect of such a decision addressed to an individual party is to prevent that party from seeking to uphold a contrary judgment under appeal at national level. Thus, references from national courts tend not to come in neat packages according to subject matter. If, therefore, competition issues do constitute a category of preliminary reference which might ideally be devolved to the CFI, the mechanism for doing so should ideally be accompanied by a provision which allows the ECJ to split the reference, so as to retain to itself for prior answer questions of this kind which simultaneously deal with constitutional and procedural issues and which merit the urgent and exclusive attention of the Court of Justice. Of course, no scheme of decentralisation is going to eliminate completely the problem of different results being reached in this manner by the Commission on the one hand and the NCAs or national courts on the other, so long as the articles have direct effect and the Commission retains the function of enforcing them, at least in cases of major Community significance. Indeed, given the wide Community interest in the practices involved, it seems likely that the ice cream

3 - Courts and Judges

563

cabinet dispute is precisely the type of case that would be retained by the Commission or taken over from the NCAs involved in any event. The cases do however, in my view, illustrate how important it is to bear in mind that this issue of decentralisation of competition rule enforcement is not merely a matter of finding the most convenient administrative solution to a problem of limited resources. Rather, it is bound up in a complex way with important questions as to the structure and quality of judicial protection at both national and Community level, and the constitutional and procedural relationships between the two.

IV Judge David Edward The Modernisation of EC Antitrust Policy: Issues for Courts and Judges

The Commission White Paper proposes to transfer to national and Community judges a significant part of the responsibility for the implementation of EC competition law and policy. This paper seeks to identify, from the standpoint of a Community judge, some of the questions that need to be asked and problems that need to be faced if such a transfer is to be effective and useful. Two caveats are necessary at the outset. First, the purpose of the paper is to raise questions and identify problems. This may give the impression that the writer opposes the Commission proposal. That is not the intention: the writer is sceptical but undecided. Second, the paper contains many generalisations about courts and their procedures. Please read most sentences as if they began with 'Broadly speaking . . .'. Competition law can arise in the courts in three ways: • In proceedings for judicial review of administrative decisions; • In private law actions between private parties; • In criminal prosecutions or quasi-penal proceedings brought against a private party in a court of law by a public prosecutor or other public authority. In Europe we have substantial experience of the first category, rather less experience of the second, and little or no experience of the third. In this respect, European experience is different from US experience. In particular, we have no procedure analogous to the antitrust proceedings brought against Microsoft in the US Federal Courts by the Department of Justice. The only 'trial courts' available in the EC for competition cases are the civil and commercial courts of the Member States. Two other differences between the US and Europe should also be stressed. First, la judiciarisation de la vie is second nature to most Americans, but it is much less part of the European culture. Second, political and public opinion in the US is more ready to accept competition as a good in itself than opinion in most European countries. The procedures for judicial review of administrative decisions and the criteria applied by the courts are broadly similar throughout the Community. This is not surprising, since in most countries the judge's function in such cases is solely to assess the lawfulness, and not the merits, of the administrative decision. Where the administrative decision has been taken under powers conferred by competition legislation, the competition law issue will be the only issue in the case and will be more or less clearly defined (and confined). In the context of judicial review of administrative action, competition law can be treated as a specialist subject, to be dealt with by specialised tribunals

566

Working Paper IV

(e.g. a specialised chamber of the CFI). By contrast, private law suits are heard in courts of general jurisdiction. The procedures and criteria applied in these courts differ widely between the Member States, particularly as regards the concept of 'evidence' and how it is provided. In this respect also, US experience is not a safe guide. The US is procedurally and linguistically homogeneous; the EC is not. Moreover, in a private law suit, the competition law issue may not be the only issue with which the judge has to deal with. Typically, Article 81 or 82 will have been raised as a defence to an action for performance of a contract or to prevent the marketing of particular goods. But the issues in the case may extend more widely, and some of them may have nothing to do with EC law, far less competition law. The judge(s) dealing with the case may have little knowledge or experience of EC law and little or none of competition law. Consequently, courts of general jurisdiction dealing with private law suits are likely, to a greater extent than specialist tribunals, to need the help of the Commission and/or of the Court of Justice through Article 234 procedure. Help must be provided as quickly as possible so that the action as a whole can be decided within a reasonable time. These considerations have a bearing on the Commission's proposal in three respects: First, the extent to which and the way in which the Commission could help national courts or intervene in national proceedings will differ considerably from one Member State to another. For example, in the UK it will be important to determine whether, and if so in what form (written, oral or both), the Commission would provide 'evidence'; if so, whether the Commission witness would be open to cross-examination; and whether the Commission evidence would be treated as 'expert evidence'. The answers to these questions could have profound procedural consequences in the UK (and also, presumably, in Ireland). The questions would not (and perhaps could not) arise in that form in other Member States, but the same would be true, mutatis mutandis, of procedural problems in those states. Second, even at present, Article 234 references do not normally come to the ECJ neatly packaged as 'competition law references'. For example, the Bosman case (C-415/93 Union Royale Beige des Societes de Football Association and Others v Bosman and Others [1995] ECR 1-4921) raised questions of discrimination on grounds of nationality, free movement of workers and competition law. Very often, the referring judge asks whether a particular national situation is compatible with a whole series of Treaty rules, including, but not confined to, Articles 81 and 82. In any event, Articles 81 and 82 have to be read in the wider context of the Treaty as a whole. The 'Rules on Competition' cover public undertakings and state aids. They appear in a Title that also covers taxation. That Title, in turn, forms part of the scheme for creating and maintaining the internal market for goods, persons, services and capital. Since Maastricht, the four freedoms are no longer 'The Foundations of the Community', but rather an aspect of 'Community Policies'. These policies

3 - Courts and Judges

567

cover a wide range of topics, including social policy, consumer protection, industry, economic and social cohesion, research and technological development and environment. The consequences of putting all these matters on the same footing in the Treaty have still to be fully worked out. It has already affected the Court's approach to some aspects of the traditional jurisprudence on free movement of goods. To a much greater extent than before, the claims of free movement of goods have to be balanced against other policy considerations, such as consumer and environmental protection. Similarly, in Albany International (Case C-67/96 Albany International v Stichting Bedrijfepensioenfonds Textilindustrie [1999] ECR 1-0000), the claims of free competition in the provision of occupational pension schemes had to be balanced against other social objectives set out in the Treaty. For all these reasons, 'competition law references' cannot be treated, a priori, as a pre-defined category capable of being assigned to a specialised chamber of the CFI (or some other specialised tribunal). The division of competence between the ECJ and the CFI in the field of state aids has already led to considerable problems and a good deal of avoidable delay. It would therefore be necessary to define with some precision what would, and what would not, be transferred to the CFI or other specialist tribunal. The desire for 'specialised' treatment of competition law must not lose sight of the latter's place in the scheme of the Treaty as a whole. Third, the resources of information, assistance and guidance promised by the White Paper, which would be necessary for the courts (national and Community) to perform their new function, must be made available within a reasonable time and in a usable form. The proposed reforms may relieve the Commission of the task of applying Article 81(3), but that does not mean that the task will have disappeared—it will simply have been moved elsewhere in a different form. It should be recognised that the courts of all the Member States, as well as the ECJ and CFI, are already suffering from severe overload, and that this situation is unlikely to change in the near future, if at all. If the proposed new system is to depend on the Commission providing national and Community courts with advice and assistance, the Commission cannot be allowed to plead lack of time, resources or manpower as a reason for not doing so. In that connection, it is disturbing that the Commission has already suggested, in a different context, that it should no longer submit observations in all Article 234 references. The principal reason for the writer's scepticism about the current proposal is indeed that the Commission has, over the past forty years, found it impossible to provide more than comfort letters in response to the overwhelming majority of individual notifications. What then are the grounds for believing that it will be able not only to monitor the decisions of national courts from Tallinn to Lisbon and from Cork to Larnaka, but also to provide those courts and the ECJ or CFI with the assistance they need to apply Article 81 (3) in concrete cases?

568

Working Paper IV

If the already overloaded courts will be left to their own devices, it would be more honest to say so, rather than promise what cannot be delivered. An alternative approach would be to make national competition authorities primarily responsible for assisting national courts, allowing them to call on the help of the Commission when required. This would be in keeping with the spirit of subsidiarity and could work in parallel with the well-established system of co-operation between national courts and the ECJ under Article 234. 'Balancing' of competing interests has, to an increasing extent, become a feature of Article 234 references. This is, in one sense, an argument in favour of the Commission's proposal. It shows that national courts applying Community law can reasonably be called on to balance non-legal considerations. There is no reason why the considerations mentioned in Article 81 (3) should be inherently more difficult to balance than the competing claims of 'commercial' human rights, consumer and environmental protection, social protection, racial and gender equality, etc. especially since the 'rule of reason' approach to Article 81 (1) already involves some degree of balancing. On the other hand, it is envisaged that Article 81 (3) would have to be interpreted and applied in the light of Community competition policy, the definition of which the Commission proposes to reserve to itself. This raises a number of questions, for example: • How would the relevant 'policy' be made known? • Would it be promulgated in binding legal instruments, in recommendations, or in some even softer form of law? • How would changes of policy be brought about and promulgated (for example, a new approach to vertical restraints)? • What would be the effect of any change of policy on the validity of existing agreements, particularly those that were already the subject of litigation? From the point of view of legal certainty and judicial operability, these are questions to which the answers cannot be left to be solved ambulando. It will be essential to define how, under the new dispensation, the three paragraphs of Article 81 are to be applied. There seem to be two possible approaches. In the first approach, any agreement, decision or practice falling within the scope of Article 81(1) would be prima facie unlawful and therefore void. It would then be for the party seeking to uphold the validity of the agreement, decision or practice, to prove positively that it qualified for exemption because it met the criteria of Article 81(3). Failing such proof, Article 81 (2) would apply. In the second approach, the court would weigh up simultaneously, on the one hand, the anti-competitive effects of the contested agreement, decision or practice (Article 81 (1)) and, on the other, its beneficial effects (Article 81 (3)). Depending on how the balance tipped, the result would be nullity (Article 81 (2)) or exemption (Article 81 (3)). From a judicial point of view, the first approach, relying on well-established criteria of onus of proof, would be easier to operate. The Cassis de Dijon

3 - Courts and Judges

569

jurisprudence as developed by the ECJ has adopted a similar approach: restrictions on free movement are prima facie unlawful unless they are shown to be objectively necessary and proportionate to achieve a legitimate public purpose. On the other hand, the second approach, involving the simultaneous balancing of all relevant considerations, would be more appropriate for an economic rather than a strictly legal assessment. It would make it unnecessary to decide whether, and to what extent, a rule of reason should be built into the interpretation of Article 81 (1), rather than Article 81 as a whole. Arbitrators are already deciding cases under Articles 81 and 82, and the Court has recently held that, at least where there is judicial control of arbitral decisions, the competition rules must be treated as d'ordre public. However, it is well established that arbitrators cannot make Article 234 references (see Case 102/81 Nordsee v Reedei Mond [1982] ECR 1055). Arbitration and Alternative Dispute Resolution (ADR) are being promoted as desirable alternatives to resort to the courts. On 13 April 2000, the Council approved the creation of a European network of national bodies for extrajudicial settlement of consumer disputes, and on 29 May 2000 the Ministers of Justice of the Member States initiated discussions at EC level on alternative methods of settling disputes under civil and commercial law. If Article 81 (3) is to have direct effect, arbitrators and other 'voluntary' tribunals will be called on to apply it. Does the Commission intend to monitor arbitral decisions and, if so, how? If not, why? Would it be feasible and/or desirable to make specific provision for arbitrators and mediators to make Article 234 references? If so, this should be done in the context of the current IGC. But the potential consequences for the effectiveness of arbitration and ADR, as well as the workload of the ECJ, cannot, at this stage, be foreseen. If Article 81 (3) is to be applied by the civil courts, giving interested parties a full right of appeal on the merits, the question will inevitably arise whether the residual powers of the Commission to take decisions in individual cases can continue to be exercised through an administrative procedure subject only to limited judicial review under Article 230. Ex hypothesi the cases dealt with directly by the Commission will be cases of major importance, possibly giving rise to very substantialfines.Would it be compatible with the principles of judicial protection that, in such cases, the enterprises concerned should have no right of appeal on the merits of their case? The same issue between the same parties may arise in parallel proceedings before the courts of different Member States. For example, the validity of a standard form of exclusive distribution agreement may be challenged by aggrieved distributors or competitors in several countries at the same time, although the nature and scope of the proceedings in each country may be different in other respects (see point 3 above). The courts of each country would undoubtedly have jurisdiction under the Brussels Convention. But they might well reach conflicting decisions as to the validity or enforceability of what is, essentially, the same agreement.

570

Working Paper IV

The rules of the Brussels Convention would not resolve the problem. Since at least one of the parties before the courts of the countries concerned (e.g. the local distributor) would be different, a decision by the courts of one country would not be resjudicata before the courts of any other country. Articles 21 and 22 of the Brussels Convention (Uspendens) would not assist, since the parties to the actions would not be the same, the cause of action would not necessarily be the same and it is very doubtful whether the actions could be said to be 'related'. It would therefore be desirable to devise a more detailed mechanism than the existing rules of the Brussels Convention to avoid multiple proceedings and conflicting decisions.

V Dan Goyder Providing Support for National Judges in Dealing with Competition Cases

I. Introduction This article proceeds on the assumption that the main elements of the European Commission's White Paper on the Modernisation of EC Antitrust are carried into force, so that it becomes necessary to consider the more extended role which national judges then could and should play in deciding private cases brought before them on competition issues. I would summarise my thoughts under four main headings, as follows: 1. Judges deciding competition law cases between private litigants have to deal with the cases that are brought before them and have no choice in their selection. The case may raise competition law issues either because a plaintiff is claiming damages, or as the result of a defendant putting forward a defence to a claim for monies otherwise due under contract. Judges do not operate in a vacuum: they can only carry out their functions satisfactorily if certain conditions apply. These include the existence of substantive law that lays down clear principles as well as consequences (e.g. prohibition and nonenforceability), when agreements or practices are found to be in breach of those principles. 2. In addition, judges in national courts deciding such private cases must operate within a framework of 'appropriate judicial architecture'. This includes a right of appeal from their decisions within a hierarchy of national courts which enables precedent to be established and any inconsistencies between decisions of courts of equal status resolved, in addition to the availability of the Article 234 preliminary reference procedure to resolve EC legal issues. For this purpose Article 234 cannot at present operate in a satisfactory way. It is clear from the recent public statements of its President that the ECJ cannot deal with the number of cases presented to it under this Article, when its other responsibilities are also taken in account, and the present delay in obtaining rulings in such references of nearly two years is unacceptable. I suggest that the time has now come when the Court of First Instance has to be brought in to deal with competition law issues, certainly on preliminary references under Article 234 and possibly also in a wider context as an appeal court in respect of national cases raising these issues. 3. The application of substantive principles of law to such private cases is always a demanding task even for a judge, quite apart from the innate ability of such cases to consume very long periods of judicial time. Specialist judges are therefore needed here, just as in other specialised areas such as

572

Working Paper V

intellectual property and tax. Even more crucial is that such judges can when necessary seek practical help from competent public authorities, whether the Competition Directorate of the European Commission or a national competition authority. 4. Whilst a broad judgment as to whether an agreement or practice falls within the prohibitions of Article 81 or 82 can normally be made by national judges (even if not specialists) the application of a system of exemption on broad policy grounds equivalent to Article 81 (3) does by contrast raise major difficulties for all judges in national courts. Ultimately the task can only be dealt with satisfactorily if there is a sensible linkage of public law and private law enforcement institutions and resources. In particular national courts should be allowed to seek assistance as mentioned above from the EC Competition Directorate or national competition authorities, and be supported by rules as to the binding evidentiary force of public law decisions or findings of fact on later private cases, as provided for example under Section 5(a) of the Clayton Act in the USA, and more recently by Section 58 of the UK Competition Act 1998.

II. The task of the national judge Turning to deal with these topics in rather more detail, my view is that in principle senior judges in national courts, particularly if allowed to specialise, could undertake the task of applying competition law rules as between undertakings and could even handle the difficult issues raised by claims for exemption under Article 81 (3). There are a number of conditions however that would need to be satisfied for this new system to work. The White Paper makes a powerful case against the continuation of the present system of notification to the European Commission as a condition precedent for an award of individual exemption; but nevertheless it is necessary to point out that we have lived happily for over 10 years with a system of mandatory notification, for mergers with a Community dimension. There cannot therefore be anything fundamentally objectionable in principle in the notification system as such. Likewise in respect of Article 82 we have lived for over 40 years with a system of direct effect (implemented in parallel by the European Commission, Community courts and national courts) without any major problems arising, so it appears that there cannot be anything per se or fundamentally objectionable in a system of laws having direct effect throughout the Community. I suggest that experience shows that either notification systems or direct effect systems can work effectively when the necessary balance between inputs and outputs is maintained; in other words, if the resources provided to the relevant institutions broadly match the number of decisions required to be made each year, then either system can work. But, if the imbalance becomes sufficiently great, neither system works. It is clear that the notification system

3 - Courts and Judges

573

under Article 81 has itself broken down mainly because of the number of agreements notified over time that required negative clearance or exemption; with a Commission monopoly provided by Regulation 17/62 for the implementation of individual exemption the task of dealing satisfactorily by way of decision with all notifications within a reasonable time limit is no longer possible; indeed the same breakdown now threatens the Merger Task Force, if additional resources are not provided for it to deal with the extraordinary increase in the number of merger cases now coming before it, often of a highly complex nature. A system of direct effect, if applied to Article 81, could likewise break down if the resources made available to national courts and national competition authorities are in their turn inadequate or if the relationships between them are not developed sufficiently so as to allow a degree of synergy to occur between their respective private and public law functions. The question therefore at the heart of the modernisation proposals is surely therefore, how to evolve a system which will in the expanded Community now, and for the future, enable a balance to be struck between inputs and outputs of relevant cases. This obviously means that more resources have to be injected into the system at some point. The question is 'how' and 'where'. Individual judges have already shown themselves in national cases quite capable of deciding whether agreements fall within Article 81 (1) or within the terms of a block exemption, and also in deciding whether interim measures should be granted or whether a fine for breach of national law is reasonable. (In other words in carrying out the functions permitted under Delimitis.1) I have however grave doubts as to whether national judges would be sensibly employed under present arrangements in seeking on their own to resolve Article 81 (3) cases, trying to balance the degree to which the agreement or practice at issue actually restricts competition against the four conditions for exemption. Under the adversarial system at least the court would be highly dependent on the resources of the parties to present a full set of arguments, and the likely imbalance of resources in a case between, say, a large multinational and a sole trader, may make this difficult. Even if the judge is able to obtain a full picture of the issues, is it a sensible use of judicial manpower which is scarce, expensive and required for many other public policy tasks? (As to the latter, society may rightly regard them as having a higher priority than resolving arguments between commercial undertakings about whether a particular restraint justifies an exemption). We have interesting views on this point from two experienced UK national judges in their evidence given to the House of Lords Select Committee on the European Union Inquiry2 into 'reforming EC competition procedures' published in February 2000. The main contention of Laddie and Ferris J.J. in their oral evidence (found at pages 63-72) to the Select Committee Report was that

1 2

Case C234/89 Delimitis [1991] ECR 1-935: [1992]: 5 CMLR 210. HL Paper 33. HMSO, London.

574

Working Paper V

it was 'wholly inappropriate for national judges to deal with Article 85 (3) cases'. At page 65 Mr Justice Ferris says '. . . judges by their training can choose between rival arguments, at any rate if the rival arguments are presented in the context of a framework which lays down rules by which the judges make their choice. They cannot make value judgments except in a very limited field, certainly not in relation to general economic questions . . . the Court should not have any part to play . . . in deciding whether an agreement or a course of conduct contributes to improving the production or distribution of goods or promoting technical or economic progress... I cannot see any court as we know it making a satisfactory job of that task. One would have to devise an entirely new beast I think or . . . give the task to something like the Competition Commission with a right of appeal to a court if they go badly wrong. I would envisage a right of appeal on a judicial review basis . . . '

III. Changes required in the Community's judicial architecture In looking at the role of judges in deciding competition law cases one also needs to look not only at the problems which they face in deciding the immediate case but also at the hierarchy of the Court system within which they operate. In the case of competition law we have both the national court system and the right under Article 234 to obtain a preliminary opinion from the ECJ on Community law issues. The ECJ as stated above is now not able to deal in a reasonable timescale with the current level of requests from national courts for rulings; it would be totally unable to deal with the additional cases that would be likely to arise following implementation of the modernisation proposals. The White Paper surprisingly seems to lack any vision or practical suggestions as to how the judicial architecture and judicial organisation of the European Community might itself have to be adapted to a new situation where national courts were asked to carry out a more substantial part of the burden than previously. Just under two years ago in the third edition of my book on EC Competition Law,3 in the final chapter—entitled 'The Future of European Competition Law'—I had made some suggestions about how European competition law might need to develop, along lines not too far away from those in the White Paper. At the time I wrote those comments (in late 1997) I could not have known, of course, that the Competition Directorate was already well advanced in its own internal thinking and planning on the same subject, though the White Paper was not published until six months after the book appeared. One of the suggestions which I made (which is not in the White Paper) was that this Article 234 function should in future not be carried out by the ECJ but should be dealt with by the Court of First Instance, possibly through chambers sitting on a regional basis. The problems of retaining all such cases for the ECJ in pres3

EC Competition law 3rd edition (1998), Oxford University Press, pp. 601-602.

3 - Courts and Judges

575

ent circumstances is well known; in a recent article (18 April 2000) in the Financial Times the President of the Court says 'the situation is far from satisfactory, particularly in the area of preliminary references [Article 234]. The risk is that the situation could seriously deteriorate to the detriment of the effectiveness of justice and the confidence of those judged'. Given the critical position in which the ECJ thus finds itself I was delighted to find support for my tentative suggestion to give a much greater role here to the CFI in evidence presented to the House of Lords Select Committee inquiry referred to above from a source as distinguished as Jeremy Lever QC; in his joint submission with Mr George Peretz,4 he says: 'in our view references by a national court or tribunal under Article 234 to the ECJ would not be a satisfactory substitute for direct appeals to the CFI (a) it was precisely the unsatisfactory nature of the ECJ as a forum for hearing Competition law appeals that substantially contributed to the establishment of CFI (b) an Article 234 preliminary reference is no substitute for the extensive judicial review exercised by the CFI (c) since the establishment of the CFI the ECJ has become much less involved with Competition law than it was and is therefore even less suitable now than in the past to exercise the primary jurisdiction of the Community level in hearing Competition law cases' (para 7.3 at page 161).

In their evidence support is also given to the suggestion made above that Article 81 (3) should be administered as public law, with all appeals on substantive issues from national courts being heard by a Community Court particularly if exemptions are to operate so as to bind all parties and third parties throughout the Community. This suggestion, though itself radical, would seem to have much value.

IV. Competition cases: generalist, or specialised judges? In deciding how to use judges to resolve cases between private parties however we should pay close attention to the nature of the issues that arise there. Judicial time and talent are scarce resources, and it seems wrong to hand out such cases more or less at random to judges without wide experience of or training in the relevant economic issues. In any substantial case of this kind there are likely to be mountains of paper, many volumes of economic analysis and statistics and endless lever archfilesof correspondence and other documents, not to mention an excess of economic experts to provide reports of great length to perplex the court. The Premier League5 case held last year in the UK Restrictive Practice 4

At pages 157-163. The judgment (which is 194 pages long) was delivered on 28 July 1999: it has not yet been published as an official report. 5

576

Working Paper V

Court, though actually a public law case in the sense that the Office of Fair Trading was there seeking to challenge an agreement made by the Premier League for exclusive TV. showing of matches in that league, was a prime example of how extravagantly such cases can expand, even when heard by a specialist judge (Mr Justice Ferris) sitting with lay members of the Court. It was three years in preparation, consumed several millions pounds of public and private funds to conduct, and moreover occupied several months of judicial time. It is of course true that not all competition cases are so complex (indeed sometimes Article 82 cases by comparison can turn on relatively narrow issues) but it is undeniable that as a general rule competition law cases tend to be both complex and lengthy, particularly where large sums of money turn on the outcome. From all this it seems clear that such cases in national courts should only be heard by specialist judges, just as they are allocated for patent or tax cases and other highly technical areas (shipping or construction disputes, for example) so that issues can be more quickly ascertained and cases decided in a shorter timescale. It is my impression with regard to the UK that we already do have a number of High Court judges who, because of their experience as advocates as well as on the bench, could make admirable specialist competition law judges and whose influence in time would be felt not only at the High Court level but at the higher appellate levels. I would hope this would also in time become the case in the courts of all Member States, even if it is not yet so.

V. Linkage of public and private enforcement To provide specialist judges is however only a partial answer. To make the modernisation proposals work properly it would seem to me necessary to integrate to a far greater degree public and private resources and enforcement; and in particular to enable judges in private cases to call in expert help from national competition authorities as well as from the Community so that technical issues could be examined and reports submitted on them to the Court without requiring it to be dealt with in the full traditional common law 'adversarial' manner, thus saving a considerable amount of Court time. An example of this process that comes immediately to mind is the decision of the ECJ in the Woodpulp6 case to call in experts to prepare two reports on the alleged parallelism of transaction and list prices. The appointment of experts to prepare a report to assist a Court is, I understand, quite common in Civil Law jurisdictions. It may therefore be sensible to consider whether the prime element in future modernisation of EC competition law should be the provision of a Community judicial system which brings closer together the enforcement of public and private law, and which would enable the process of private law enforcement to be aided by the 6

Cases 89, 104, 114, 116-17, and 125-129/85 Ahlstrom and Others v Commission [1993] ECR 1-1307: 4 CMLR 407.

3 - Courts and Judges

577

ability of courts to call on the services of experts appointed by the European Commission or, when appropriate, by national competition authorities. These would help for instance in the analysis of issues, definition of relevant markets and in the production of evidence drawn from the relevant competition authority's experience. Some elements of linkage between public and private enforcement is already found in many national systems. The Clayton Act7 Section 5(a) provides in the USA that final judgment rendered in any anti-trust case brought by the US Government to the effect that a defendant has violated such laws is primafaciae evidence against such a defendant in any civil action 'as to all matters respecting which said judgment . . . would be an estoppel as between the parties'. Likewise, Section 58 of the UK's new 1998 Competition Act8 provides that all findings of fact made by the Director General of Fair Trading in cases taken by him under the Act are binding upon the parties unless subsequently overruled by the Competition Commission Appeal Tribunal or any higher court. But there is a need to take this process a step further, as suggested, by enabling competition law authorities to participate actively in private cases to help a court reach its decision inter panes with access to the policy and public law aspects of the case and of the particular sector. The European Commission itself in its own proposals does envisage some move in this direction so as to involve the Commission in private cases, which would go far beyond the limited involvement of the Commission in such cases contemplated by the original 1993 Notice on Co-operation with national courts. What is yet to be clarified however is the legal basis for such a reform and how national competition authorities could properly be brought into such a scheme. The need in principle for such a reform is clear.

7 8

Enacted in 1955. The rule does not apply to consent decrees. Which came fully into force on 1 March 2000.

VI Judge Jochem Groning National Judges in a Modernised Community Law System: A Special View to Procedural Aspects

I. General issues This article is based on the assumption that the adoption of a directly applicable exemption system will be achieved, be it as the result of an amended Regulation 17/62 or be it that the Treaty itself will be changed. Resulting from this change, the existing notification system will be abolished. The Commission will give up its monopoly of exemption. The article discusses whether the judges of the national courts will be suited to master the difficulties that switching to a directly applicable exemption system may cause. One of the main concerns in this context is that judges might fail to deal properly with the complex economic analysis that is generally required in antitrust cases, especially if this is accompanied by the task of applying a rule they are not at all familiar with, i.e. Article 81 (3) EC Treaty. Could this be a source of disturbance of such alarming proportions that it would be preferable to put the application of Article 81 as a whole into the hands of administrators in the national competition authorities, even though they might not be immune to national interest considerations? Personally I consider the application of Article 81 as a whole by the national judges not only a manageable task, but first and foremost an extremely interesting one. I admit that my optimistic point of view might be influenced by the fact that for six years I have been sitting in a court specialised in hearing antitrust cases. The Kammergericht Berlin is, in the German hierarchy of courts, a court of appeal. As a court of appeal, its 'cartel panel' (panel of three judges) decides according to civil procedural law in civil lawsuits with a competition background (boycott, discrimination by dominant undertakings, entrance to essential facilities, etc.). The cartel panel also decides on the appeals of undertakings against administrative actions of the antitrust authorities, especially the federal cartel office (Bundeskartellamt).' These administrative antitrust cases are decided according to the rules of administrative law, although the court itself is traditionally for reasons of supposed economic competence settled in the civil jurisdiction. A substantial number of these cases are merger cases. I refer to this aspect here because the decision of merger cases is based on an analysis that is, to a certain extent, comparable to the analysis 1 After the Bundeskartellamt moved to Bonn last autumn, the jurisdiction for these cases is no longer in Berlin.

580

Working Paper VI

required by the application of Article 81 (3). It concerns the question of whether the merger may establish or increase the dominant position of an undertaking on the relevant market. Although generally not as complex as in Article 81 (3), this consideration also requires certain economic assessments. A decision in a merger case is, in the words of Mr Justice Ferries, a value judgment in relation to general economic questions.2 Therefore I do not see the same fundamental obstacles in dealing with Article 81 (3) as substantive law, as does Mr Justice Ferries. It only seems to me that his concerns—and this is a very important aspect—refer to a reform assigning the leading roles to the litigants and not allowing the judges to investigate. This point touches upon the main doubts I entertain with regard to the national courts and judges in a modernised EC competition law system. These concerns are mainly procedural. From the point of view of a judge, the application of substantive law is always only one side of the coin. The other side is the procedural frame that has to be applied in the individual case. If substantive and procedural provisions are not completely compatible, the realisation of substantive law might be hindered. Therefore I would like to emphasise, in my remarks, some procedural aspects of the decentralised application of Article 81 as a whole. To come to the point directly, the application of Article 81 (3) is essentially an administrative procedure. There is no problem in verifying the application of Article 81 (3) by the national competition authority in the course of law in administrative courts (especially special competition courts such as exist in Germany). But civil courts are not completely suited to the application of this rule. In addition to this, it does not seem to be desirable that a national competition authority and a civil court should be able to deal with the same case in parallel proceedings.

II. The application of Article 81 as a whole by national courts: general problems 1. The jurisdiction of administrative and civil courts After EC competition law has been modernised, both administrative and civil courts are supposed to have jurisdiction in antitrust cases. According to the concept of the White Paper, national competition authorities will be competent to pursue restrictive practices, and undertakings might take the initiative to challenge these practices in civil courts. After a national competition authority has issued, on complaint from a third party, an administrative decision against the parties to a restrictive agreement, the parties affected will have the right to appeal against this decision. Their 2

See Goyder (2000) at p. 5.

3 - Courts and Judges

581

appeal should lead to administrative courts. This course should also be open if the competition authority rejects the third party's complaint. In my opinion, the national competition authorities in a decentralised EC competition law system should also be competent to grant exemptions from the Article 81 (1) prohibition to notified agreements.3 In other words, I tend to believe that the possibility of notifying authorities of a restrictive practice should not be abolished at all, but only the notification as in the actual system (especially the rules in Articles 4, 6 Regulation 17/62) should be eliminated. It is not inconceivable that undertakings might be interested in 'notifying' of their restrictive practices in order to obtain an exemption from the national competition authority, and thus obtain legal security rather than waiting for a complaint against their agreement to be filed by a third party. If competition authorities are given the competence to give such exemptions, then the course of appeal to administrative courts should be open as well. According to the White Paper, the modernised EC antitrust law system will (also) empower the national competition authorities to apply Article 81 as a whole, in order to enable undertakings to invoke the direct applicability of Article 81 (3) as an argument in their defences before (civil) courts. Thus undertakings would obtain immediate civil enforcement of their restrictive practices not infringing Article 81 (3), and those damaged by illegal agreements could obtain compensation more quickly.4 As to the procedural aspects of the application of Article 81 as a whole, and the special problems of the judges in this context, it has to be taken into account that there is a significant difference between a judge who is allowed/obliged to conduct investigations himself and a judge of civil lawsuits where only the parties present the facts. According to the German codes of procedure, for example, the judge in an administrative action is empowered and obliged to conduct investigations on the facts ex officio. By contrast, civil lawsuits are generally dominated by the litigants and they decide which facts shall be presented (or not), and the judges cannot correct this. With respect to the quality of the judgments, the application of Article 81 could be distorted in civil lawsuits if the litigants were to be unable (or unwilling) to disclose all the economic and competitive consequences of a restrictive practice. It is true in this context that the Commission claims it is empowered to intervene in trials as amicus curiae, in the same way as the Bundeskartellamt can do under German law.5 The Bundeskartellamt may especially contribute to the statement of facts and means of evidence in civil lawsuits. But it is doubtful that this would be a satisfactory contribution to the protection of competition as an institution. It would absorb a lot of the Commission's resources to be continuously informed about competitive conditions in a huge variety of different 3

I do not think the system of 'positive decisions' as described in White Paper para, 84 will be satisfying; see, as opposed to this, Ehlermann (2000), at para. 92 fT. White Paper para. 100. 5 White Paper para. 107.

582

Working Paper VI

(national/regional) relevant markets and to send representatives to give judgments dealing with all kinds of restrictions in 15 to 20 Member States. The aim of the White Paper, to improve the conditions for the Commission to act rather than react,6 would immediately be undermined.

2. The types of civil lawsuits Civil courts would have to apply Article 81 in different types of lawsuits: • Contractual liability proceedings. One of the parties to an agreement might stop performing its contractual obligations if they are no longer favourable to its own interests. If the other party sues for (further) enforcement the contract or to obtain damages, the first party could defend itself with an argument that the agreement infringes Article 81. • Non-contractual liability proceedings between a third party and one or more parties to a restrictive agreement. The third party considers itself burdened by this agreement and might initiate an action in order to get an injunction and/or to obtain compensation.7 In such cases, the defendant(s) could deny that Article 81 (1) is infringed, or they might claim that Article 81 (3) is applicable. In order to anticipate an action of the latter kind, the parties to the restrictive agreement (or one of them) could bring an action for a (negative) declaratory judgment against the third party. In such cases, the judgment could state whether the agreement infringes Article 81 or not. The most frequent type of trial would probably be the application for an injunction.8 The undertakings will be much more interested in putting a stop to restrictive practices that could affect their own activities in the future rather than getting compensation for damage caused in the past. This is due to the fact that it is generally difficult to prove causality between particular damage and a certain restriction of competition. The undertakings will be reluctant to run the financial risks of such trials and will prefer to improve their market position in the future.

6 7

White Paper para. 44. Applications for interim injunctions are not of a special interest in this context and can8 be neglected here. Another question is, how the absolute number of such cases may develop. I agree with Mr Martinez Lage (Martinez Lage and Brokelmann 2000, p. 5-7) that injunctions generally could better be expected in cases based on Art. 82. But if the civil course of law would be opened for the application of Art. 81 (3), the undertakings, or better, their legal advisers, could make increasing use of this opportunity too.

3 - Courts and Judges

583

3. Territorial jurisdiction, 'forum shopping', inconsistent judgments The problem of 'forum shopping' is frequently brought up in discussions about the problems that a modernised EC competition law system might create. It is a problem of efficiency and legal security as well as a procedural problem, and therefore I would like to consider it too. 'Forum shopping', or the attempt of undertakings to bring their case before a court with a presumably 'indifferent' legal attitude to the meaning of competition, is closely connected to the rules of the territorial jurisdiction. 3.1. Contractual liability proceedings Contractual liability proceedings do not cause special problems as far as territorial jurisdiction is concerned. If both parties have their legal seat in the same Member State, jurisdiction is determined according to the national procedural law of this State. When the parties have their seats in different Member States, the Brussels and Lugano Conventions on jurisdiction and the enforcement of judgments in civil and commercial matters have to be applied. According to these Conventions, the courts of the State where the seat of the defendant is located will first have territorial jurisdiction (Article 2 of the Conventions). At the same time, the territorial jurisdiction goes to the courts of the place where the contract has to be fulfilled (Article 5 (1) of the Conventions). The problem of 'forum shopping' is rather reduced in these cases. As to the expected entrance of new Member States to the EU and the jurisdiction of their courts, these States will probably also accede to the Brussels and Lugano Conventions. If antitrust cases come to their courts, it will therefore be on the basis of the Conventions, and not because one of the litigants had an opportunity to take unfair advantage of the assumed lack of experience with such disputes. 3.2. Non-contractual liability proceedings In non-contractual liability proceedings there is, with regard to Article 81, a different point of contact for territorial jurisdiction. According to the jurisprudence of the European Court of Justice, the prohibitions of Article 81 tend to produce direct effects in relations between individuals and create individual rights for the parties concerned.9 For a third party, whose competitive actions are hindered by a restrictive practice, the infringement of Article 81 is a tort (delict). In such a case, the territorial jurisdiction is at the place where the tort was caused (Article 5 (3) of the Conventions). As the Court of Justice has ruled in several cases, this is either at the place of the causal happening (or the place where the parties of the restrictive agreement have acted), or at the place where

Court of Justice Case 127/73 BRTx Sabam [1974] ECR 51.

584

Working Paper VI

the damage occurred.10 This rule does not leave space for forum shopping manipulations. It will generally also lead to satisfying results; the problem of inconsistent judgments will probably be rather theoretical, as the following example might illustrate. Suppose a German travel agency is operating close to the Dutch border and is quite dominant in the demand of Spanish hotel beds, and it puts Spanish hotelkeepers under an obligation not to give bed contingents to certain other (cheaper) German and/or Dutch travel agencies.'' The aim of this practice is to prevent German clients from booking their trips with the cheaper German/Dutch competitors and to force Dutch tourists not to make their reservations with those suppliers. The market directly affected by this restriction is the Spanish market for hotel beds. With regard to the Spanish hotel owners, the Spanish courts have jurisdiction. Indirectly affected are the German and Dutch travel markets. A German travel agency affected by this practice may, referring to Article 5 (3) of the Brussels Convention, appeal to a German court for remedy, whereas a Dutch agency affected by this practice can appeal to the courts of its own native State.12 Both the German and Dutch courts are competent because the damage occurred in Germany and the Netherlands. Article 5 (3) of the Convention excludes only places where no more than a financial loss has occurred.13 It is not completely inconceivable that inconsistent judgments of different courts in different Member States might enter into effect. But it is more likely in such cases that participants in the different trials will engage in some communications and that the courts will hear about the parallel cases. Therefore they will harmonise their judgments, at least by preliminary rulings.

4. The jurisdiction of national courts and the competence of the Commission The Commission assumes that the prosecution of certain cases will remain to be assigned to it, whereas others should be pursued by the national authorities/courts. This raises the problem of how the different competencies have to be delimited. The White Paper's proposals in this respect start from whether the effects of a disputed practice are felt primarily in one Member State or 10 Case 21/76 Mines de Potasse, NJW 1977, 493; Case C 364/93 Antonio Marinari v Lloyds Bane pic a. Zubaidi Trading Company, EuZW 1995, 765. 1 ' Example leaned on the TUI and NUR cases pursued by the Bundeskartellamt; see BGH Urteil vom 7.10.1997 KVR 14/96, WuW DE R 89 Selektive Exklusivitat. 12 Although it may be preferable to go directly to a German court in order to avoid the procedure of recognition of the Dutch judgment in Germany. 13 Art. 5 (3) is not understood as a 'money pocket rule' ('damage was suffered in my pocket'); see Court of Justice Case C 364/93 Antonio Marinari v Lloyds Bane pic a. Zubaidi Trading Company, EuZW 1995, 765.

3 - Courts and Judges

585

Community wide.14 However, this principle will not necessarily be compatible in each case with the current rules for the territorial jurisdiction of the civil courts, and especially not with the Brussels and Lugano Convention. A certain court is either competent according to these rules or not, but its given jurisdiction cannot be cancelled because the effects of the case are felt Community wide. The jurisdiction of the civil courts instead follows the principle that, if civil courts are invested with the competence to apply Article 81 as a whole, their jurisdiction cannot simultaneously be limited to cases where the effect of a restrictive practice is felt in only one Member State. Apart from this procedural aspect, one should take into account the fact that the markets of several (or at least two) Member States will very often be affected and yet it will not always be justified to assume a Community wide interest. For example, in the TUI and NUR cases the Commission denied the existence of a Community wide interest although several markets in different Member States were involved.

5. Inconsistencies (courts and Commission/national authorities) The White Paper also takes up the problem of inconsistent decisions.15 The principles of resolution pointed out by the Commission16 will generally be satisfactory. Yet some aspects should be further considered. If a case with an international dimension has been brought to a national civil court, there is no possibility for the Commission to intervene. Consequently, in the White Paper the Commission only demands the right to take a case out of the jurisdiction of the national competition authorities, but not to take it out of the jurisdiction of the courts.17 With regard to the national courts, the Commission envisages only the possibility of intervention in individual cases as amicus curiae.l& Stepping into this role definitely excludes the right to take the case out of the jurisdiction of the court because otherwise the Commission would play the part of a participant to the trial and the role of the amicus curiae at the same time. On the other hand, there can be no doubt that the Commission is not prevented from opening an antitrust proceeding against parties to a restrictive agreement for the simple reason that this agreement is already subject to a civil action. But what would the outcome be if the parties to a restrictive agreement, that is already the subject of a complaint before to the Commission, brought an action against the complainant for a (negative) declaratory judgment to the national civil court? In the worst case, the civil lawsuit on the one hand and the 14 15 16 17 18

White Paper para. 97. Para. 101. White Paper para. 102. White Paper para. 105. White Paper para. 107.

586

Working Paper VI

Commission's antitrust proceeding on the other could end up with contradictory results. It is true that the Commission, referring to the Delimitis judgment, asserts that the national courts were obliged to avoid conflicting decisions as soon as the Commission initiated procedures or even adopted a decision.19 Yet it has to be taken into consideration that the Delimitis case was ruled under the conditions of the present legal system, which prevented the national courts from applying Article 81 (3). As soon as they are empowered to apply Article 81 as a whole and the present boundaries of the have disappeared, the national courts will apply this rule along with the Commission. Even if it is not very likely that national courts will seek confrontation with the Commission, it has to be taken into account that, under the conditions of a reformed Community law system, there would be no justification for the civil court to suspend the proceedings (apart from requests for preliminary rulings) or even to 'ask the Commission for information'.20 To guarantee that such inconsistencies do not occur, an amended Regulation 17/62 should provide for a 'suspension' of a concurrent application of Article 81 by the Commission and the national courts in certain cases. Similar inconsistencies might occur in the relations between the national competition authorities and the civil courts.

6. Influence on the application of the Community law by the Commission As far as the juridical problems of the application of Article 81 by the national courts are concerned, the Commission can have a direct and obligatory influence through regulations only, because they are part of the national law (Article 249 (3) EC Treaty).21 The notices and guidelines that the Commission intends to draw up to explain its policy and provide guidance on the application of Community competition rules will not be binding on the national courts.22 The Commission intends to reserve the possibility of adopting individual decisions that are not prohibition decisions ('positive decisions').23 These will, supposedly, be decisions within the meaning of Article 249 (4) EC Treaty. As these are only binding on the parties addressed by the decision, the national courts are not forced to 'apply' such decisions in similar cases involving different parties.

19

White Paper para. 102. White Paper para. 102. 21 T h e controversy surrounding the question of the extent to which the block exemption regulations really are obligatory (see Rehbinder in Immenga and Mestmacker (1997) para. 17 ff./24 ft".), can be neglected in the context of interest here. 22 This is obviously the opinion of the Commission too, see White Paper para. 86. 23 White Paper para. 88. 20

3 - Courts and Judges

587

III. The main problems in the application of Article 81 (1) by national courts 1. General aspects National (civil) courts are not completely unfamiliar with the singular elements of Article 81 (1). They already have jurisdiction to directly apply Article 81 (I),24 and they might even have gained some practical experience in applying this rule. If not, they are at least familiar with most of the legal aspects of Article 81 (1). Many agreements whose compatibility with Article 81 stands in question will, at the same time, be doubtful as far as national competition law is concerned. Article 81 applies only to agreements distorting inter-state trade, and this element is different from national competition laws. Yet its application should not cause insurmountable problems for national courts. In this context there is a problem of contradictory judgments that could lead to unsatisfying results. A restrictive practice that has no effect on inter-state trade can be prohibited without the possibility of exemption if national competition law is infringed. But as soon as the effects of this practice are manifest in inter-state trade—which will generally have to be considered as an even more serious infringement—the parties to this practice may successfully claim the benefit of Article 81 (3).

2. The burden of proof Under civil procedural rules, the problems of applying Article 81 (1) are principally for the litigants—first and foremost for the plaintiff—rather than for the court. The plaintiff has to specify and—if necessary—provide evidence that all the requirements for the application of Article 81 (1) are fulfilled. As previous experience has shown, private litigants often find it difficult to fulfil these requirements although they act on the markets. This is due to the fact that they do not have the same possibilities to investigate as competition authorities and courts in administrative proceedings. Contrary to the private litigants, the competition authorities are equipped with the staff and have the legal competence necessary to conduct investigations as to the competitive conditions within individual markets.

If it is beyond doubt, Art. 81 (3) is not applicable to the case in question.

588

Working Paper VI

3. The relevant market and market shares The difficulties in this context start with defining the relevant product and geographical market. Difficulties increase if solving the case means ascertaining the actual market shares. From my experience in dealing with antitrust cases— either in civil or administrative proceedings, or applying the national or the Community law—definition of the relevant market is normally one of the central, and often delicate, factual problems. For the private parties it is a serious procedural burden to present all the facts that are required to judge the market conditions. Moreover, the current economic developments will bring about a variety of completely new relevant markets, difficult to define and delimit by the undertakings as well as the courts. Another factual difficulty for the claimant concerns the restrictive practice/agreement itself. If the parties to this agreement deny its existence, the claimant has to prove that the contrary is true. Parallel to the problems of the parties to the civil lawsuit, it is likely that the majority of the national civil courts are only to a certain extent experienced in dealing with the specific problems connected with the demarcation of the relevant markets and assessing the actual competitive conditions, even if they are used to applying national competition law. This is partly due to the fact that the territorial jurisdiction is probably wide spread. Even in Germany, where the law provides for a concentration of jurisdiction in antitrust cases (see next paragraph), individual courts will only have dealt with antitrust cases once in a while. In addition, the application of national antitrust law does not necessarily include a detailed knowledge of market conditions in every case. A good number of the cases are disputes between the parties to an agreement, and they deal with the question of whether the agreement is void because it infringes national antitrust rules. In these cases, market definition is generally not an essential problem for the judgment. The same holds true for cases of boycott. As far as German antitrust law is concerned, market data are required (apart from merger cases) especially in cases of discrimination by an undertaking that holds a dominant position on the relevant market and in diverse types of horizontal cartels. But this latter kind of restriction is more often the objective of a complaint brought before the (federal) antitrust authority, rather than object of a civil lawsuit; and thus the complainant delegates all problems to the competition authority. Therefore, if the national civil courts are to be empowered to apply Article 81 as a whole, it is desirable that national procedural laws provide for the highest degree of concentration in jurisdiction over civil competition cases. The smaller the number of competent courts, the faster they will accumulate experience in dealing with these cases and the more quickly a uniform application of the Community law will grow. The German procedural law already empowers the Lands of the Federal Republic to concentrate the jurisdiction in competition cases (including the application of the Community law). The Lands have availed themselves of this right and therefore already in the first degree only a

3 - Courts and Judges

589

small number of courts are competent. At the level of courts of appeal, the degree of concentration is even higher.

IV. Special problems in the application of Article 81 (3) by national courts The central problems for the civil courts in applying Article 81 as a whole relate to paragraph 3. It seems to me that civil lawsuits are not completely suitable for the application of Article 81 (3), for two reasons.

1. Basic unsuitability of civil lawsuits for the application of Article 81 (3) First, the application of Article 81 (3) requires a thorough consideration of the circumstances, starting from conditions of the relevant market and going through the benefits of an eventual exemption in terms of improving production or promoting technical and economic progress, and benefits to the consumers concerned. With regard to this, the main disadvantage of civil lawsuits is that civil courts do not conduct investigations on their own, but only make use of the facts the parties have presented. Therefore, the consideration might not cover all the circumstances if the parties do not state the facts completely. Or, even worse, the judgments passed might be incorrect because, for some reason, one party has not contested the facts stated by the other. A counter-argument to the above could be that—in the present system—the undertakings have to convince the Commission or the judges in Luxembourg that the conditions of Article 81 (3) are met,25 and that my concerns only refer to a change in the burden of proof among equal private undertakings. This might be true, but I wonder in how often the claims of undertakings were crowned with success if the Commission maintained the contrary? The second procedural problem seems to be more delicate. Exemptions under Article 81 (3) are granted a temporal limitation. This is not only due to the fact that Art 8 (1) of Regulation 17/62 provides such a limitation. The existence of this rule is rather the urgent consequence of the nature of exemptions. It is in the nature of things that an exemption from Article 81(1) can be granted only for a certain period of time.26 Conditions in the relevant market might develop in such a way that a future cancellation of an exemption can become inevitable. Thus, as a matter of principle, exemptions cannot be granted 25

Ehlermann (2000) para. 74 f. Exceptions are claimed only for certain constellations, see Ritter in Immenga and Mestmacker (1997) para. 3. 26

590

Working Paper VI

without temporal limitations (and without the possibility of withdrawal). It therefore seems to me that decisions with temporally limited effects can be taken only by a competition authority (and so can be open to attack only in administrative courts, not in civil ones). Determining the duration of an exemption in an individual civil case would be foreign both to the nature of exemptions and the nature of civil law. The application of 8 (3) Regulation 17/62 or a similar rule would be foreign to the basics of civil procedural law. As for non-contractual proceedings—probably the most frequent type of lawsuit in a decentralised Community competition law system—the incompatibility between the nature of civil lawsuits and the application of Article 81 (3) is quite obvious as well. In a civil lawsuit, the plaintiff will assert that an exemption cannot be granted (at all) and therefore he will generally demand a refusal of an exemption. But if a temporal limitation is in the nature of exemptions, he would be obliged to claim the refusal of a temporally limited exemption. Otherwise his claim would prove not entirely successful, and consequently he would have to be ordered to pay an appropriate part of the costs. By contrast, the defendant will claim indefinite exemption, but for the same reasons he should only claim for a limited one. Yet it seems to me problematic to impose upon the parties of a civil action the duty to set such a kind of limitation to their claims. The duration of an exemption is not part of the enforceable claim of a party. How can the parties know what is an appropriate length for an exemption? Apart from the procedural impossibility of deciding on the temporal limitation of an exemption in a civil lawsuit, it would not even be desirable to empower the parties in such a way at all. To me it does not seem to be within the meaning of upholding competition to empower either the plaintiff or the defendant with the authority to decide on the duration of an exemption from Article 81 (1). This would better be left to the national competition authorities (whose competence to establish temporary limitations or withdrawal of exemptions would be in accordance with the spirit of competition law).

2. Procedural implications The partial incompetence of the civil courts to deal with Article 81 (3), as described above, could have implications for what the European Court of Justice ruled in Delimitis:21 • If Article 81 (1) is obviously not infringed, the civil court can decide itself. • If Article 81 (1) is infringed, but a block exemption regulation is applicable (especially the new type of regulation described in the White Paper at para. 71), the jurisdiction of the civil court is also sufficient. 27

Case C 234/89 ECR I 935 paras. 43-54.

3 - Courts and Judges

591

Problems occur if the granting of an individual exemption is in question. In my opinion, the civil court has to suspend the proceeding because it does not have jurisdiction. The court has to leave it to the competition authority to decide whether a—temporarily limited—exemption should be granted or refused. In practical and procedural terms, this means that the decision of the competition authority replaces the judgment of the civil court. As soon as the authority has granted or refused the exemption,28 the main issue of the civil action is settled. The civil court is only left to decide on the costs of the trial. Another aspect should be taken into consideration: suppose a civil court had granted an (un)limited exemption, whereas the national authority comes to the conclusion that the practice in question should be prohibited? If both the courts and the competition authorities are enabled to apply Article 81 at the same time, contradictory decisions could result. An undertaking willing to take action against a restrictive practice should think carefully about how to proceed. It should only take a civil lawsuit if the application of Article 81 (3) is undoubtedly unnecessary. Otherwise, it would do better to complain to the national competition authority from the outset. This would avoid loss of time, and the risk of being penalised with additional costs will be minimised.

V. Increasing number of requests for preliminary rulings? Being authorised to apply Article 81 (3) might cause some disconcert among national judges. The litigants will contribute a good part to this. Antitrust disputes are very often placed in the hands of experienced and skilful lawyers, defending the interests of their parties with remarkable competence. Such lawyers will petition the court to requests preliminary rulings as soon as this seems to be favourable to their own party's position. The judges, owing to the above mentioned disconcert, will often find it difficult to refuse, even if the request is not urgent. In order to avoid a 'collapse' of antitrust jurisdiction, the judges should bear in mind that preliminary rulings are not necessary if the right application of the Community law is beyond doubt. As for the application of Article 81, a wide range of precedents (decisions by the European Court of Justice) already exist. Apart from this, regulations, Commission notices and guidelines,29 and decisions will contain a lot of valuable indications for interpreting Community law. The judges should, by all means, examine carefully whether the correct interpretation of the Community law really requires a preliminary ruling. In particular, this includes an examination of whether a dispute really concerns the interpretation of Community law or just factual aspects of its 28 29

Legal remedy included. See White Paper Paragraph 86.

592

Working Paper VI

application. It is not unusual, while reading the judgments of the European Court of Justice, to get the impression that the national court was only unsure about problems of the application of the Community law rather than about juridical questions. Factual problems, however, have to be solved according to national law. In this context it should be emphasised that the application of Article 81 (3) primarily demands a consideration of, and a prognosis on, present and future market conditions. Therefore national courts should only make reserved use of the request for preliminary rulings, especially as long as their own judgment can still be appealed (Article 234 (2) EC Treaty).

VI. Conclusions It appears that civil proceedings are not completely suited for the application of Article 81 as a whole. In my opinion, the competence of civil courts fails at least when individual exemptions must be granted. Neither temporary limitations nor the possibility of withdrawing exemptions is compatible with substantive and procedural civil law. Therefore, one should be advised against assigning the civil courts with the application of Article 81 (3); one should instead be in favour of the exclusive competence of the national competition authorities to apply Article 81 as a whole. This concentration of competence would support a quick and uniform application of Article 81. The national competition authorities would gain experience in dealing with Article 81 (3) in a comparatively short period of time without being obliged to request preliminary rulings. Even though I plead for such an assignment of competencies, it should be left to the individual Member States (and, in a mediated way, to undertakings) to decide which system to install. From my point of view there is no legitimacy to excluding from the outset recourse to civil courts in Article 81 cases only because of the assumed lack of competence to apply Article 81 (3). It is only a question of procedural economy to assign these cases to national competition authorities. The problem of susceptibility to national interests on the part of administrators should not be disquieting as long as the competition authorities' decisions can be reviewed by (independent) administrative or (better) specialised courts. These courts would also become familiar with the application of Article 81 (3) relatively quickly. Thus it should be possible to bring the cases to an end within an acceptable period of time. This would be an additional advantage to the widespread jurisdiction of first-degree civil courts and the perspective of appeals to several instances.

VII Santiago Martinez Lage and Helmut Brokelmann* Article 81 (3) Before National Courts: The CAMPSA Doctrine of the Spanish Supreme Court and Articles 84 and 85 Revisited

I. Introduction The Commission's White Paper on the modernisation of the rules implementing Articles 81 and 82 proposes to replace the current notification and authorisation system for restrictive practices with a legal exception regime in which the third paragraph of Article 81 would 'become' directly applicable in the Member States. This revolutionary proposal implies that the Commission will abandon the monopoly to grant exemptions, which Article 9 (1) of Regulation 17 has conferred on it since 1962, and gives both national competition authorities as well as national courts the power to apply Article 81 (3). The relinquishment of the Commission's exemption monopoly in favour of a decentralised application of this provision by national competition authorities would not seem to raise fundamental problems as these authorities are generally well prepared to apply Article 81 (3). However, the direct applicability of this provision by national courts raises problems that call for a detailed assessment of the implications of such direct application of Article 81 (3) by national judges. The purpose of this article is to analyse the specific problems arising in the context of the new role attributed to national courts in the Commission's White Paper and to explore possible solutions to ensure a coherent decentralised application of Article 81 (3) by the Member States in the context of the legal exception regime proposed in the White Paper. To this end, specific reference will be made to the Spanish Supreme Court's case law on the inapplicability of the competition rules by ordinary courts without a prior administrative pronouncement, as well as to Articles 84 and 85 (former Arts. 88 and 89) which, before the adoption of Regulation 17/62, seriously limited the jurisdiction of ordinary national courts to apply Article 81.

II. The role of national judges under the present system Under the current authorisation system, the ECJ has only recognised the direct effect of the first and second paragraph of Article 81 (as well as to Article 82).' * Martinez Lage & Asociados, Madrid. 1 See, e.g., Case 127/73 BRT v Sabam [1974] ECR 51 and Case C-234/89 Delimitis [1991] ECR 1-935.

594

Working Paper VII

Although, to our knowledge, the Court has not yet had the opportunity to pronounce its opinion about the possible direct effect of Article 81 (3),2 the ECJ has denied national courts the power to declare Article 81(1) inapplicable pursuant to Article 81 (3), on the grounds that Article 9 (1) of Regulation 17/62 confers exclusive competence on the Commission to adopt decisions in implementation of Article 81 (3). It is noteworthy—and we shall return to this issue at the end of this article— that prior to the enactment of Regulation 17/62 (as well as in those sectors in which an implementing regulation does not yet exist) the ECJ imposed significant limits on the jurisdiction of national courts to apply thefirstand second paragraphs of Article 81 pursuant to the direct effect of these provisions. According to the Bosch judgment—later confirmed in the Asjes and Ahmed Saeed rulings3 relating to the air transport sector—in the absence of a procedure giving effect to Article 81 (3), ordinary national courts do not have jurisdiction to apply Article 81 (1) and (2) unless either the competent national authority [pursuant to Article 84 (former Art. 88)] or the Commission [under Article 85 (2) (former Article 89)] has previously declared that the agreement in question is contrary to Article 81 (1) and cannot be exempted pursuant to Article 81 (3). Since the BRT /judgment it is clear that the 'national authorities' within the meaning of Article 84 are either the administrative competition authorities of the Member States or specialised courts specifically entrusted with the task of applying national competition law. In no case they are ordinary national courts.4 Since the entry into force of Regulation 17/62, the (horizontal) direct effect of paragraphs one and two of Article 81 enables the national judge to establish whether an agreement is caught by Article 81(1) and thus void pursuant to the second paragraph of this provision. Furthermore, the national judge also has the power to decide on the applicability of block exemption regulations, which according to Article 249 are directly applicable due to their very nature as regulations. The Court's rather broad interpretation of Article 81 (1) has given national judges a certain room for the application of this provision without having to resort to the third paragraph of Article 81. Such a broad—and thus 'decentralised'—interpretation of Article 81 (1) (that contrasts with the Commission's hitherto narrow interpretation based on the so-called freedom of 2 The ECJ was invited to discuss this issue in the context of two preliminary references by the German Kammergericht questioning the compatibility of Article 9 (1) of Regulation 17/62 with Article 81 EC on the basis that the first and third paragraph of Article 81 constitute and indivisible whole. Both preliminary references were, however, later cancelled from the Courts' Register (Cases C-365/96 Ruhrgas-Thyssengasl Bundeskartellamt OJ C 9/12 of 11.1.1997, cancelled by Order of 26.3.1998; C-34/97 RWE\ Bundeskartellamt, OJ C 94/7, of 22.3.1997, cancelled by Order of 13.10.1998). 3 Case 13/61 Bosch [1962] ECR 89; Cases 209 to 213/84 Asjes [1986] ECR 1425; Case 66/86 Ahmed Saeed [1989] ECR 803. 4 The BRT /judgment, supra note 1, in this respect overruled the Bilger judgment, Case 43/69 [1970] ECR 127, Rec. 1970, p. 127, para. 9.

3 - Courts and Judges

595

action theory, which ignored any analysis of the economic and legal circumstances surrounding the case) can be seen in early judgments such as Societe Technique Miniere and Consten/Grundig as well as in Pronuptia, the Metro cases, Nungesser and Delimitis.5 These judgments give national courts some room for manoeuvre in the interpretation and application of the notions of 'restriction' and effect on 'trade between Member States' contained in the first paragraph of Article 81. Nonetheless, the ECJ has imposed certain restrictions on the national courts' powers to apply Article 81(1) and the block exemption regulations. Among the various judgments addressing this question, the most systematic precedent is the Delimitis judgment6, in which the Court established that the general principle of legal certainty obliges courts to avoid conflicting decisions which might arise as a consequence of the shared competence between the Commission and the national courts to apply these provisions of Community competition law. Hence, in Delimitis the ECJ held that national courts must not give decisions on restrictive practices which may subsequently be the subject of a decision by the Commission. For these purposes the following limits on the jurisdiction of national courts were established by the ECJ. Where the conditions for the application of Article 81 (1) are clearly not satisfied, a national court may proceed and rule on the agreement as there is scarcely any risk of the Commission taking a different decision. The same applies if the agreement's incompatibility with Article 81 (1) is 'beyond doubt' and the agreement may 'on no account' be the subject of an exemption decision under Article 81 (3), regard being had to the exemption regulations and the Commission's previous decisions. If, on the other hand, the national judge considers, in the light of the Commission's rules and decision-making practices, that the agreement before it may be the subject of an exemption decision—and the agreement fulfils the formal requirement of notification (or those of Article 4 (2) of Regulation 17/62)—the judge should stay the proceedings or, at the most, adopt interim measures pursuant to its national rules of procedure. The same applies if there is a risk of conflicting decisions in the context of the application of Articles 81 (1) or 82. In these circumstances the national court may seek information from the Commission on the state of any procedure that the Commission may have set in motion and as to its likely outcome. The national judge may contact the Commission if the specific application of Articles 81 (1) or 82 raises particular difficulties, 'in order to obtain the economic and legal information which that institution can supply to it'. 5

Cases 56 and 58/65, Consten and Grundig v Commission [1966] ECR 429; Case 56/65 Societe Technique Miniere v Maschinenbau Ulm [1966] ECR 337; Case 26/76 Metro v Commission I [1977] ECR 1875; Case 75/84 Metro v Commission //[1986] ECR 3021; Case 161/84 Pronuptia [1986] ECR 353; Case 258/78 Nungesser [1982] ECR 2015. 6 Supra, note 1. See also Case 48/72 Brasserie de Haecht [1973] ECR 77; Case BRTv Sadam, supra note 1; Case 37/79 Marty Lauder [1980] ECR 2481.

596

Working Paper VII

This case law of the ECJ shows that under the present authorisation regime there are serious limitations on the national courts'jurisdiction to apply Article 81 (1) and the (directly applicable) block exemption regulations, which derive from the impossibility of applying the third paragraph of Article 81. Although they have jurisdiction to apply these provisions—subject to the possibility (or duty) to refer to the ECJ for a preliminary ruling under Article 234 if a question of interpretation within the meaning of the Court's CILFIT judgment7 arises—the impossibility of applying the third paragraph of Article 81 has hitherto hindered the effective application of the first two paragraphs of Article 81 by national courts.

III. The new role of national courts as envisaged in the White Paper With the abolition of the Commission's exclusive right to grant exemptions pursuant to Article 81 (3), the White Paper—assuming that the ECJ indeed confirms that this provision does confer rights which are directly enforceable by private parties in proceedings before national courts—proposes to entrust the application of this provision to both national competition authorities and national judges. The Commission argues that the parallel competence of national courts to enforce Article 81 in its entirety will allow both the Commission and national competition authorities to focus on important cases affecting the public interest. The White Paper therefore clearly favours the concurrent enforcement of Article 81 (3) by national courts. This can also be seen in the Commission's rejection of the option to limit the decentralisation of Article 81 (3) to an enforcement of the provision by national competition authorities. Regrettably, the White Paper only discusses this option in the context of maintaining the current authorisation regime and does not contemplate the possibility of adopting a legal exception system without giving ordinary national courts the power to apply Article 81 (3). There are undoubtedly advantages of allowing the direct applicability of Article 81 (3) by national courts. It would allow judges to settle private disputes with speed and immediacy as parties would no longer be able to thwart court proceedings by notifying to the Commission the agreement in question. This un-blocking of national judicial proceedings would increase the attractiveness of private enforcement as it would enhance the possibilities for companies that are victims of restrictive practices to get judicial relief. At the same time, undertakings would have a new defence against claims of infringement of Article 81 (1), claiming that the conditions of Article 81 (3) are fulfilled. Hence, under the new regime companies would immediately be able to enforce clauses containing restrictions of competition capable of being exempted. Case 283/81 CILFIT[\982] ECR 3415.

3 - Courts and Judges

597

Through the grant of interim relief, national courts would also be able to act quickly to assist complainants. Nonetheless—with the exception of cartels, which in any case will be less affected by the abolition of the authorisation system—injunctions are more likely to be applied in Article 82 situations than in the context of the enforcement of agreements under Article 81. National courts will also be able to apply Article 81 in its entirety in the context of private actions for damages sought by victims of violations of competition law. Here again, such actions for damages would seem more likely in the context of Article 82 or of cartels, rather than in respect of agreements which have hitherto been notified to the Commission. These advantages must, however, be weighed against the risks of entrusting the application of Article 81 (3) to ordinary, i.e. non-specialised, national courts, which, as will be discussed below, lack the expertise and means to enforce this provision from a competition policy perspective and cannot guarantee its coherent application throughout the Community.

IV. Administrative decentralisation vs. 'judicialisation' Before discussing the specific problems faced by judges in the new enforcement regime proposed in the Commission's White Paper, a fundamental distinction should be made between, on the one hand, the decentralised administrative enforcement of Community competition law by national competition authorities and, on the other hand, the judicial application of these rules by the courts of the Member States. The White Paper aims at establishing a new system of concurrent jurisdiction in which the burden of enforcing Article 81 is shared between the Commission and the Member States. As regards the latter, the White Paper distinguishes, albeit only to a certain extent, between national competition authorities and national courts, which would both have jurisdiction to apply a directly effective Article 81 (3). This distinction remains formal without reaching its ultimate consequences because national competition authorities and courts are ultimately treated alike for the purposes of the decentralisation envisaged in the White Paper, even though their functions differ in various respects. In the first place, there is a significant difference in the perspective of competition authorities and judges when confronted with issues of competition law. Whereas competition authorities are primarily (if not only) concerned with questions of competition policy, a national court is focused on resolving the dispute between the parties before it—mostly in the context of contractual litigation between the parties to an agreement—without having regard to third parties' interests or even to the wider implications of a restrictive practice for competition in the marketplace. The White Paper expressly acknowledges the public policy role to be carried out by competition authorities as opposed to the role of judges:

598

Working Paper VII

'The actions of competition authorities, at both [a] national and Community level, are guided by considerations of public policy in the economic sphere: unlike the national courts, they do not set out to decide disputes between parties, but rather to guarantee the maintenance of a system ensuring that competition is not distorted'. (White Paper para. 91).

This different perspective also has to do with the different moment at which competition authorities and courts deal with a given agreement addressing issues under Article 81. Under the new regime proposed in the White Paper, national competition authorities would apply Article 81 either ex officio or following a complaint by third parties once the agreement has been made public. At this stage (as is the case under the present system following the agreement's notification), the parties to the agreement are generally in full harmony defending their contract against third parties' and the authorities' objections. Only very exceptionally will one party to an agreement approach the competition authority (instead of an ordinary court) when in disagreement with its contractual partner over the interpretation or implementation of the agreement in question. By contrast, courts are usually confronted with issues of competition law at a later stage, when the parties to an agreement disagree on specific issues. Most cases reaching national courts concern contractual litigation in which one party claims (sometimes artificially) the unenforceability of a contract due to its incompatibility with Article 81. In our experience, non-contractual litigation launched by third parties on the grounds of violation of the competition rules instead concerns infringements of Article 82, and only exceptionally of Article 81, for instance in the case of cartels (which are not affected by the White Paper's reform proposals). Even there, it is much more likely that third parties direct themselves to the competition authority in order to discover the cartel before subsequently claiming possible damages before an ordinary court. In this context, it is worthwhile to discuss another issue that is frequently mentioned by the defendants of an enforcement of the competition rules by ordinary courts. It is often argued that the EC Treaty generally uses national courts for the enforcement of EC law and that there should, therefore, be no obstacle to having recourse to national courts for the purposes of enforcing Articles 81 and 82. Against this reasoning it must be recalled that the vertical direct effect of certain provisions of Community law ultimately deals with the compatibility of national law with the Treaty. Although it is certainly true that national courts play an important role in enforcing directly applicable Community law, there is, however, a difference between such vertical direct effect of (primary and secondary) EC law and the horizontal direct effect of Article 81 and 82. The ultimate reason for conferring vertical direct effect on certain provisions of EC law is the Member States' failure to duly implement the respective provision into national law. The enforcement of such provisions by private parties against the State thus usually lies in the public interest of respecting the Community legal order. At the same time, the ECJ's refusal to

3 - Courts and Judges

599

give such provisions horizontal direct effect8 has much to do with its reluctance to make them enforceable also against private parties, i.e. to extend the doctrine of direct effect to situations in which a burden can be imposed on the citizen. Apart from Article 141 (former Art. 119), until now the only exception to this principle have been Articles 81 and 82 which, by their very nature, do have horizontal direct effect, although with important limits. Since the adoption of Article 9 of Regulation 17/62 it is clear that national courts cannot apply Article 81 (3). As we have seen above, this severely limits their ability to enforce the first and second paragraph of the provision and always leaves litigating parties with the possibility of obtaining an administrative pronunciation by notifying the Commission of their agreement. Before the entry into force of Regulation 17/62, the Bosch jurisprudence hindered national courts from applying Article 81 (2) without the prior pronunciation of the competent competition authority, be it the Commission (pursuant to Article 85) or the national administrative or—specialised—judicial authority (under Article 84). Until now—that is, before and after the adoption of Regulation 17/62—companies confronted with the horizontal enforcement of Article 81 before national courts have thus always had the possibility of involving the (Community or national) administration to ensure a coherent and uniform application of Article 81 throughout the Community. This provided the necessary legal certainty to the companies concerned. In the Ahmed Saeed judgment the Court held that, subject to the application of Articles 84 and 85 (former Arts. 88, 89), a price-fixing agreement completely eliminating price competition between airlines (and thus contrary to Art. 81(1) (a)) is not liable to be automatically void under Article 81 (2) until after the entry into force of the rules adopted with a view to organising the Commission's powers to grant exemptions under Article 81 (3) 'and hence to conduct the competition policy sought by the Treaty'.9 This shows that the enforcement of private rights before national courts does not necessarily coincide and can even run into conflict with the public interest of enforcing competition law. In contrast to situations of vertical direct effect where private parties enforce the rights conferred on them by Community law against a Member State, in situations dealing with issues of Article 81 private parties will not necessarily act in the public interest as the resolution of their dispute might well be contrary to the interest of competitors and other third parties not involved in the judicial proceedings. Having made this distinction between the different forms of decentralised enforcement at the level of Member States, it is clear that judicial enforcement by national courts cannot replace the administrative enforcement by competition authorities. It is therefore regrettable that the White Paper treats both methods under the same heading of 'decentralised enforcement', without 8 See, e.g., Case 152/84 Marshall /[1986] ECR 723; Case C-91/92 Faccini Dori [1994] ECR 1-3325; Case C-192/94 Cone Ingles v Cristina Blazquez [1996] ECR 1-1281. 9 Ahmed Saeed, supra note 3, para 20.

600

Working Paper VII

distinguishing between their different functions and without analysing their respective limitations. The White Paper, which only devotes a few paragraphs to the enhanced role of national courts in the application of the competition rules, seems to understand that for the purposes of decentralised enforcement it does not matter whether a given agreement is analysed before the national administrative authority or a judicial authority, thus ignoring the specific problems faced by national judges when applying Article 81 (3). Due to the different perspectives of judges and competition authorities in applying the competition rules, the judicial enforcement of these rules can by no means be an alternative to its administrative enforcement by authorities having the necessary skills and means to apply these rules in their economic and legal context. Not only can the judicial application of Article 81 not replace its administrative enforcement, but it might even run counter to the objectives of competition policy due to the different perspective of national courts when applying this provision. The question is therefore whether, apart from the administrative decentralisation to national competition authorities, there are real advantages of also entrusting the enforcement of Article 81, and in particular its third paragraph, to ordinary national courts.

V. Problems in the context of national judges applying Article 81(3) 1. Economic Analysis by National Courts The first and probably most important question as regards the application of Article 81 (3) by national courts is whether national judges are in a position to carry out the complex economic analysis which this provision requires to be properly enforced. Under the present system national courts have jurisdiction to apply Article 81 (1) as well as the block exemption regulations, albeit within the above-mentioned limits of the Delimitis judgment. Although it is true that the application of the first paragraph of Article 81 also entails—or at least should entail according to the case law of both Community courts—a good deal of economic analysis, it is also true that compared with Article 81 (3) the first paragraph contains legal concepts that can be easily applied and interpreted by national courts. This is all the more so if one were to follow the Commission's practice, according to which any limitation of the contractual freedom of economic operators is deemed to restrict competition within the meaning of Article 81 (1), subject only to the de minimis rule. For these purposes, the necessary legal guidance for national courts to apply Article 81(1) can be provided through notices, guidelines, and the precedents established by the Community institutions. This applies equally to the old system of blockexemption regulations (i.e. leaving aside the new generation of block exemp-

3 - Courts and Judges

601

tions discussed below), as their application does not require any economic analysis. In contrast, if Article 81 (3) 'became' directly applicable as envisaged in the Commission's White Paper, it would demand that national judges carry out complex economic analysis in order to determine whether the four conditions of this provision are met. As a consequence, the Commission enjoys a large margin of administrative discretion in evaluating whether a given agreement fulfils the four conditions of Article 81 (3). As the ECJ itself recognises in its case law since the Remia judgment, 'where complex economic facts are involved, judicial review of the legal characterisation of the facts is limited to the possibility of the Commission having committed a manifest error of assessment'.10 Even so, the difficulty of such assessment is underlined by cases such as European Broadcasting Union or Eurotunnel, in which the CFI annulled the Commission's exemption decision for contravening Article 81 (3). 11 The Commission itself acknowledges the large margin of discretion it enjoys as a consequence of the complexity of applying Article 81 (3) in its 23rd Competition Report: '(T)he grant of a derogation from the ban on restrictive agreements requires assessment of complex economic situations and the exercise of considerable discretionary power, particularly where different objectives of the EC Treaty are involved. This task can only be performed by the Commission.'12 In this respect, the question that necessarily has to be raised is whether ordinary national courts are able to carry out such economic analysis, which in many cases is not limited to the agreement or restriction in question but rather to the whole context of the agreement, its effect on competitors, clients and on the entire market affected by the agreement. This can be seen particularly in respect of the fourth requirement of Article 81 (3)—the elimination of competition on the relevant market—but it also applies to the other three conditions. The answer to this question is already given in the above-mentioned case law of both Community courts, as it limits their review of the economic assessment carried out by the Commission to verifying whether the rules of procedure and the obligation to state the reasons on which the decision is based have been complied with, whether the facts have been accurately stated and whether there 10 See, e.g., Case 42/84 Remia [1985] ECR 2545, paras. 34 and 48; Case T-17/93 Matra Hachette [1994] ECR 11-595, para. 104; Case C-7/95P John Deere [1998] ECR 1-3111 paras. 34; Cases 142 and 156/84 BATv Reynolds [1987] ECR4487, para. 62; Case T-39/92 Cartes Bancaires [1994] ECR 11-49, para. 109. This case law has been recently confirmed in Case T-29/92 SPO v Commission [1995] ECR 11-289, on appeal case C-137/95P [1996] ECR 1-1612. 11 Case T-528/93 Metropole Television [1996] ECR 11-649; and Case T-374/94

European Night Services [1998] ECR 11-3141. 12 XXIIIrd Report on Competition Policy (1993), point 190.

602

Working Paper VII

has been any manifest error of appraisal or misuse of powers in carrying out the analysis.13 In our view, this case law clearly shows that the European Courts do not even regard themselves as being in a position to replace the Commission's assessment of the four conditions of Article 81 (3). Bearing in mind that national courts are less familiar with both Community law in general and European competition law in particular, it would be no surprise if the ECJ also held national courts unfit to apply Article 81 (3). These limitations would also seem relevant in the context of the new block exemption regulation regime, both as regards vertical and horizontal restraints. Compared with its forerunners, the new generation of block exemptions relies heavily on economic analyses and is no longer limited to a legalistic application of black and white clauses. This can be seen in the new block exemption on vertical restraints, which often entails complex economic analyses, such as the delimitation of the relevant market or determination of the parties' market share. Interestingly, the new block exemption on vertical agreements reserves the withdrawal of the benefit of the block exemption to the antitrust authorities—both Commission and national—depending on whether the relevant market goes beyond the boundaries of a Member State. This means that the Community legislator has opted against conferring the power of withdrawal to national judges who do, however, have jurisdiction to apply the regulation as such. While they are competent to disapply the block exemption because of the inclusion of a black-listed clause in the agreement or to deny exemption for a non-compete clause exceeding thefive-yearlimit, they must not assess whether in a specific case the conditions of Article 81 (3) are not fulfilled (Art. 7 Regulation 2790/1999). Furthermore, in civil litigation, national courts can only act on the case and the facts put before them by the parties. They lack the necessary information on the structure of the market and are usually barred from carrying out their own investigations in civil law proceedings under national procedural law. National courts therefore work on the basis of a limited perspective when applying Article 81 (3) that does not permit them to properly assess the public policy considerations that are implicit in the four criteria of Article 81 (3). It has been argued that the European Courts' self-imposed restrictions as regards the scope of their review of exemption decisions derive from the present system of judicial review of the administration and that these limitations would therefore no longer apply under the proposed legal exception system. Although it is true that the principle of limited judicial review is a consequence of the present authorisation system, it is also true that it is much easier for any non-specialised court to review an administrative decision than to apply Article 81 (3) in the first place.

13 See, e.g., Remia, para. 34 and, more recently, SPO, para. 288, both quoted supra note 10.

3 - Courts and Judges

603

2. Jurisdiction to consider competition policy issues? Another issue closely related to the Commission's wide margin of discretion as regards the application of Article 81 (3) is the question—which the White Paper does not address—whether national judges would be able to consider issues of competition policy in the context of their assessment of the agreement's compatibility with Article 81 (3). Under the current system 'the Commission is responsible for the implementation and orientation of Community competition policy. [...] The performance of that task necessarily entails complex economic assessments, in particular in order to assess whether an agreement falls under Article 85(3)'.14 The application of the four criteria of Article 81 (3) clearly has public competition policy implications and decisions like VWIFordxs show that the Commission may even take other policy issues, such as industrial or environmental policy, into account for the purposes of exempting an agreement. In the context of private litigation it would, however, seem difficult to imagine a national court adequately considering the competition policy implications underlying Article 81 (3). Thus, even assuming that ordinary national courts were able to apply the four criteria of Article 81 (3), the problem remains that they can neither replace the competition authorities' permanent role of orientating and developing competition policy, nor are they in a position to enforce the competition rules from a public policy perspective.

3. Lack of power to amend a restrictive agreement Another important detriment of the direct applicability of Article 81 (3) before national courts would be the practical impossibility of amending and modifying an agreement the validity of which has been questioned before a national court or to impose limits on its duration. Whereas the vast majority of formal exemption decisions or exemption comfort letters issued by the Commission are granted following amendment of a 'notified' agreement or the assumption of certain commitments by the notifying parties, the national court applying Article 81 (3) will be unable to suggest, impose or accept modifications to the agreement in question. In civil proceedings before national courts it would therefore be impossible to carry out the necessary 'fine-tuning' of an agreement, which is today one of the essential functions of the notification process. Should a judge declare an exclusivity clause void pursuant to Article 81 (2) in the first year of the agreement's term if he considers that the overall seven year duration is excessive and does not fulfil

14 15

Delimitis, supra note 1, para. 44; see also para. 91 of the White Paper. Decision 93/49/EEC Ford-VW of 23.12.1992, OJ L 20/14, of 28.1.1993.

604

Working Paper VII

the conditions of Article 81 (3), although a five year duration would be compatible with Article 81? While national judges could only hold that a given agreement which—in its unmodified version does not fulfil the conditions of Article 81 (3)—is void pursuant to Article 81 (2), the decentralised application of Article 81 (3) by national competition authorities could assume this function of 'fine-tuning' hitherto reserved to the Commission. Such 'fine-tuning' by administrative competition authorities does not depend on the existence of an authorisation regime but can also be carried out in the framework of the new legal exemption system, be it in the course of a procedure leading to a decision accepting commitments, a non-infringement decision (following the agreements modification) or to avoid the adoption of a prohibition decision.

4. Lack of consistency and uniformity The risk of a lack of coherence and uniformity in the application of Article 81 (3) by national courts and a consequent re-nationalisation of competition policy and danger of fragmentation of the Internal Market is obvious. Basically two scenarios should be addressed in this respect. First, the risk of conflicting decisions adopted by different institutions on the same matter and, second, the incoherent enforcement of the competition rules by different courts or authorities in the absence of such conflicting decisions. There is a risk of contradicting decisions adopted by different bodies in respect of one and the same agreement that a company applies in one or more Member States. Such diverging decisions could stem from the Commission or national competition authorities but also from various courts in one or more Member States. As regards these conflicts, the Commission's White Paper limits itself to stating that the judgments of national courts (which certainly only have effects inter partes) would have the force of res judicata and would be recognised by the courts of other Member States under the Brussels and Lugano Conventions. However, other questions—which the White Paper only addresses superficially—arise as regards the effects of a declaration of validity pursuant to Article 81 (3) by a national judge or the contrary declaration that an agreement is void as it does not fulfil the requirements of the provision. Will such declarations be binding upon the national competition authority, competition authorities from other Member States, or even the Commission, if any of these authorities is faced with a complaint against an agreement previously held compatible with Article 81 by a national court? Will any binding effect of judicial resolutions depend on such a resolution being non-appealable? Or conversely, would a national judge be bound by a previous pronouncement of a national competition authority or the Commission on the agreement in question?

3 - Courts and Judges

605

In this context, the impact of decentralisation on the primacy of Community law should also be addressed. There is a risk that the legal exception system undermines the primacy of Community law in thefieldof the competition rules as the primacy rule only applies vis-a-vis national law. That is, Article 81 or decisions adopted in the application of this provision enjoy primacy over national competition law, but not over other decisions—be they by national courts or administrative authorities—applying the same provision of Community law. The Fotofrost case law of the ECJ,16 according to which national courts cannot question the validity of secondary Community law except through a reference for a preliminary ruling to the ECJ, only resolves the question of the binding nature of the decisions adopted by the Commission under the new enforcement system. It does not address, however, all those cases in which Article 81 is applied by national competition authorities or national courts. Such decisions will only enjoy primacy over contradicting acts applying national competition law but not over decisions by other courts and competition authorities applying Article 81 to the same agreement. It is therefore clear that only some of these conflicts between decisions relating to the same matter can be addressed with the principles of resjudicata, the recognition and execution of judgments in other Member States pursuant to the Brussels and Lugano Conventions, the principle of primacy of Community law as well as—in the case of conflicts with decisions adopted by the Commission—the Fotofrost case law of the ECJ. The easing of the administrative burden on undertakings derived from the abolition of the authorisation system could thus easily be replaced by a judicial burden to get approval of an agreement in several Member States. Even more serious questions of coherent and uniform application of the competition rules arise as regards the enforcement of these rules by national courts, leaving aside the problem of decisions adopted by the national or Community competition authorities that conflict with national courts' judgments. The new enforcement regime proposed by the Commission bears a serious risk that Article 81 will be applied to some companies but not to others and that, if applied by a national judge, such application will be substantially different both between the courts of one and the same Member State and between the courts of different countries. It is not difficult to imagine that, under the new regime, agreements that are equivalent from a competition law perspective will be enforceable before one national judge but not before another, within the same Member State or that the agreement is enforceable according to the courts of Member State A but not before those of Member State B. However companies need coherence and a level playingfieldin respect of what they are allowed to do and what not, particularly if they compete on the same geographical market. In this context there is also a certain danger of forum-shopping by potential complainants which would—to the extent that procedural rules allow—try to 16

Case 314/85, Fotofrost [1987] ECR 4199.

606

Working Paper VII

pursue claims before courts known to apply the most strict interpretation of Article 81. There is also a risk of multiple litigation, i.e. of companies having to defend an agreement before the courts of various Member States, particularly where a network of contracts (e.g. in the distribution sector) extends over several Member States. The danger of inconsistent applications of Article 81, and in particular its third paragraph, by national courts can, to a certain extent, be addressed through preventive measures such as unified training, assistance, guidance for courts, etc. In this respect, a reinforced role of preliminary references to the ECJ under Article 234 would also be helpful to ensure a coherent and uniform application of Article 81 throughout the Community. For these purposes one could think of making preliminary references mandatory for any national court confronted with a 'question of interpretation' concerning Article 81 within the meaning of the CILFIT judgment,17 that is, to extend the duty to make such preliminary references from courts of last instance to the lower courts. There are, however, several disadvantages of such an increased role of preliminary references. First, both Community courts are (as are national courts) already heavily overloaded. Even the assignment of preliminary references on the interpretation of Community competition rules to the CFI instead of the ECJ would therefore not diminish the long duration of this procedure and the consequent stay of the proceedings before the national judge. Second, and more importantly, the above-mentioned case law of both Community courts shows also that these courts do not regard themselves as best placed to carry out the complex economic assessment required for a proper application of Article 81 (3). To a certain extent the problems faced in this respect by national judges would also apply to the Community courts. Another possible measure could therefore be the institution of consultation procedures at the Commission or national competition authorities. In this respect the White Paper contemplates the possibility of the Commission acting as amicus curiae in proceedings before national courts. However, the danger of such consultation procedures, be it to the Commission or the national competition authority, is that the national judge might tend—also due to a lack of an own in-depth knowledge of competition matters—to simply follow the Commission's opinion, even if this were delivered on a non-binding basis. It would therefore be essential that either the parties have the opportunity to fully take part in the procedure initiated before the administrative authority as if it were an ordinary administrative application of Article 81, or that the Commission is obliged to act as an intervener in favour of one of the litigating parties before the national court. This solution seems preferable, as the parties to the judicial procedure would not be excluded from the factual assessment of their agreement by the Commission. Otherwise a mere consultative function of the Commission could raise issues of due process under Article 6 of the ECHR. 17

Supra, note 7.

3 - Courts and Judges

607

In any case, it would seem much easier to ensure a coherent application of Article 81 by the decentralised enforcement of 15 national competition authorities (or a limited number of specialised courts) than it would be to achieve coherence among hundreds of national judges whose decisions will also depend very much on the financial means each litigating party is able to invest in the judicial proceedings. Competition authorities are much less influenced by such imbalances. The White Paper contemplates withdrawal by the Commission of a case from a national competition authority. Such a safeguard would, however, seem unthinkable in the case of national courts due to the independence of judges under the constitutional laws of any Member State.

VI. Possible solutions As discussed in the previous section, there are several factors that advise against the direct applicability of Article 81 as a whole—and in particular of its third paragraph—by ordinary national courts: the difficulties ordinary courts would encounter with the application of Article 81 (3); the impossibility to apply this provision from a public competition policy perspective and even the risk of running into conflict with the competition policy developed by the Commission; and the resulting risks for a coherent and uniform application of this provision and of conflicting decisions by administrative and judicial authorities. To avoid these problems without questioning the fundamental reform envisaged in the White Paper, which is the abandonment of the authorisation system in favour of a legal exception system, we suggest several alternative proposals: the institution of specialised national courts; the need of a prior administrative pronouncement in line with the Spanish Supreme Court's case law; or a decentralised enforcement system inspired by Articles 84 and 85—which could remedy the above-mentioned disadvantages and risks of making Article 81 (3) directly applicable.

1. Specialised courts One possible remedy to the difficulties of having ordinary national courts enforce Article 81 (3) would be the creation of specialised courts in each Member State that have the necessary skills and knowledge to apply competition law. These courts would have exclusive competence to apply both national and Community competition law. Other courts confronted with the application of these provisions would have to inhibit themselves in favour of these specialised courts if the application of competition law would be relevant for the solution of the legal dispute in question. To ensure a coherent application of the competition rules by such courts and national competition authorities,

608

Working Paper VII

these courts should also be competent to review the administrative enforcement by these authorities. Such a system of concentrating cases, the outcome of which depends on the application of competition law, can already be found in the German Competition Act (§§ 87 et seq. GWB), also in respect of Articles 81 and 82 (§ 96 GWB).

2. The CAMPSA doctrine of the Spanish Supreme Court In Spain there is no possibility of directly seeking judicial relief in competition matters by bringing an action before the ordinary jurisdiction due to the socalled 'CAMPSA doctrine' of the Spanish Supreme Court.18 According to a judgment handed down by this tribunal in 1993 (a decision confirming an earlier ruling dating back to 1985), the competition rules (both national and EC) may only be applied by ordinary tribunals19 if the competition problem in question constitutes a mere incidental issue. Although the case decided in 1993 only referred to the judicial application of Article 82 and the equivalent provision of the Spanish Competition Act, thefindingsof the Supreme Court can be interpreted as referring to the competition rules in general (i.e. Articles 81 and 82 and the respective provisions of national competition law). According to this ruling, if a party's principal claim before an ordinary court (in this case, cessation of the abusive conduct and a claim for damages) is based either on Articles 81 or 82 or the corresponding provisions of national competition law, the civil court does not have jurisdiction to hear the action brought before it. According to the Supreme Court, in these cases jurisdiction to apply the competition rules is reserved to the administrative authorities in charge of enforcing these provisions; that is, the Commission pursuant to Art. 9 (3) Regulation 17/62 and the Spanish Tribunal for the Defence of Competition ('TDC which, in spite of its name, is a purely administrative authority) pursuant to the Spanish Competition Act. In reaching this conclusion, the Supreme Court held that Articles 84 and 85 (former Arts. 88, 89) entrust the competence to apply Articles 81 and 82 to the Commission and the specialised (administrative or judicial) national authorities, as does the Spanish Competition Act with respect to the TDC. Hence, the direct applicability of Article 81(1) and (2), as well as Article 82 established by the ECJ in the BRTI case, had to be understood as being limited to the incidental application of these provisions by ordinary courts, although the Supreme Court did not define the difference between such 'incidental' applica18

Case 211/1991, CAMPSA, judgment of 30.12.1993. For a detailed analysis of this judgment see Martinez Lage S. (1995): 'La aplicabilidad directa del Derecho de la competencia comunitario y nacional por los tribunales ordinarios espanoles', in Perspectivas juridicas actuates. Homenaje a Alfredo Sanchez-Bella Carswell, pp. 236-250. 19 This case law does obviously not affect the administrative tribunals' competence to judicially review the resolutions adopted by the TDC.

3 - Courts and Judges

609

tions and the 'principal' application of the competition rules. The court reached this conclusion on the grounds that the administrative enforcement of the Spanish Competition Act was one of the instruments through which the State could conduct its public economic policy, its application therefore being exclusively reserved to the State. As regards Articles 81 and 82, the court's reasoning rests on Articles 84 and 85, which will be analysed below. In a more recent judgment handed down in November 1999,20 the Supreme Court has confirmed this case law and has arguably abandoned the previous distinction between 'incidental' and 'principal' enforcement of the competition rules, so that it must now be understood that the application of these rules is in any case reserved to the competent administrative authorities. The abandonment of this distinction is to be welcomed as there is no material difference between the invocation of the competition rules by the plaintiff for the purposes of founding its claim and their invocation by way of defence to an action by the defendant. There is certainly an argument against the compatibility of the Supreme Court's doctrine with the ECJ's case law on the (horizontal) direct effect of Articles 81 (1) and (2) and Article 82. Nevertheless, the CAMPSA doctrine puts the enforcement of Articles 81 and 82 on an equal footing with that of national competition law and we have already seen that the limitations imposed in respect of the direct applicability of Article 81 in the Delimitis judgment (as well as in the Bosch judgment as regards the period prior to the adoption of Regulation 17/62) in practice make it quite difficult for ordinary national courts to apply Article 81. In any case, as things stand in Spain, no ordinary judge will be prepared to hear an action which is based on a claim of incompatibility with the competition rules. This shows that in a Member State like Spain, which has had a competition act that is seriously enforced since 1989, ordinary tribunals are not regarded as being fit to apply either national or European competition rules. If all these limitations already seem relevant for the national courts of the present Member States, the Union's enlargement to include new countries from Eastern Europe, in which judges educated in centralised economies are even less familiar with competition law issues, makes the judicial enforcement of Article 81 (3) look even more difficult.

3. Administrative decentralisation within a legal exception system Given the difficulties national courts would face when having to apply Article 81 (3) and the risks to a coherent and uniform pattern of application of Article 81 that such decentralised judicial application of Article 81 (3) would imply, the best solution would be to limit the envisaged reform to a decentralisation of this 20

Case 589/1995 United International Pictures, judgment of 4.11.1999.

610

Working Paper VII

provision to national competition authorities and/or specialised courts. This does not mean that the burdensome and inefficient notification and authorisation system should not be abandoned. Although the White Paper only contemplates—and discards—the option of administrative decentralisation within the framework of the existing authorisation system, it would also be possible to entrust the application of Article 81 (3) to national competition authorities in the framework of the new legal exception regime. This alternative reform proposal is somehow inspired by the transitional rules laid down in Articles 84 and 85 (former Arts. 88 and 89) and the ECJ's interpretation of these provisions in the context of the direct effect of the first and second paragraph of Article 81. This case law can be summarised as saying that, in the absence of a procedure giving effect to Article 81 (3), ordinary national courts do not have jurisdiction to apply Article 81 (1) and (2) unless either the competent national authority [pursuant to Article 84 (former Art. 88)] or the Commission [under Article 85 (2) (former Article 89)] have previously declared that the agreement in question is contrary to Article 81 (1) and cannot be exempted pursuant to Article 81 (3). According to the Court's case law, 'national authorities' within the meaning of Article 84 are either the administrative competition authorities of the Member States or specialised courts specifically entrusted with the task of applying national competition law, but in no case are they ordinary national courts.21 Article 81 (3), instead of generally 'becoming' directly applicable for both national courts and competition authorities, could be made applicable for national competition authorities only. In other words, the application of Article 81 by ordinary national courts could be made conditional upon a prior pronunciation of an administrative competition authority or a specialised court (depending on the provisions existing in each Member State) along the lines of the Spanish Supreme Court's CAMPSA case law described above. Such 'limited' direct applicability of Article 81 (3) is envisaged in Article 84 of the Treaty (which does not refer to ordinary national courts) and sits well with the ECJ's case law on the limited direct applicability of Article 81 (2) in the absence of implementing regulations, established in Bosch and later confirmed in the Asjes and Ahmed Saeedjudgments relating to the air transport sector.22 This case law is summarised in the Asjes judgment: 21 See BRTI, supra note 1, overruling the Bilger judgment, Case 43/69, Rec. 1970, p. 127, para. 9. 22 Case 13/61 Bosch [1962] E C R 89; Cases 209 to 213/84 Asjes [1986] E C R 1425; Case 66/86 Ahmed Saeed [1989] ECR 803. For a detailed analysis of this case law see Schroter, H., in Groeben, Thiesing, and Ehlermann eds. (1999), Kommentar zum EU-IEG- Vertrag 211 and II, Artikel 87 paras. 3-5; Artikel 88, paras 3-7; Pernice, I., in Grabitz and Hilf eds. (1997): EGV, Artikel 87, paras 4 - 8 , Artikel 88; Ritter, in Immenga and Mestmacker eds. (1997): EG-Wettbewerbsrecht-Kommentar II, Artikel 87, paras. 1-4 and Artikel 88, p a r a s 1-4; for a critical appraisal see Cienfuegos, M . (1991): 'L'application de la nullite de l'article 85.2 du Traite CEE p a r lesjuridictions nationales (avec un examen particulier du domaine des transports aeriens)', Cahiers de Droit Europeen, pp. 317-353.

3 — Courts and Judges

611

'[I]n the absence of a decision taken under Article 88 [now Art. 84] by the competent national authorities ruling that a given concerted action [.. .] is prohibited by Article 85 (1) and cannot be exempted from that prohibition pursuant to Article 85 (3), or in the absence of a decision by the Commission under Article 89 (2) [now Art. 85 (2)] recording that such a concerted practice constitutes an infringement of Article 85 (1), a national court f. . .] does not itself have jurisdiction to hold that the concerted action in question is compatible with Article 85 (1 ) . ' 2 3

In Bosch the Court made the equivalent finding in respect of Article 81 (2): '[...] up to the time of entry into force of thefirstregulation implementing Article 85 and 86, the nullifying provision [la nullite deplein droit] had operated only in respect of agreements and decisions which the authorities of the Member States, on the basis of Article 88, have expressly held to fall under Article 85 (1), and not to qualify for exemption under Article 85 (3), or in respect of which the Commission has taken the decision envisaged by Article 89 (2)'.24 If, however, either the national authority or the Commission has ruled on a given agreement pursuant to Article 80 or 85 (2), '[...] the national court must draw all the necessary conclusions therefrom and in particular conclude that the concerted action [. . .] in respect of which such a ruling or recording has been made is automatically void under Article 85 (2).'25 In practice this means that whenever an ordinary (i.e. non-specialised) national court is confronted with the application of Article 81 it would have to inhibit itself in favour of the national authority or the Commission for the purposes of obtaining a binding declaration on the applicability of Article 81 as a whole to the agreement or practice in question. National authorities or the Commission (depending on the nature of the case) would have to give a declaratory positive or negative decision on the applicability of Article 81 as a whole' before the respective ordinary court could proceed with the case before it, drawing all the necessary conclusions (as set out in the Asjes judgment) from the Commission's or the national authority's declaratory decision. The Court basically mentions two reasons for this important limitation of the direct effect of Article 81 (1) and (2) that, at least to a certain extent, also apply within the framework of a legal exception system. First, the principle of 23

Para. 68 of the judgment. Cf. also the analogous—albeit non-transitional- situation under Article 65 of the ECSC Treaty, Case C - l 28/92 Banks v British Coal [1994] E C R 1-1209, para. 17. 24 These limitations d o not apply in the context of Article 82 E C as n o exemption may be granted as regards abuses of a dominant position (see Ahmed Saeed, paras. 30-32). Although it is true that the prohibition of Article 82 cannot conflict with a n exemption decision, the potential for conflicting decisions therefore being m u c h m o r e limited than in the context of Article 81, if the enforcement of Article 81 w e r e reformed along the lines proposed here, it should also be discussed t o subject the direct application of Article 82 t o a prior administrative pronouncement (or t o entrust its judicial enforcement to specialised courts). 25 Para. 69 of the Asjes judgment.

612

Working Paper VII

legal certainty, that impedes the risk that undertakings are exposed to the threat of having their agreements declared void under Article 81 (2) if they do not, at the same time, have the possibility of ascertaining whether the conditions of the third paragraph are fulfilled and Article 81 hence applies as a whole to their agreement. To the extent that national courts would have the power to apply Article 81 (3), the new system would not seem to contravene the principle of legal certainty, although under the legal exception system there would be no 'procedure' to give effect to Article 81 (3) as companies no longer have the right to obtain an administrative decision on the applicability of Article 81 (3). Also, it is not clear whether the Court had an administrative rather than a judicial declaration on Article 81 as a whole in mind when referring to the principle of legal certainty. The second reason adduced by the ECJ is that the application of Articles 84 and 85 (former Arts. 88 and 89) is justified in the circumstances described above because it is through the institutions that have been given jurisdiction to grant or refuse exemptions pursuant to Article 81 (3) 'that competition policy develops'.26 It is this function that national courts cannot fulfil and that therefore justifies the administration's prior pronouncement in any case involving application of Article 81 (3) before an ordinary court. Under the regime proposed here, reference to an administrative authority (or specialised court) would be compulsory for ordinary national courts. The competition issue would in principle be referred to the competent national authority unless the Commission sought to keep jurisdiction in those cases already now contemplated in the White Paper as having sufficient Community interest to be dealt with by the Commission. In contrast to the amicus curiae proposals contained in the White Paper, the parties to the litigation would be fully involved in the proceedings before the respective authority. Member States could avoid the duty to refer a case to an administrative authority by instituting specialised national competition courts (which are also recognised as 'national authorities' under Article 84) and would therefore have a choice in the implementation of this system. In our view, the mandatory pronouncement by a specialised authority—be it the national competition authority, a specialised court or the Commission in cases having a Community interest—would guarantee the coherent and uniform application of Article 81, and in particular its third paragraph, throughout the Community.

26

Ahmed Saeed, para. 32.

VIII Judge Anne Spiritus Dassesse1 On the Modernisation of EC Antitrust Policy

EC competition policy is based on certain provisions of the Treaty of Amsterdam that are binding on all authorities entrusted with the application of this policy at national level.2 As far as judges are concerned, the relevant Treaty provisions are not limited to Articles 81 and 82, but also include, for example, Articles 15, 85, 86, etc. EC competition policy encompasses a single geographical entity, the European Economic Space, and a single legal and jurisprudential system. Thus, Treaty rules take precedence over national laws, and the jurisprudence of the European Court of Justice has authority for national courts and individual Member States. These characteristics reinforce the independence and improve the efficiency of the role played by the national judges vis-a-vis (sometimes) arbitrary and protectionist decisions of their own Member States. These elements are therefore conducive to the enforcement of a uniform policy going behind purely national interests. At present, national judges can enforce the prohibition contained in Article 81 (1) irrespective of the existence of a notification to the Commission. In such circumstances, national judges must of course take into account the possibility that the Commission could apply Article 81 (3). The proposed reform does not substantially change the present setting, except in one respect: it expressively gives competence to national judges to apply, if need be, Article 81 (3). This power was previously within the exclusive competence of the European Commission. Because of the method by which they recruit judges and the matters over which they have jurisdiction, the Belgian commercial courts are very well placed to make use of the possibility of directly applying Article 81 (3). They will be helped to do so by the existence of certain Community legal instruments upon which they can rely. The Belgian commercial courts of first instance have 'professional judges', with law degrees and appointed for life (that is, until they reach the retirement 1

Outline of oral presentation made at the Fifth EU Competition Law and Policy Workshop, EUI Florence, 2-3 June 2000. 2 Preliminary remarks: the questions addressed below should be looked at against the background of the following issues: the scope of the application of articles 81 and 82 of the Amsterdam Treaty: what is their position in the pegging order of legal norms? how are they enforced by national courts? would it be appropriate for national judges to be specialised? How does one best ensure coherence and uniformity in the enforcement of competition rules and meet the wish for legal certainty of commercial operators?

614

Working Paper VIII

age, which is generally 67 years), as well as 'lay judges', appointed for renewable five-year terms—although normally they would keep the appointment for life—and recruited from all sectors of the economy. Many of the 'lay judges' do not have a law degree and did not practice law before their appointments. The President of the Commercial Court is a professional judge. Each chamber of the Commercial Court consists of a presiding professional judge and two lay judges. Decisions in all chambers are taken by simple majority vote. The President (or, in his or her absence, a substituting professional judge) has exclusive jurisdiction over certain kinds of proceedings, such as summary proceedings and unfair competition proceedings. As a result of their composition, Belgian commercial courts are sensitive to economic realities as well as the need for speedy decision-making. The composition of the commercial courts is designed to cover both legal specialisation in economic and competition law (for which the professional judges are primarily responsible) and contact with economic life generally (for which the lay judges are primarily responsible). Belgian, commercial courts of first instance ('tribunaux de commerce' or 'rechtbanken van koophandel') have jurisdiction over all disputes between or against companies or traders regarding their business activities.3 These courts are separate from the civil and criminal courts of first instance. Appeal against the judgments of the commercial courts is to be directed to the Belgian Courts of Appeals. The jurisdiction of the commercial courts is not limited to the application of Articles 8land 82. In fact, these courts have a much broader dominion. This can be a significant advantage in contributing to the enforcement of an effective competition policy in the European Economic Space. For example, in a specific case, the claimant may fail to establish the existence of an illegal cartel. As a result, Article 81 cannot be applied. However, in such a case a commercial court may come to the conclusion that the defendant benefits from state aid that is illegal because it has not been 'notified'. In accordance with Article 88 (3),4 the court then has jurisdiction to declare the illegality of this state aid and (if requested by the claimant), to order its reimbursement by the defendant.

3 Business in general (international and national trading—banking—shipping—intellectual property—competition—company law—bankruptcy—insolvency etc. . . .)— Spiritus-Dassesse A. (1996): 'The role of national courts with respect to European Community competition law', in CD. Ehlermann and L.L. Laudati, eds.: Robert Schuman Centre Annual on European Competition Law 1996, London, Kluwer Law International, 355-000. 4 Article 88 §3 (re. Aids granted by States) of the Treaty : 'The Commission shall be informed, in sufficient time to enable it to submit its comments, of any plans to grant or to alter aid. If it considers that any such plan is not compatible with the common market having regard to Article 87, it shall without delay initiate the procedure provided for in paragraph 2. The Member State concerned shall not put its proposed measures into effect until this procedure has resulted in a final decision.'

3 - Courts and Judges

615

One has to bear in mind that the European prohibition on state aid is of public interest for the Community as a whole. Thus, in accordance with the case law of the European Court of Justice,5 the Belgian judge must scrutinise whether the case brought before him includes illegal state aid. The judge must institute this examination on his own initiative, irrespective of whether any of the parties to the proceedings have raised this question of possible illegality. If the Belgian judge, after having heard the arguments of the parties, then comes to the conclusion that prohibited state aid is involved, he has to draw all implications of that illegality. Thus, for example, a national judge may not decline to declare bankrupt a company on the ground that its main creditors, such as the Social Security or the VAT Authorities, have agreed to reschedule their claims more or less indefinitely. Indeed, by taking these elements into account the judge would allow the company to benefit from prohibited state aid.6 Similarly, a national judge must exclude a contestant from a tendering process if he concludes that the contender, which had submitted a bid in response to a tender invitation from public authorities, has benefited from prohibited state aid that makes its bid more competitive.7 Individual companies, and their technical and legal advisers, endeavour to ensure that the agreements into which they enter are compatible with Article 81. They will amend them, if necessary, in order to meet this objective. However, experience teaches that the unearthing of significant anticompetitive agreements or behaviour having substantial negative effects on competition, and of which the Commission has not been notified, or which do not respect the conditions pursuant to which an exemption was granted, usually comes about by accident. Experience also teaches that a national judge can play a significant role in this respect. The proposed reform will require companies to be even more attentive to the need to respect Article 81 in connection with the agreements they enter into. This can only benefit a truly competitive environment. As past experience has shown, the notification procedure does not solve all problems in any event. Indeed it is far too rigid, as evidenced by the problems linked to the distinction between 'pre-notification' and 'past-notification', and having regard to the absence of a method by which the notification procedure can be invoked retroactively. National judges will be able to be more flexible once they can apply both Article 81 (1) and Article 81 (3): they will be able to weigh up all relevant 5 Case C-126/97 Eco Swiss China Time Ltd. v Benetton International NV[\999] ECR 1-3079; see Belgian proceedings rules (Article 774 of the Belgian Judicial Code which provides that the judge must raise, of his own initiative, legal issues of Belgian public policy. 6 D M Transport: 1998 Journal des Tribunaux de Droit Europeen 90 (Brussels Commercial Court)—Case C-256/97 D M Transport SA [1999] ECR 1-3926. 7 Manoir v Breda, 1995 Journal des Tribunaux de Droit Europeen 72 (Brussels Commercial Court).

616

Working Paper VIII

elements of the situation put before them and, in this respect, to take into account the requirements of an ever changing and dynamic economic environment. This should benefit both commercial enterprises and consumers. Companies have claimed that the proposed reform will bring about increased legal uncertainty. This fear appears to be unfounded and, under the present rules, legal uncertainty already exists. It is only in a very small minority of cases that the Commission reacts to a notification in a formal manner, thus invoking all the benefits of the legal safeguards that attach thereto. In most cases, its reaction takes the form of a comfort letter and has no binding effect on the national judges. One should also note that there are already a number of Treaty provisions, the application of which is shared between the Commission and the national judges (and which are not therefore the exclusive preserve of D G COMP). And there are other instances where commercial companies' desire for total legal certainty with regard to Article 81 can be defeated (the field of illegal state aid comes to mind in this respect). A few words about 'forum shopping': in my opinion, forum shopping cannot be avoided: the ingeniousness of lawyers (and their determination in this respect) has proved this point on many occasions and in various situations. Competition law is no exception. When a national judge has to declare an agreement null and void under the Article 81 (2), the judge is bound by Community law to assess the effects of such invalidation in accordance with his national law. Thus an element of forum shopping is inevitably present on a geographical basis as a result of the lack of harmonised rules in this field (for example, with regard to the amount of damages). This is a situation that already exists under the present rules, which give the Commission a monopoly over the application of Article 81 (3). Transferring the jurisdictional power to apply Article 81 (3) to national judges will in no way change this situation.

IX Judge Sarah S. Vance Judicial Application of Article 81 (3): Are the Fears Justified?

The White Paper on Modernisation of the Rules Implementing Articles 81 and 82 of EC Treaty ('White Paper') contains proposals that would make fundamental changes in the roles of the European Commission, national courts, and national enforcement authorities in the Community's system of competition enforcement. It is not surprising, then, that the proposed changes, including the new role envisioned for national courts in applying Article 81 (3) of the EC Treaty, have provoked anxiety in the legal, political and business communities. This article gives the perspectives of a U.S. judge on the new role proposed for national courts in the White Paper's decentralised enforcement regime. This article asserts that fears about the ability of judges to apply Article 81 in its entirety, while not unfounded, are exaggerated. Competition law principles are not beyond the ken of judges, and mechanisms exist or have been proposed to minimise the risk of inconsistent applications of Article 81 (3). The article also asserts that opportunities for judicial education would enhance the accuracy of national courts in applying Article 81 (3). Greater legal harmonisation would also facilitate substantive consistency.

I. The changes The White Paper proposes changes in three areas to make enforcement of Community competition law more effective. First, it would end the notification and authorisation system under which undertakings must apply to the Commission for an affirmative decision that agreements falling under Article 81 (1) are not illegal.1 Second, the White Paper would decentralise the application of the competition rules.2 Last, the Commission would intensify ex post control over anti-competitive activity.3 Article 81 (1) of the EC Treaty prohibits agreements that 'may affect trade between Member States' and that have the 'object or effect' of restricting competition. Article 81(2) makes such agreements automatically void, but paragraph 81 (3) provides that paragraph 81 (1) can be declared inapplicable when the agreement has pro-competitive benefits, is not unnecessarily restrictive, and 1 European Commission: White Paper on Modernization of the Rules Implementing Articles 81 and82 of EC Treaty, OJ C 132 of 28.4.1999, paras. 76-81. 2 Id. paras. 82-108. 3 Id. paras. 108-128.

618

•

Working Paper IX

does not impose the risk of eliminating substantial competition. Regulation 17/62 established the current system in which the Commission alone can apply Article 81 (3) to grant an exemption and in which undertakings are required to give prior notification of the agreements for which an exemption is requested.4 The White Paper would eliminate the requirement for an administrative exemption decision. Instead, it would make Article 81 (3) directly applicable so that any national court or national competition authority could apply it.5 Thus, agreements would be legally enforceable until a competent authority found that they did not satisfy the criteria of Article 81 (3). These proposed changes are designed to free the Commission from the administrative burdens imposed by the ex ante enforcement system, so that it may focus on the most severe violations of competition rules, which companies rarely bring to the Commission's attention through notification. The changes are also designed to make the enforcement of Article 81 more effective by decentralising responsibility for its enforcement. Few would question the need to lighten the administrative burden on the Commission imposed by the notification and exemption system.6 Not everyone agrees, however, on the means to modernise the EU competition enforcement system.7 Those who fear the enhanced role proposed for national courts in applying Article 81 (3) express concerns that national courts lack experience and training to deal with the complex legal issues and sophisticated economic principles involved in applying Article 81 (3).8 They argue that this inexperience, together with the multiplication in the number of decision-makers, will bring about inconsistent interpretations of EU competition law, to the detriment of coherent competition policy and business certainty. Another concern is that in national courts parochial concerns may override competition principles, possibly resulting in a re-nationalisation of competition policy. Still another concern is that the divergence in procedures and remedies from one Member State to the next will undermine coherence in interpretation and promote forum shopping.9 These 4

Council Regulation No. 17 of February 6,1962, Art. 9(1), O.J. 35, 10.5.1962, at 118. See White Paper para. 82. 6 See id. paras. 24—40 (describing administrative burdens of centralised notification and authorisation system and efforts to minimise them). 7 White Paper on Reform of Regulation 17/62, Summary of the Observations (February 29, 2000), passim. The American Bar Association's Section of Antitrust Law, for example, would prefer the adoption of a Rule of Reason under Article 81 (1). See Comments of the Section of Antitrust Law of the American Bar Association on the European Commission's ' White Paper on Modernisation of the Rules Implementing Articles 85 And 86 of the EC Treaty' (1999) at 4-5 (hereinafter ABA Comments). Recognising that this approach may not be feasible at this juncture, the Section supports making Article 81 (3) directly applicable but has reservations about decentralised enforcement. 8 See Ehlermann C.-D. (2000): 'The Modernisation of EC Antitrust Policy—A Legal and Cultural Revolution', EUI-RSC Working Paper No. 17/2000; ABA Comments, supra note 7, at 13. 9 See ABA Comments, supra note 7, at 15-16; Kon, S. (1999) 'The Commission's White Paper on Modernisation: The Need for Procedural Harmonisation', Fordham 5

3 - Courts and Judges

'

619

concerns are far from trivial, but they do not justify foregoing the flexibility offered by the new, less formalistic system.

II. Are judges up to the task? U.S. experience suggests that judges are capable of grasping the legal and economic issues presented in competition cases. In the U.S., generalist trial judges hear competition cases brought by private litigants seeking treble damages or by enforcement agencies seeking injunctive relief or criminal penalties. Generalist appellate judges hear appeals from trial court decisions or of final administrative actions. Competition cases comprise a very small portion of the dockets in U.S. federal courts. Nevertheless, U.S. judges define markets, balance economic factors under the Rule of Reason and assess anti-competitive effects. While competition cases are complex, they have not proved to be beyond the abilities of judges to handle with a reasonable degree of competence. The decisions of Judge Thomas P. Jackson in United States v Microsoft10 illustrate how a judge with no specialised economic training can master sophisticated economic principles and apply them to complicated facts. EU competition rules are not more complicated than their U.S. counterparts, and national court judges should be able to apply them competently. Judges on national courts are not strangers to the application of EU law. As Temple Lang has observed, 'national courts probably apply Community law more often than the two Community Courts do.'11 Temple Lang also notes that many of the most important issues of Community law 'have been raised first by national judges.'12 In the competition law arena, national courts already apply Article 82, which is at least as difficult to apply as Article 81 (3).13 National court judges also deal with complex legal issues in other areas like intellectual property, which involves complex economic evaluations, and human rights, which involves balancing conflicting interests and values.14 National judges will not be without guidance in applying Community competition principles. Over the past thirty years, the Commission and the Court Corporate Law Institute, 26th Annual Conference on International Antitrust Law and Policy (forthcoming). 10 84 F.Supp.2d 9 (D. D.C. 1999) (findings of fact); 87 F.Supp.2d 30 (D. D.C. 2000) (conclusions of law). 11 Temple Lang, J. (1996): 'The Duties of National Courts under Community Constitutional Law', paper given at Cambridge on Institute of European Law, Feb. 23 1996, at 1 (mimeo). 12 Id. at 15. 13 Temple Lang, J. (1999) 'Decentralised Enforcement of Community Competition Law', paper given at Dublin on Conference on Competition Enforcement, Nov. 11, 1999, at 6-7 para. 7 (mimeo). 14 Ehlermann, supra note 8, para. 146.

620

Working Paper IX

of Justice have gone a long way to flesh out Community competition rules. Through court decisions and the Commission's decision-making practices, block exemption regulations, guidelines and notices, national courts have the benefit of more recent and authoritative guidance on competition law than U.S. courts have received from Congress or the Supreme Court in recent years. The new block exemptions and guidelines for vertical restraints will add further certainty to the law in this area, as will proposed new rules and guidelines on horizontal agreements between competitors.15 The White Paper envisions that the Commission will continue to issue group exemption decisions, prohibition decisions, guidelines, as well as declaratory decisions to serve as precedents in new areas.16 The quality of judicial decision-making in competition cases can be enhanced by educating judges in competition law and economics. In the U.S., the Federal Judicial Center offers educational programs taught by legal experts in variousfields,which judges can attend in person or view on tape. Bar associations like the ABA Section on Antitrust Law provide educational programs and materials on competition issues from the most elementary to the most complex. These types of focused educational experiences give judges the 'big picture' and enable them to approach a complicated case at a much higher level of competence. Judges on national courts would likewise benefit from educational programmes designed to teach them the fundamentals of EU competition principles and economics. One writer has suggested the possibility of appointing a Community Law Adviser or National Advocate General in each Member State to assist in individual cases.17 At the request of national courts, this official would advise national courts on competition law issues, economic issues, or on the application of law to fact. Other suggestions for improving the performance of national courts include allowing national competition authorities to make submissions to courts in competition cases and appointing economic advisers to sit with judges.18 In the U.S., federal judges have-the authority to appoint independent experts and to tax the costs on the parties.19 The parties have the right to discover the expert's opinions before trial and to cross-examine him. Judges do not invoke this rule often because the parties usually hire their own experts, and judges typically find the adversary system adequate to expose the biases 15 See European Commission: Regulation (EC) No. 279011999 of 22 December 1999 on the Application of Article 81 (3) of the Treaty to Categories of Vertical Agreements and Concerted Practices, O.J L 336 1999 pp. 21-25; also Draft Commission Regulation on the application of Article 81 (3) of the EC Treaty to categories of research and development agreements; Draft Commission Regulation on the application of Article 81 (3) of the EC Treaty to categories of specialisation agreements; and Draft Guidelines on the Applicability of Article 81 to horizontal cooperation. OJ C 118 of 27.4.2000. 16 White Paper paras. 14, 78, 85-89. 17 Temple Lang, supra note 13, at 6 para. 7. 18 Id. 19 Fed. R. Evid. 706.

3 - Courts and Judges

621

held by partisan experts. The rule has, however, been used in a number of competition and complex products liability cases. Judges on national courts would undoubtedly benefit from the input from economic experts, in one form or another, in cases involving the application of Article 81 (3).

III. The inconsistency problem The sheer number of decision-makers under the reform proposal poses a more realistic threat to the coherent application of Article 81 (3) than does the inexperience of judges. Increasing the number of decision-makers inevitably increases the risks of inconsistent decisions. In the U.S., where nearly 750 federal trial judges apply the Sherman Act, inconsistent decisions do occur. Federal appellate review provides harmonisation, at least within the jurisdictions of the courts of appeals. When U.S. appellate courts disagree, only Congress or the Supreme Court can provide a national rule. While higher courts do not resolve every conflict, and the system is far from perfect, the U.S. system does not produce intolerable levels of uncertainty or inconsistency in competition enforcement. As long as avenues exist to resolve significant issues of national importance, a certain amount of inconsistency, particularly in evolving areas, can be viewed as healthy experimentation. The White Paper acknowledges the risk of inconsistent decisions, and it proposes several information sharing and co-operation mechanisms to minimise these risks.20 A significant mechanism for assuring coherence in the interpretation of EC competition law already exists in Article 234 of the EC Treaty.21 Under this article, national courts are permitted, and courts of final appeal are required, to apply to the European Court of Justice for preliminary rulings on questions of interpretation of Community law.22 Although this mechanism is 20

White Paper para. 107. There is apparently no avenue for judicial review of a final decision of a national court or a national competition authority on a Community law issue by the C o m m u n i t y courts in Luxembourg. A t least one writer has suggested judicial review of national competition authority decisions by the Court of First Instance and, on appeal, by the Court of Justice. See Cooke, J.D. (1999): 'Changing Responsibilities and Relationships for Community and National Courts: the Implications of the White Paper', paper given at a conference on The Modernisation of the Rules Implementing Articles 81 and 82 of E C Treaty (mimeo). While an appeal to a Community court of last resort would appear to be an effective way to enforce Community-wide rules, this suggestion runs afoul of the general decision-making framework established by the E C Treaty, which apparently accepts the risk of inconsistent decisions. See Ehlermann (2000) at para. 114. For this reason, such appeals proposals are probably too radical at this time. 22 This procedure is akin to the U.S. procedure for certifying issues for interlocutory appeal. A federal court may submit an issue to an appellate court for resolution by certifying that it involves a controlling question of law about which there exists a substantial ground for difference of opinion and that a n immediate appeal may materially 21

622

Working Paper IX

effective, national courts apparently vary in the frequency to which they resort to this procedure.23 The preliminary ruling process is also slow, taking an average of about 21 months in 1998.24 To the extent that decentralisation produces more requests for preliminary rulings, the process may become even slower. The Courts of the EC have made proposals to improve the preliminary ruling process.25 One proposal is to give the Court of First Instance subject matter jurisdiction over preliminary references to assist the Court of Justice.26 Given the importance of the preliminary reference process to the coherence of the Community legal scheme, reforms aimed at improving its responsiveness are worth exploring. The White Paper does not rely on the preliminary ruling process alone to deal with the risks of inconsistent decisions by national courts. It proposes that national courts be required by regulation to inform the Commission of cases in which Article 81 or 82 are invoiced and that the Commission be allowed to intervene as amicus curiae in these proceedings.27 The White Paper also proposes to formalise by regulation the rules set out in the Notice on Co-operation between the Commission and National Courts.28 The Notice provides that national courts may ask the Commission for information of a procedural, economic or legal nature during the course of proceedings before them.29 Specifically, the Notice states that national courts may consult the Commission on its customary practice on the Community law question at issue and receive a non-binding opinion.30 Further, national courts can obtain information like statistics, market studies and analyses from the Commission.31 The ability of the Commission to intervene as amicus curiae is a powerful tool for assuring consistency. While the Commission will probably not intervene often,32 this mechanism provides the means for the Commission to act when there is a significant threat to Community interests or legal consistency. As to the ability of national courts to obtain information from the Commission advance the termination of t h e litigation. The appeals court has the discretion to refuse to decide the certified issue. See 28 U . S.C. § 1292(b). Courts of appeal may certify issues to the Supreme C o u r t when they desire instructions on the law. Id. § 1254. 23 See Kon, supra note 9, at 1 2 - 1 3 . 24 See C o u r t of Justice a n d C o u r t of First Instance, 'The Future of the Judicial System of the European Union (Proposals and Reflections)', submitted to Council of Ministers of Justice and H o m e Affairs on May 28, 1999, at 29. 25 Wat 21-27. 26 Id. at 25-26. 27 White Paper p a r a . 107. 28 Id. 29 E u r o p e a n C o m m i s s i o n : Notice on Cooperation between National Courts and the Commission in Applying Articles 85 and 86 of the EEC Treaty, OJ C 39 of 13.2.1993, at p. 6. 30 Id. paras. 38-39. 31 Id. para. 40. 32 Indeed, K o n asserts that the C o m m i s s i o n already has this authority a n d has never used it. See Kon, supra n o t e 9, at 16.

3 - Courts and Judges

623

in pending cases, this can be effective only to the extent that courts avail themselves of this opportunity, and the Commission has the resources to provide the information. Early reports suggest that the impact of the Notice on Co-operation has not been great.33 National courts are not accustomed to asking any authority for information and prefer the preliminary ruling process to get authoritative rulings on legal issues.34 Greater levels of co-operation between national courts and the Commission may result, however, if national courts see increasing numbers of Article 81 (3) cases. As a practical matter, concerns over inconsistent national court decisions under Article 81 may be exaggerated to the extent that they are predicated on an expectation of large-scale use of private litigation. Europe does not have the private litigation tradition or the incentives to sue found in the U.S., where there are treble damages, class actions and contingency fees. Even though damages are available from national courts, national courts have not issued many competition decisions.35 Indeed, some writers argue that the historic lack of enthusiasm for domestic enforcement before national courts may not be purely the product of the Commission's exclusive jurisdiction over the application of Article 81 (3).36 To the extent that private litigation in national courts increases with decentralisation, change is likely to be gradual. This suggests that while private enforcement of Article 81 in national courts may not be as rigorous as desired at first, the time lag will allow breathing room to see what problems develop under the new system and to make adjustments. In any event, if the U.S. experience is any guide, a certain amount of inconsistency in court decisions does not render an enforcement system ineffective.

IV. Harmonisation issues Several commentators see a lack of substantive and procedural harmonisation as an obstacle to consistent application of EC competition law and as an inducement to forum shopping.37 They note that differences in the availability of discovery and expert testimony, as well as differences in remedies, burden of proof, rules of causation and rules on standing can undermine the coherent application of Article 81 (3).38

33 See Temple Lang, J. (1998): 'General Report on the Application of Community Competition Law on Enterprises by National Courts and National Authorities', paper given at F I D E Congress, Stockholm on June 3, 1998, at 38. 34 Id. 35 See Kon, supra note 9, at 5-7. 36 Id. at 7; see also Ehlermann, supra note 8, para. 151. 37 Kon, supra note 9, at 6 & passim; ABA Comments, supra note 7, at 15-16; Temple Lang, supra note 33, at 41-46. 38 See note 37.

624

Working Paper IX

It is difficult for a foreign observer to evaluate the magnitude of this problem. This is especially true since national courts apply Community law in other areas, and many of the same procedural differences appear to exist. In the U.S., federal courts apply uniform rules of civil procedure and evidence, and impose the same set of remedies for federal antitrust violations.39 Nevertheless, U.S. businesses can also be sued in state courts under state antitrust laws, which are generally similar to, but do not always converge with, federal antitrust rules. On procedural issues, the states differ in varying degrees from the federal system and from each other. While this makes for a vigorous enforcement environment, differences in state and federal approaches to antitrust enforcement are a source of frustration and added expense for those subject to regulation. Some of the issues that divide national courts are more central to the concerns of Community competition policy than others. For example, the questions of whom can sue for infractions of Community competition rules and for what types of relief are more related to substantive competition rules than are rules on discovery or the use of experts. The ECJ has stated that national courts must give effective protection to Community law rights.40 Temple Lang argues that this principle will require the Court 'to define the extent of these rights, and what effective protection necessitates,' which means answering questions concerning 'who may sue . . . , in what circumstances, and for what compensation.'41 Harmonisation on the issue of whom can sue and for what relief, whether by decision of the ECJ or by regulation, is desirable sooner rather than later. Greater harmonisation on other procedural issues is probably too ambitious an undertaking at this juncture, especially since it would affect the procedures national courts use in cases outside the realm of competition law. If European national courts are anything like their U.S. counterparts, institutional resistance to an overhaul of court procedures is likely to be substantial. Over time, however, greater procedural harmonisation will probably occur as a result of pressure from the bar and litigants. In sum, while procedural and substantive law harmonisation is desirable, its absence should not be a reason to reject the reforms proposed by the White Paper. Greater harmonisation over time is likely to be a consequence of decentralisation.

V. Conclusion The White Paper's proposals are designed to produce greater involvement of national courts and national authorities in EC Competition enforcement. If 39

See, e.g., Fed. R. Civ. Proc. (West 1999); Fed. R. Evid. (West 1999); 15 U.S.C. §§ 1, 2, 4, 15a, 26 (1994) (establishing criminal penalties and authorising treble damages and injunctive relief). 40 See, e.g., Case C-271/91 Marshall // [1993] E C R 1-4367; Case C-213/89 Factortame [1992] E C R 1-2433; Temple Lang, supra note 33, at 38-39. 41 Temple Lang, supra note 33, at 38-39.

3 - Courts and Judges

625

successful, this will do more than free the Commission to focus on the big cases and to set its own enforcement agenda. Investing national institutions with greater enforcement authority under the supervision of the Commission and the European Court of Justice will spread a common culture of competition and increase commitment to competition principles in Member States. These developments appear to be essential to an effective single market. For these reasons, this article argues that a greater role for national courts in competition enforcement is generally a healthy development, the risks of which are manageable.

X Diane P. Wood* Techniques of Judicial Federalism

Abstract This paper explores three questions. First, it looks at the question whether the national courts within Europe can be expected to have the technical ability to handle competition cases, should the White Paper's system be adopted in a relatively unchanged fashion. It concludes that they can, using analogies to similar problems in U.S. law and discussing the techniques U.S. courts use to handle cases that are complex and that arise out of difficult areas of the law. Second, it takes an extensive comparative look at the mechanisms U.S. courts use in allocating responsibility among courts within a federal system. Principal attention is paid to systems that ensure that state courts adhere faithfully to federal law, but some examples are drawn from cases where the question is whether a federal court has properly understood state law, and others are drawn from co-ordination mechanisms that are available within the federal court system. Finally, the paper looks briefly at certain problems in ensuring substantive consistency. It concludes that this is an overstated concern, and that there is every reason to expect that the national judiciaries within Europe will rise to the new challenge presented in the White Paper's proposals.

I. Background With the advent of the White Paper on Modernisation of the Rules Implementing Articles 85 and 86 of the EC Treaty, the European Union has embarked on an ambitious programme of decentralising the enforcement and implementation of competition law. As Part II.A.3 of the White Paper specifically recognises, an important consequence of this decentralisation will be the enhanced role that national courts will play in the application of the EC Treaty's competition rules. This will occur against the backdrop of a mosaic that looks vaguely familiar to an observer from the other side of the Atlantic: a central authority (the Commission) exists with primary authority for articulating European-level competition rules and policies and for taking enforcement actions; courts at the European level (the Court of First Instance and the European Court of Justice) hear cases brought by the parties to European * The views expressed in this paper are purely personal.

628

Working Paper X

proceedings, by the Commission, or by Member States; national authorities exist with responsibilities both to apply European law faithfully and to implement and apply their own set of national competition rules; and national courts are open to hear either European or national law cases. There are obviously many differences between the European and the American systems for competition law and (of greatest concern here) the role that state and national judiciaries have played in that area. Those differences in most cases are important reflections of divergent histories, structures, and goals, and no one would argue that a direct transplant from one system to the other would make any sense. On the other hand, it is also true that there are useful insights to be gained from a study of one another's experiences. With respect to the role of courts in competition law enforcement, and more generally with the co-ordination of state and national judiciaries, the United States may have something useful to offer Europe as Europe deepens the interactions between the courts and judges of its Member States and those of the EU. On the American side, this is sometimes referred to as the problem of 'judicial federalism,' and I will occasionally use that term in this paper. For Europe, a better term might be something like 'bi-level judicial responsibility,' though at present I shall leave the coining of such phrases to others. The concerns expressed in the White Paper over the proposed greater role for national courts and the commentary the White Paper has inspired suggest three areas deserving of special attention: (1) the competence of national courts to take on the 'directly applicable exception system' described in the White Paper; (2) the need for, and design of, mechanisms that will make possible effective coordination among national courts, and between national courts and the European courts; and (3) the need for, and design of, mechanisms to ensure that competition policy remains coherent and consistent across all of Europe after decentralisation is achieved. Drawing heavily on American experience, I look at each of these in turn.

II. Ability of national courts to decide competition cases No one can doubt that competition cases, or antitrust cases in the United States, entail complex questions of fact, economic policy, and law. Rare is the U.S. federal district court judge who, coming to the bench from a background of a federal prosecutor, or a general litigation practice, or a state court, does not quail when the first antitrust case shows up on the docket. The same could be said of a judge on a state court of general jurisdiction, who first encounters a case that might have been brought under the federal antitrust laws, but in which the plaintiff instead invoked rights under state contract laws (e.g., for a terminated distributorship), or state laws of unfair competition, or state antitrust laws. With very few exceptions, the United States has not adopted the system of specialised courts that prevails across much of Europe, and so the problems

3 - Courts and Judges

629

these judges face are as severe as any the national judges in the Member States of the EU may encounter in the post-decentralisation world. The first problem facing our hypothetical judge is to distinguish between proper questions of law and questions of policy. Questions of law will first include pertinent procedural issues, including matters like whether the court has proper jurisdiction over the parties to the case, whether certain information can be collected in aid of the proceeding, what privileges might be available to the defendants, the order in which various legal points will be explored, and the range of possible remedies. There is nothing unique in competition cases for that kind of problem, and there is thus no reason to believe that the national judges in Europe will have any more difficulty with the protection of the rights of the defense or with the administration of the cases than they do with any other matter properly before them. Indeed, the national court judges are likely to bring more expertise on those matters than any administrative authority possibly could, precisely because they deal with the processing of cases every day. The other questions of law relate to the substance of the matter. Is it a case dealing with horizontal agreements? Vertical agreements? The conduct of a large and possibly dominant firm? A U.S. judge would look principally to two sources of law in deciding such a case: the text of the antitrust statutes themselves (vague though they are) and the accumulated decisional law from the U.S. Supreme Court and the relevant courts of appeal. To a far lesser degree, the U.S. judge might look to guidelines and statements from the U.S. enforcement authorities (i.e. the Antitrust Division of the U.S. Department of Justice and the Federal Trade Commission—almost never the state authorities), but those statements would be of persuasive value only in the antitrust area. Once the trial judge has identified the sources of law, the remainder of the judicial task is to compile the proper factual record on which to base a decision. In the United States, there is nothing that distinguishes antitrust cases from many other potentially complex matters, ranging from securities, to environmental, pension, labour relations, intellectual property, or complex product liability matters. There are techniques for acquiring the assistance of expert witnesses and for managing the litigation that are helpful, but a discussion of those techniques is beyond the scope of the present paper. What about policy? Given the broad commands of the antitrust laws (e.g., do not enter into anti-competitive agreements; do not monopolise or attempt to monopolise), there is certainly room for policy judgments. The prosecuting authorities, which include not only the Antitrust Division and the FTC, but also the state attorneys general, exercise policy discretion every day when they decide which cases are worth pursuing. The various guidelines and speeches they make available to the public are an important source of current thinking about the correct way to interpret the antitrust laws. But, because of the importance of the private right of action and the direct access U.S. plaintiffs have to the courts, the enforcement authorities do not have the last word on antitrust policy. It is for the courts to glean exactly how far the law will go, in the light of traditional principles of statutory interpretation and stare decisis. The fiction—

630

Working Paper X

and it may be only that in thefieldof antitrust—is that Congress has provided the last word on policy, and the courts are simply applying the law as it has been given to them (including the delegation by Congress to the courts to exercise quasi-common law authority to elaborate this particular field of law). That aspect of American antitrust law is least likely to be useful to those in other countries and regions currently developing their own competition laws. It is in part an historical accident in the United States, and in part it reflects much broader truths about the American legal system. Once again, a full exploration of these characteristics would take us far beyond the present topic and would in any event be of doubtful utility to a European audience. More helpful analogies for European competition law are present, however, in areas of U.S. economic regulation that use a more administrative model, such as securities regulation, environmental regulation, or labour regulation. In these fields, Congress has also enacted legislation that is rather general, but it has delegated to the competent administrative agency the responsibility to spell out particular rules of conduct that are consistent with the broad outline Congress has provided. In areas like this, the statements of the competent administrative authorities are entitled to deference by the federal courts. In a recent case, for example, the Seventh Circuit had to decide whether to uphold an order of the Commodities Futures Trading Commission that penalised certain traders for engaging in non-competitive, pre-arranged (and hence illegal) trading practices on the exchange. See Elliott v Commodities Futures Trading Comm'n, 202 F.3d 926 (7th Cir. 2000). By a 2-1 vote, the panel ruled in favour of the CFTC, in an opinion that made clear the majority's belief that it was required to give substantial deference to the CFTC's expertise in regulating commodities markets. The doctrine under which courts must defer to legitimate policy choices made by competent administrative agencies is known as the Chevron rule, after the Supreme Court decision that explained it in greatest detail, Chevron U.S.A. Inc. v Natural Resources Defense Council, Inc., 461 U.S. 837 (1984). It is quite possible that something like Chevron deference would be appropriate for national courts reviewing EC competition cases, where the Commission has issued guidelines or other official statements designed to provide definitive rules for all the Community. Before leaving this subject, it is also important to note what role remains for the courts in a Chevron situation. It can be a significant one, as the recent decision of the U.S. Supreme Court in Food & Drug Admin, v Brown & Williamson Tobacco Corp. 120 S.Ct. 1291 (2000), illustrates. The question before the Court in that case was whether the federal Food and Drug Administration had acted within the scope of its authority when it attempted to regulate the nicotine in tobacco products as a 'drug.' The Court concluded that the FDA had exceeded its powers and thus it struck down certain regulations that the FDA had promulgated. The opinion discusses the way in which a Chevron analysis must proceed: first, the reviewing court must ask whether Congress directly spoke to the precise question at issue; if so, then the only job for the court is to give effect to the legislative command; if not, 'a reviewing court must respect the agency's

3 - Courts and Judges

631

construction of the statute so long as it is permissible.' (See 120 S.Ct. at 1300.) The latter rule, which requires deference, is justified both because the responsibilities for making policy choices are not proper judicial functions and because the agency has greater familiarity with 'the ever-changing facts and circumstances surrounding the subjects regulated.' (Id.) The same functions could be performed by either EU or national courts in considering regulations, directives, or less formal statements in the competition area. It would be open for the court to decide that the Commission had exceeded its powers under the Treaty (as in fact the ECJ did in the case that challenged the original version of the U.S.-EC Antitrust Cooperation Agreement), but as long as the Commission had chosen to implement the rules in a way not forbidden by more fundamental legal instruments, judicial deference would be required. In this way, as well as through the ordinary process of relying on earlier decisions of the Commission, the EU courts, and (over time) the courts of other Member States, it seems that the legal content of the competition rules would be just as accessible to the national courts as the content of any other rules that fall within the particular court's jurisdiction. Especially in the early stages, the advocates appearing before the courts in competition cases will bear a heavy responsibility to ensure that the relevant legal principles are explained as clearly as possible, to call to the court's attention the most helpful prior decisions and scholarly writings, and to keep the record as clear as it can be. That leaves the question whether the judges of the national courts will have the capacity to deal with the theoretical and factual complexity of competition cases. I share Professor Ehlermann's view that the answer is yes. This is not to say that such cases will always be easy, but as he points out, neither are cases dealing with industrial property rights, complex international joint ventures, or environmental catastrophes. It will be up to each Member State to decide whether its courts are currently equipped with all the tools they need to secure expert advice, to organise large quantities of information, and to command the participation of those who know the most about the case. To the extent that Member States have programs for the continuing education of their judiciaries, competition law will be a topic that they will probably add to the curriculum. (Such programs are common in the United States, and they are most helpful to both federal and state judges.) In conclusion, therefore, it appears to me that the task the White Paper contemplates assigning to the national judiciaries is one that they will be able to undertake. Indeed, it may seem less unfamiliar to the national judges than many who are steeped in competition law as a speciality may fear.

III. Mechanisms for effective co-ordination among courts The greatest risk that the White Paper (and its critics) have identified for the proposed new system is the loss of the consistency and coherence that

632

Working Paper X

competition law in Europe has enjoyed thus far. The Commission's monopoly over the granting of exemptions under the article now numbered 81 (3) has had the practical effect of ensuring that all of Article 81 has developed under its own firm hand, as corrected from time to time by the Court of First Instance or the European Court of Justice. While the Brussels monopoly over Article 82 cases has not been as absolute, the only fair assessment of that law as well is that it has evolved almost exclusively through the actions of European institutions as opposed to those in the Member States. The question on everyone's mind is how that desirable consistency can be maintained once responsibility for enforcement and application of the law becomes decentralised. The same problem exists both for antitrust law and, even more acutely, for other areas of law, in the United States. Antitrust law is a bit peculiar, because it is one of the few areas in which Congress gave exclusive jurisdiction to the federal courts for all cases presenting federal antitrust claims. Far more typical are the federal civil rights statutes, such as 42 U.S.C. § 1983, which give a plaintiff the option to begin her action in either state court or federal court. The general rule in U.S. law is that the state courts have concurrent jurisdiction with the federal courts for claims arising under federal law. Federal courts do not have concurrent jurisdiction with the state courts for claims arising under state law, unless all plaintiffs are from different states than all defendants (and foreign countries are treated as a single state) and the matter in controversy exceeds $75,000. In fact, major commercial disputes governed by state law often end up in federal court because of this 'diversity'jurisdiction. The net result of this system is that both state courts and federal courts are quite accustomed to dealing with legal claims that arise under the law of the other system. Before discussing the different mechanisms for co-ordination that exist within and among the federal and state courts in the United States, it is worth highlighting one additional and critically important distinction between the American situation and its European counterpart. As everyone well knows, in the United States there is a comprehensive system of lower federal courts (there are 94 federal districts, following state lines and subdividing larger states; the total number of authorised positions for U.S. district judges is 652; there are 29 other lower court judges of miscellaneous courts such as territorial courts and the Court of International Trade; and there are 13 federal courts of appeal, with a total of 179 authorised judgeships). Europe has taken a significantly different approach with respect to its own courts; thus far, it has only two courts. Thus, many of the co-ordination mechanisms that the U.S. federal courts use internally are quite unnecessary at a literal level in Europe and will remain so unless or until the Member States decide to create more European courts of first instance. That day may never come, and so this paper proceeds on the assumption that the European system will remain unchanged. The mechanisms that exist in the U.S. system to ensure that the state courts are following federal law faithfully exist as a matter of general law, which is to say that they are not limited to any particular substantive area like antitrust. Many of them, in fact, have direct counterparts already in European law. It may

3 - Courts and Judges

633

not be too much to say they are necessary features of judicial federalism, or bilevel judicial responsibility. First in importance is the Supremacy Clause of the U.S. Constitution, which appears in Article VI, ^J 2. It is worth setting out in full, just to see once again the particular attention the framers of the Constitution paid to the role of judges: This Constitution, and the Laws of the United States which shall be made in Pursuance thereof; and all Treaties made, or which shall be made, under the Authority of the United States, shall be the supreme Law of the Land; and the Judges in every State shall be bound thereby, any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.

Putting that language together with the first sentence of Article III, § 1, of the Constitution, which vests the judicial power of the United States in one Supreme Court and 'in such inferior Courts as the Congress may from time to time ordain and establish,' it is plain that the American Founding Fathers expected that the state judges would from the start be responsible for enforcing the new federal laws. This was why they took care to spell out the fact that the Constitution and other federal laws would be binding on those judges. The only federal court that the Constitution required was the Supreme Court itself, and in this respect the U.S. Constitution is similar to the original treaties establishing the European Communities, which also required only one European Court of Justice. True, in the United States the First Congress immediately passed the Judiciary Act of 1789, which exercised the power to create inferior federal courts. But nothing compelled it to do so. The second principal device for ensuring that state courts adhere to federal law is one that Europe consciously chose not to adopt: direct judicial review by the U.S. Supreme Court of decisions from the highest court of a state. The governing statute is 28 U.S.C. § 1257, which gives the Supreme Court the power to review by writ of certiorari (i.e. discretionary review) 'final judgments or decrees rendered by the highest court of a State in which a decision could be had,' where the validity of a treaty or statute of the United States is drawn into question or whether the constitutionality of a statute of any state is drawn into question, or a right secured by federal law is claimed. Thus, logically enough, the U.S. Supreme Court has no power to review state court decisions that rest on state law; it can act only if a question of federal law was addressed by the relevant state court. In practice, such review is exceedingly rare: out of a total of approximately 7,000 petitions for review that reach the Court from all 50 state supreme courts, the District of Columbia and territorial courts, the thirteen federal circuits, and special courts like the Court of Appeals for the Armed Forces, the Court gives plenary review to only about 90 cases per year (most of which come from the federal courts of appeal). A procedure much more like the European practice of Article 234 references (formerly Article 177) from national courts to the European Court of Justice exists for federal courts who need to obtain a definitive answer on a question of

634

Working Paper X

state law that arises while they are adjudicating a diversity case. For example, Rule 52 of the Circuit Rules for the United States Court of Appeals for the Seventh Circuit permits the court to certify questions to a state court (when the state has in place a certification procedure) when those questions 'will control the outcome of a case pending in the federal court.' Each of the three states within the Seventh Circuit (Illinois, Indiana, and Wisconsin) in fact provides a certification procedure, and it is used regularly, if not frequently. Once the certified question is answered by the state supreme court, the case returns to the federal court for final resolution. Another way in which federal interests are protected for litigants in state court is the process of 'removal' of cases from the state court to the federal court. If a case is filed in a state court that falls within the jurisdiction of the federal court, the defendant normally has the right to remove it to the federal court for the area where the state court is sitting. (See 28 U.S.C. § 1441.) This is done by the simple filing of a petition for removal in the federal court. Any challenge to the removal must be brought before the federal court, but if it finds that the case was improperly removed, then it remands the case back to the state court. Although removal as it is known in the U.S. federal courts is a matter of right, one could also envision the possibility of discretionary petitions to remove. If a case brought before a Member State court were sufficiently important for the development of European law as a whole, it would be possible to give the Commission the right to intervene in the Member State court proceedings and also to give it the right to remove the case to the Court of First Instance, or to take over the proceedings itself (assuming problems with different types of remedies could be resolved). In the area of criminal law, another mechanism by which the federal courts ensure that fundamental rights guaranteed by the U.S. Constitution are being observed by state courts is the federal writ of habeas corpus. State prisoners, even after they have taken all their appeals within the state court system, have the right tofilea petition for a writ of habeas corpus under 28 U.S.C. § 2254 and, if they make a sufficiently strong prima facie showing of a constitutional violation, they can receive a full hearing on their claim in the federal court. This mechanism has probably created as much tension as it has served to facilitate intersystem co-ordination; its utility is probably limited to the criminal law area. Although the next power is little known and rarely used, U.S. law also permits the Attorney General of the United States to send an appropriate officer of the Department of Justice 'to any State or district in the United States to attend to the interests of the United States in a suit pending in a court of the United States, or in a court of a state, or to attend to any other interest of the United States.' 28 U.S.C. § 517 (emphasis added). One could easily imagine a similar power on the part of an appropriate official of the EU to appear in any case implicating the interests of European competition law (or any other area of European law, for that matter), to ensure that the Member State court is fully informed on the relevant issues.

3 - Courts and Judges

635

Another statute leads to much the same result, although it applies to litigation in federal court that affects state interests rather than litigation in state courts that has an impact on federal interests. When a suit is pending in federal court to which a State is not a party (either directly or through the concerned agency), and the constitutionality of any statute of the state is drawn into question, the federal court is required to certify that fact to the attorney general of the State and to permit the State to intervene 'for presentation of evidence, if evidence is otherwise admissible in the case, and for argument on the question of constitutionality.' (28 U.S.C. § 2403(b)) The State at that point has all the rights of a party. I am not aware of any statute that would require the state courts to notify the Attorney General of litigation that affects federal interests (and it would go very much against the grain for Congress to assume that it had any power whatever so to instruct the state courts). In Europe, however, if there is no such procedure in place already, it might be possible to imagine a directive requiring Member States to put in place a procedure under which their courts notified the Commission of significant cases arising under the competition laws. There are a host of other doctrines applicable to the federal courts which, in one way or another, are designed to ensure that they display comity and respect for their state court counterparts. Given the fact that there are only two European courts, these mechanisms probably have little to offer even in the way of broad analogy. They include everything from the jurisdictional limitations on federal court action, to the abstention doctrines (requiring federal courts to dismiss or stay certain cases so that the state may act first), to several acts forbidding (with narrow exceptions) the federal courts to enjoin state court proceedings, to full faith and credit obligations requiring federal courts to recognize and enforce state court judgments. Hardly an argument day goes by in the Seventh Circuit without at least one issue in this family arising in one of the six cases that are being heard. Keeping each set of courts within its proper boundaries and making 'judicial federalism' work is more than a full time job. Within the federal court system, there are a host of tools that are designed to ensure uniformity of result that do not depend on ultimate Supreme Court review. That is possible, of course, but as indicated above such review is granted each year only in a tiny number of cases. If two federal cases dealing with essentially the same subject are pending in two different districts, the parties may ask one of the courts to transfer the case to the other district in the interest of justice, and for the convenience of the parties and the witnesses. (See 28 U.S.C. § 1404(a).) Once both cases are securely lodged in one court, that court may consolidate them for further handling, and thus make sure that inconsistent results do not come about. (See Fed. R. Civ. P. 42.) In larger cases, including over the years many antitrust cases, it often happens that private actions are filed all over the country. Either the judges before whom these cases are pending or one (or more) of the parties may file a request with the Judicial Panel on Multi-district Litigation to consolidate all the cases in one district for purposes of all pre-trial handling. See 28 U.S.C. § 1407. If the Panel

636

Working Paper X

grants the request, the cases all end up in one district. Most often, that is where they stay: parts of the case will be dismissed or resolved without trial, and the rest will often be resolved by settlement. In the rare situation in which they are not resolved in the transferee district, once the cases are ready for trial, they are sent back to the originating courts for trial and entry of whatever final judgment or remedial order is necessary. Comparable mechanisms do not exist among the state courts, although there is a National Center for State Courts (web site ) that is engaged in numerous projects to improve the ability of the state courts to work together. In addition, the chief justices of the state supreme courts have established the Conference of Chief Justices, which also facilitates both administrative and substantive co-ordination. Even with all of this, no honest observer would claim that the highly decentralised U.S. system of courts manages to avoid all conflicts and inconsistent results. The surprise is that things work as well as they do. Information about cases, about judicial reasoning, and about legal developments disseminates with lightning speed through the system, especially in this era of the Internet and computer-assisted legal research. This all bodes well for anyone else who aspires to decentralise to a greater degree, thereby obtaining the benefits of more enforcement resources.

IV. Coherent and consistent substantive law Much of what has been said about the mechanisms for co-ordinating the procedures and results in different courts applies with equal force to the problem of how to avoid conflicting and inconsistent substantive results. Consolidation of cases in a single court (whether at the national/EU level, or at the State/Member State level) allows a single tribunal to make a definitive ruling on a particular matter. The nightmarish prospect of the same merger being reviewed by multiple authorities (a phenomenon that is beginning to reach crisis proportions according to many informed observers), and worse, having those authorities reach different legal conclusions not because they are analysing different markets or business practices, but simply because they have a different perspective on the identical case, is largely eliminated if one tribunal or authority can be given sole or leading responsibility. Mechanisms like the Article 234 reference, or final judicial review in the Supreme Court of the United States, also give afinaland definitive legal answer to questions of general interest. What else, if anything, beyond the procedural mechanisms already discussed should be considered in order to minimise the risk of substantive divergences? Very little, in my view. In fact, I suspect that decentralisation may lead to far less in the way of substantive differences than the pessimists fear. This is so because the national courts of the Member States, very like the state courts in the

3 - Courts and Judges

637

United States, already have internal systems in each country to ensure that they are applying the law properly. Those safeguards differ, to be sure, depending on whether the country follows a common law or a civil law tradition, and depending on how much the traditional models have been adapted to modern-day needs. But they exist in every country of Europe for the same reason they exist throughout the civilised world: the rule of law cannot survive when law itself is uncertain. This is true whether we are speaking of the law of contract or the law of competition; the law of tort or the law of trademarks; the law of patents or the law of pensions. If we can count on the national courts of each member state to keep their own houses in order, then the role the European institutions must play in ensuring substantive consistency of the law becomes far more manageable. Through its analogue to the U.S. Supremacy Clause, the European Court of Justice has already made clear that the national courts are obliged to follow European law. The highest courts of each Member State have complied with their obligation to refer questions of community law to the ECJ, even if the discretionary practice of the lower courts has varied more from country to country. The more difficult question is whether, in the post-decentralised world, the Commission will want to avail itself of a greater number of regulations or other legal instruments that will be entitled to binding effect in the Member State courts. With the sunset of the notification and exemption system, the Commission will have lost one of the ways in which it has communicated its policy understandings to the broader business community. Something else might need to take the place of those rulings, so that the law can continue to evolve in a transparent fashion. The last point, which can only be a question, is to wonder whether decentralisation will occur most effectively if the principal enforcement responsibility at the Member State level remains with the competition law authority, or if it will actually work better if the EU takes this step to its logical conclusion and encourages Member States to expand the private right of action. More litigation means more court decisions, and more information in the system about what is permitted and what is forbidden. Private actions also diminish the opportunity for the competition law bureaucrats to steer only certain cases to the courts. This is obviously a mixed blessing, and a mixed curse. Responsible enforcement discretion plainly keeps the silly cases out of the courts, but on the other hand, strong private rights of action can act as a counterweight to fears about strategic use of competition policy as an industrial policy tool. The debate over the desirability and extent of private actions will affect the development of EC law and the extent to which the White Paper's innovations cause repercussions in the national courts. In the United States, this kind of discussion would take place under the auspices of organisations like the Federal Judicial Center, the National Center for State Courts, think-tanks, and universities. Europe has a similar institution in the Advisory Committee on Restrictive Practices and Dominant Positions, which works with the Commission.

638

Working Paper X

In summary, from the American perspective it appears as if Europe has much to gain and little to apprehend if it chooses to implement the version of decentralisation described by the White Paper. By saying this, I am not necessarily endorsing every proposal within the White Paper. To the contrary, there are some ideas that I would find very distasteful as a judge, both from the standpoint of traditional American separation of powers doctrine and from the standpoint of efficiency. It would be quite unusual, for example, to imagine a court interrupting its proceedings to ask the Commission for an answer to a complex economic question; if the Commission's input is needed, then finding a way to include the Commission as a party or an amicus curiae in the proceeding is preferable. Once a case is properly before the court, the judicial process as it is constituted in each Member State must be permitted to run its course. The final effect of the White Paper may be to reveal the true nature of the rules surrounding competition policy: do they deserve to be called laws, or are they a collection of looser policies that are under constant readjustment because of changing circumstances. To the extent the system has matured over the last 40-some years into a truly legal system, there is no reason to believe that the national courts cannot absorb their new duties. If, or to the extent, the system remains heavily policy-oriented (even if the policy tends to favour open and free competition), the national judges will have trouble precisely because they will have been asked to do something judges cannot do. From more than six thousand kilometres away, in a courthouse near the shores of Lake Michigan, it looks as if the system has reached the kind of stability that supports characterising it as 'law.' If that is right, then the future looks bright for the White Paper in general, and for the quality of work that can be expected from the national courts in particular.

BIBLIOGRAPHY

I. Publications: Baker S. and S. Bishop: 'Market Definition in Non-merger Inquiries', Research Paper for the Office of Fair Trading (forthcoming). Bartosch A. (2000) 'Von der Freistellung zur Legalausnahme—Was geschieht mit der Rechtssicherheit?', 5 WuW, 462-473. Baudenbacher C. (1980): 'Zur funktionalen Anwendung von Paragraph 1 des deutschen und Art. 1 des schweizerischen UWG', ZHR, 145 ff. Baudenbacher C. (1981): 'Machtbedingte Wettbewerbsstorungen als Unlauterkeitstatbestande', GRUR, 19-00. Baudenbacher C. (1997): 'Between homogeneity and independence: The legal position of the EFTA Court in the European Economic Area', 3 Columbia Journal of European Law 2, 169-000. Baudenbacher C. (1999): 'Some remarks on the method of civil law', Texas International Law Journal, 333-000. Behrens P. et al., eds. (1994): EEC Competition Rules in National Courts, Baden-Baden, Nomos. Bellamy C.W. and Child G.D. (1993) Common Market Law of Competition, 4th edition, London, Sweet & Maxwell. Blaise J.B. (1997): 'L'article 10 (2) de l'ordonnance et la contribution au progres economique: du bilan economique a la regie de raison', Gaz. Pal., 13 February, p. 9. Blaise J.B. (1997): 'L'utilisation de la regie de raison en droit interne de la concurrence', in Le droit de I'entreprise dans ses relations externes a lafindu XXieme siecle, Melanges en l'honneur de Claude Champaud, Paris, Dalloz, 85-99. Blanchet T. et al. (1994): The Agreement on the European Economic Area (EEA): A Guide to the Free Movement of Goods and Competition Rules, Oxford Clarendon Press. Boge U. (2000): 'Die Rolle der mitgliedstaatlichen Kartellbehorden bei der Anwendung des EU-Kartellrechts', Siebentes St. Galler Internationales Kartellrechtsforum, 27/28 April 2000. Bos P.-V. (1997): 'Towards a clear distribution of competence between EC and national competition authorities', 16 ECLRev 7, 410-16. Braakman, ed. (1997): The Application of Articles 85 and 86 of the EC Treaty by National Courts in the Member States, European Commission, Brussels. Bruzzone G. and Todino M. (2000): 'Modernisation from a NCA perspective— Interpretation of substantive rules and impact of the reform on decentralised application of EC competition law', paper presented at the INC Conference in Rome, 30-31 March 2000. Bull H. (1994): 'The EEA Agreement and Norwegian law', 5 ECLRev 12, 291-000. Calkins S. (2000): 'California Dental Association: Not a quick look but not the Full Monty', 67 Antitrust Law Journal 3, 495-000. Campbell A. H., Skinner A.S. and Todd W.B. eds. (1976): Adam Smith: An Enquiry into the Nature and Causes of the Wealth of Nations, Oxford Clarendon Press.

640

Bibliography

Cienfuegos M. (1991): 'L'application de la nullite de Particle 85.2 du Traite CEE par les juridictions nationales (avec un examen particulier du domaine des transports aeriens)', Cahiers de Droit Europeen, 317—353. Commission europeene, DG X et Conseil Universitaire Europeen pour l'Action Jean Monnet (1999): 40 ans des Traites de Rome ou la capacite des Traites d'assurer les avancees de la construction europeenne, Actes du Colloque universitaire de Rome organise a la memoire d'Emile Noel, Rome, March 26-27, 1997 (Bruxelles, Bruylant), at p. 63. Cooke J.D. (2000): 'Changing responsibilities and relationships for Community and national courts: the implications of the White Paper', paper given at a conference on 'The Modernisation of European Competition Law: The Next Ten Years', organised by CELS Cambridge and Linklaters & Alliance on December 4, 1999 (not yet published). Deringer A. (1961): 'Rapport fait au nom de la commission du marche interieur ayant pour objet la consultation demandee a PAssemblee parlementaire europeenne par le Conseil de la Communaute economique europeenne sur un premier reglement d'application des articles 85 et 86 du traite de la C.E.E. (document 104/1960-1961)— M. Arved Deringer, Rapporteur', Assemblee Parlementaire Europeenne, Documents de seance 1961-1962, 7 septembre 1961, document 57. Deringer A. (1963): 'Les regies de la concurrence au sein de la CEE', 54 Revue du marche commun, 256-66. Deringer A. (2000): 'Stellungnahme zum WeiBbuch der Europaischen Kommission iiber die Modernisierung der Vorschriften zur Anwendung der Arts. 85 und 86 EG-Vertrag (Arts. 81 und 82 EG)', 1 EuZW, 5-11. Deselears and Obst (2000): 'Weissbuch zum Europaischen Kartellrecht Rechtssicherheit ade? Europaisches Wirtschafts und Steuerrecht', EWS, 41-46. Doherty B. (1994): 'Community exemptions in national law', 15 ECLRev 6, 315-21. Easterbrook F. (1986): 'Antitrust law enforcement in the vertical restraints area: Vertical arrangements and the rule of reason', 53 Antitrust Law Journal, 135-000. Editorial comments (1999): 'European Economic Area and European Community: Homogeneity of legal orders?', 36 CMLRev 4, 697-000. Ehlermann C.-D. (2000): 'La modernisation de la politique antitrust de la CE: une revolution juridique et culturelle", 1 Revue de Droit de l'Union Europeenne, 13-74. Forman J. (1999): 'The EFTA Court five years on: Dynamic homogeneity in practice and its implementation by the two EEA Courts', 36 CMLRev 4, 751-000. Forrester I. (1994): 'Competition Structures for the 21st Century', in Fordham Corporate Law Institute, Matthew Bender & Company, Inc. Forrester I. (1999): 'Modernization of EC Competition Law', in Hawk B., ed.: Fordham Corporate Law Institute, 26th Annual Conference on International Antitrust Law and Policy, forthcoming. Geiger A. (2000) 'Das Weissbuch der EG-Kommission zu Art. 81, 82 EG—eine Reform, besser als ihr Ruf, 6 EuZW, 165-169. Goyder D. (1998): EC Competition law 3rd edition, Oxford, Oxford University Press. Grabitz E. and Hilf M., eds. (1997): EGV, Miinchen, Beck. Graupner R. (1965): The Rules of Competition in the European Economic Community, Martinus Nijhoff. Groeben H., Thiesing J. and Ehlermann C.-D., eds. (1999): Kommentar zum EU-IEGVertrag 211 and II, Baden-Baden, Nomos. Hart H.L.A., Raz J. and Bulloch P.A. eds. (1994): The Concept of Law, Oxford Clarendon Press.

Bibliography

641

Hawk B. (1999): Administrative/Prosecutorial Discretion of Antitrust Authorities', in Hawk B, ed: Fordham Corporate Law Institute 1994.

Hellman, Arthur D. (1998) 'Light on a Darkling Plain: Intercircuit Conflicts in the Perspective of Time and Experience', Supreme Court Review, 247-000. Hoeren T. (1999): 'Europaisches Kartellrecht zwischen Verbots und MiBbrauchsprinzip: Uberlegungen zur Entstehungsgeschichte des Art. 85 EGV, in Huebner U. and Ebke W. F. (eds) Festschrift fur Bernhard Grossfels zum 65 Geburtstag, Verlag Recht und Wirtschaft, p. 405-000. House of Lords (2000): Select Committee on the European Union, 'Reforming EC Competition Procedures', Session 1999-2000, 4th Report, 15 February 2000, London, The Stationary Office. HSH Prince Nikolaus von Liechtenstein (2000): 'Szenarien zur integrationspolitischen Zukunft Liechtensteins', 4th Liechtenstein Europe Symposium of May 5, 2000. Idot, L. (1994): 'L'application du principe de 'subsidiarite' en droit de la concurrence', Dalloz, 37-44. Idot, L. (1996): Arbitrage et droit communautaire', 5 RDAI, 561-581. Idot, L. (1998): 'Subsidiarite et droit de la concurrence: les regies applicables aux entreprises', Law & European Affairs, p. 114-000. Idot, L. (1999): 'La revision du reglement 17/62 ou l'ouverture de la boite de Pandore', Europe, aout-septembre, reperes. Immenga U. and Mestmacker E.-1(1997): Kommentar zum EG Wettbewerbsrecht VO 17, Miinchen, Beck. Jalabert-Doury, N. (1999): 'Livre blanc sur la modernisation de l'application des articles 81 et 82: quel avenir pour le droit de la concurrence communautaire?', 5 RDAI, 497-508. Jones C.A. (1999): Private Enforcement of Antitrust Law in the EU, UK and USA, Oxford University Press. Kapteyn P. J. G. and van Themaat P. V. (1998): Introduction to the Law of the European Communities 5th edition, London, Kluwer Law International, 1334-000. Keyte J. (1998): 'The risks and rewards of Business Review Letters', 12 SPG Antitrust 28, American Bar Association, 107-000. Klimisch A., and Krueger B. (1999): 'Decentralised application of E.C. competition law: Current practice and future prospects', 24 ELRev, 463-482. Kon S. (1982): Article 85 para. 3: a case for application by national courts', CMLRev, 541-000. Kon S. (1999): 'The Commission's White Paper on modernisation: The need for procedural harmonisation', in Hawk B., ed: Fordham Corporate Law Institute 1999, 26th Annual Conference on International Antitrust Law and Policy (forthcoming)Korah V. (1997): An Introductory Guide to EC Competition Law and Practice, Hart Publishing, Oxford. Kraft A. (1980): 'Gemeinschaftsschadliche Wettbewerbsstorungen als unlauterer Wettbewerb?', GRUR , 966-000. Lenaerts K. and van Nuffel P. (1999): Constitutional Law of the European Union, Sweet and Maxwell, London. Lever J. (2000): 'Private Antitrust Enforcement in the European Community', conference paper, Washington, D.C., June 23-24, 2000. Marsden P. (1997): 'Inducing Member State enforcement of European competition law: A competition policy approach to «antitrust federalism))', 18 ECLRev 4, 234-41.

642

Bibliography

Martinez Lage S. (1995): 'La aplicabilidad directa del Derecho de la competencia comunitario y nacional por los tribunales ordinarios espanoles', in Perspectives juridicas actuates. Homenaje a Alfredo Sanchez-Bella Carswell, 236-250. Mestmacker E.-J. (1999): 'Versuch einer kartellpolitischen Wende in der EU.—Zum Weissbuch der Kommission iiber die Modernisierung der Vorschriften zur Anwendung der Art. 85 und 86 EGV a.F. (Art. 81 und 82 EGV n.F.)\ 17 EuZW, 523-529. Mohr Mersing E. (1999): 'The modernization of EC competition law—The need for a common competition culture', in Hawk. B, ed.: Fordham Corporate Law Institute 1999, 26th Annual Conference on International Antitrust Law and Policy, forthcoming. Mok M. R. (1999): 'Naar "modernisering" van het Europees kartelbeleid', 9 SEW, 314-000. Monopolkommission (1999): 'Kartellpolitische Wende in der EU? Zum Weissbuch der Kommission', 28 Sondergutachten. Moschel W. (1976): 'Le controle des prix d'apres le droit allemand et europeen sur les restrictions de la Concurrence', RMC, 263-000. Moschel W. (1999): 'Systemwechsel im Europaischen Wettbewerbsrecht?', Juristenzeitung, 61-67. Mtiller-GraffP. C. and Selvig E., eds. (1999): EEA-EURelations. Nazerali J. and Cowan D. (1999): 'Modernising the enforcement of E.U. competition rules—Can the Commission claim to be preaching to the converted?', ECLRev 8, 442^45. Nehl H. P. (1999): 'Changes in legislation: Constitutional reform and the role of the administrator' (mimeo). Neven D., Nuttal R. and Seabright P. (1992): Merger in daylight; the economics and politics of EC merger control, CEPR, London. Norberg et al. (1993): EEA Law—A Commentary to the EEA Agreement, Stockholm, Fritzes. Norberg S. (2000): 'Perspektiven der kiinftigen Entwicklung des EWR', 4th Liechtenstein Europe Symposium of May 5, 2000. Paulweber M. and Kogel R. (1999): 'Das europaische Wettbewerbsrecht am Scheideweg. Die Reformvorhaben der Kommission zur Modernisierung des europaischen Kartellverfahrens in der Kritik', 11 AG, 500-515. Rittner (1999): 'Zuriick zum Missbrauchsprinzip im EG-Kartellrecht?', Der Betrieb, 1485-000. Rodger B. J. (1995): 'Decentralisation, the public interest and the pursuit of certainty', 16 ECLRev 7, 395-9. Rogne I. (2000): Forklaringsplikt i konkurransesaker, Oslo. Schaub A. (1999): 'Modernization of EC Competition Law: Reform of Regulation No. 17', 23 Fordham Int.L.J., 752-764. Schaub A. (2000) 'Das Europaische Kartellrecht im Jahre 2000', paper presented at Siebentes St. Galler Internationales Kartellrechtsforum, 27/28 April 2000. Schaub A. and Dohms R. (1999): 'Das Weissbuch der Europaischen Kommission iiber die Modernisierung der Vorschriften zur Anwendung der Artt 81 und 82 EG-Vertrag. Die Reform der Verordnung Nr. 17', 11 WuW, 1055-1070. Schroter H. (2000): 'Diskussionsvotum', Siebentes St. Galler Internationales Kartellrechtsforum, 27/28 April 2000. Sevon L. and Johansson M. (1999): 'The protection of the rights of individuals under the EEA Agreement', 24 ELRev 4, 373-000.

Bibliography

643

Shuz R. (1986): 'Controlling forum shopping: The impact of MacShannon v. Rockware Glass LTD', 35 International and Comparative Law Quarterly 2, 374-412. Siragusa M. (2000): 'A critical review of the White Paper on the reform of the EC competition law enforcement rules', 23 Fordham Intl. L. J., 4-000. Siragusa M. (1998): 'Decentralised application of EC competition law—Current practice and future prospects', Discussion Paper prepared for the meeting of the Working Group on Competition Law on October 8 and 9, 1998 in Berlin. Siragusa M. (1999): 'A critical review of the White Paper on the reform of the EC competition law enforcement rules', in Hawk B., ed: Fordham Corporate Law Institute 1999, 26th Annual Conference on International Antitrust Law and Policy, forthcoming. Spiritus—Dassesse A. (1996): 'The role of national courts with respect to European Community competition law', in CD. Ehlermann and L.L. Laudati, eds.: Robert Schuman Centre Annual on European Competition Law 1996, London, KJuwer Law International, 355-000. Sundstrom G.O. Z. and Kauppi M.R.S. (1999): Access to justice, the rule of law and due process', paper presented at the Nordic Conference on the European Union, 6—8 November 1998. Temple Lang J. (1998): Application nationale du droit europeen de la concurrence, Rapport General, FIDE XVIII Congres II, Stockholm, 3-6 June 1998. Temple Lang J. (1996): 'The duties of national courts under EC constitutional law', paper given at Institute of European Law, Cambridge on Feb. 23, 1996 (published at http://europa.eu.int/comm/dg04/speech/indexl9.htm.) Temple Lang J. (1998): 'General Report on the application of Community competition law on enterprises by national courts and national authorities', paper given at FIDE Congress, Stockholm on June 3, 1998 (published at http://europa.eu.int/comm/ dgO4/speech/index98.htm.) Temple Lang J. (1999): 'Decentralised enforcement of Community competition law', paper given at Conference on Competition Enforcement, Dublin on Nov. 11, 1999 (published at http://europa.eu.int/comm/dg04/speech/theme7.htm.) Temple Lang J. (1999): 'Decentralised application of Community competition law', 22 World Competition 4, 3-20. Van Miert K. (2000): 'Modernization of the Competition Rules', BLI Issue 2, International Bar Association. Wesseling R. (1997): 'The Commission Notices on decentralisation of EC antitrust law: In for a penny, not for a pound', 18 ECLRev 2, 94-7. Wesseling R. (1999): 'The Commission White Paper on modernisation of E.C. antitrust law: Unspoken consequences and incomplete treatment of alternative options', ECLRev 8,420-433. Wesserling R. (1997) 'The Commission Notices on decentralization of EC antitrust law: In for a penny, not for a pound', 18 ECLRev, 94-97. Whish R. (1994): 'The Enforcement of EC Competition Law in the Domestic Courts of Member States', 15 ECLRev, 60-67. Wils W.P.J. (1999): 'Notification, Clearance and Exemption in E.C. Competition Law: An Economic Analysis', 24 ELRev, 139-156. II. Official documents: American Bar Association, Section of Antitrust Law (1999): Comments of the Section of Antitrust Law of the American Bar Association on the European Commission's

644

Bibliography

' White Paper on Modernisation of the Rules Implementing A nicies 85 And 86 of the EC Treaty' (published at http://www.abanet.org/antitrust/eumodern.html.) Austrian Government (1999): 'Osterreichische Stellungnahme zum Weissbuch des Europaischen Kommission iiber die Modernisierung der Anwendung der Article 81 und 82 EGV, dated 15 November 1999. European Commission (1993): Notice on Cooperation Between National Courts and the Commission in Applying Articles 85 and 86 of the EEC Treaty, OJ C 39 of 13.2.1993, pp. 6-11. European Commission (1997): Notice on cooperation between national competition authorities and the Commission in handling cases falling within the scope of Articles 85 and 86 EC, OJ C 313/3 of 15/10.1997. European Commission (1999): White Paper on Modernisation of the Rules Implementing Articles 85 and 86 of the EC Treaty, Commission Programme No. 99/027, OJ C 132 of 12.5.1999, pp. 1-47. European Commission (2000): Draft Commission Regulation on the application of Article 81 (3) of the EC Treaty to categories of research and development agreements, OJC 118 of 27.4.2000,4-9. European Commission (2000): Draft Commission Regulation on the application of Article 81 (3) of the EC Treaty to categories of specialisation agreements, OJ C 118 of 27.4.2000,10-13. European Commission (2000): Draft Commission Regulation on the application of Article 81 (3) of the EC Treaty to horizontal cooperation, OJ C 118 of 27.4.2000, 14-39. European Commission, Directorate General Competition (2000): 'White Paper on Reform of Regulation 17, Summary of the Observations', Feb. 29, 2000 (published at http://europa.eu.int/comm/dg04/antitrust/others/wp_on_modernisation/summary_o bservations.htm.) Court of Justice and Court of First Instance (1999): 'The Future of the Judicial System of the European Union (Proposals and Reflections)', submitted to the Council of Ministers of Justice and Home Affairs on May 28, 1999, 1-32. German Government (1999): 'Stellungnahme der Regierung der Bundesrepublik Deutschland zum WeiBbuch der Europaischen Kommission iiber die Modernisierung der Vorschriften zur Anwendung der Artikel 81 und 82 EGV, dated 20 October 1999, at http://www.bundeskartellamt.de/weissbuch.htmWBundesregierung. Monopolkommission (1999): 'Kartellpolitische Wende in der Europaischen Union', Sondergutachten der Monopolkommission gemaB §44 Abs. 1 Satz 4 GWB. OECD (1998): Recommendation of the Council Concerning Effective Action Against Hard Core Cartels, C (98) 35/Final.

INDEX

Arbitration, 510, 511 enforcement, 13 European Court of Justice, 14 Article 81(1): economic analysis, 50, 51 interpretation, 50, 56, 499 relationship with Article 81(3), 96-8 Article 81(3): conditions, 500 direct application, 33-4, 37, 38-9, 42, 45-6, 304,479-81 exemption decisions, 50-1, 95-6 justiciability, 40, 43, 243-5, 484 national courts application of 539—43, 593-612,617-25 see also, Judiciary; National courts relationship with Article 81(1), 96-8 scope, 50 temporary exemptions, 499-500, 504 use of provisions, 44-5 Associated countries: association agreement, 391-3 generally, 387 Romania: competition enforcement system, 387-91 proposed reforms and, 393-7 Block exemptions: compatibility, 173 modernisation of EC antitrust policy: basis, 59-60 market share thresholds, 59-60 withdrawal, 142 Brussels Convention, 296-7, 300 Business community: modernisation of EC antitrust policy, 548-50 Certainty, legal, see Legal certainty Co-operation, 217-21 Commission and national authorities, 15 need for, 18-19 Coherence, 512-13, 547-8 action, 326-9 allocation of cases, 275-6 application: national competition authorities, 338—44 national courts, 344-8 associated countries 387-97, see also, Associated countries Commission, role of, 289-92 conditions for ensuring, 298

conflict of jurisdiction, 325 consistency, 294 Delimitis, 419-22 desirability, 471-3 dual control, 275 economics, 278-80 enforcers, 302-3 European Agreements, 283 forum shopping 329—33 see also. Forum shopping framework conditions, 284-5 German view, 277—8 judicial views on, 271—3 meaning, 273-5, 294-6 mechanisms to ensure, 381-2 national competition authorities, 282-3 non-action, 325-6 parallel competencies, 348-51 broader context of enforcement, 401-2 co-operation by national courts, 419-22 Commission decisions, effect of, 422—5 conflicts between two different legal orders, 413-18 Delimitis, 419-22 discrimination created by Commission's exemption monopoly, 400-1 framework to enhance coherence, 403—10 harmonisation of national procedures, 425-7 importance of issue, 399-400 national competition authorities, specific measures applicable to, 410-12 national courts, specific measures applicable to, 412-13 primacy, 413-18 same legal order, conflicts in, 419-22 Walt Wilhelm, 413-18 proposed measured, 471-3 United States, 285-8 see also. Decentralisation; National competition authorities Comfort letters, 37, 49 modernisation of EC antitrust policy, 63-4 Commission: fact-finding by, 508 focus of, 15 jurisdiction, 292^1 means at disposal of, 303 national authorities: co-operation with, 15 policy, 481-3

646

Commission (cont.): positive decisions, 43-4, 46, 48, 63-4, 171-2, 304 prohibition of agreement, 500 role, 21-2, 43 soft law, 505-6 status, 292-3 Compatibility: accepting commitments, decisions, 173-4 adoption of new Regulation, 18 amendment of EC Treaty, 145—76 administrative decisions, 151-64 control of abuse, proposed system not, 145-9 generally, 145 implementing legislation, 150-1 interpretation by court, 164 justiciability of Article 81(3), 164-7 range of permissible systems, 149-50 authorisation, 245-7 block exemptions, 173 damages, 235-6 direct applicability of Article 81(3), 33-4, 37, 38-9, 42, 45-6 ECJ jurisprudence, 229-32 ECSC, example of, 94-5 exemption decisions, 95-6 giving effect to Article 81 principles, 228-9 historical background, 77-94 injunctions, 235-6 internal consistency, 171-4 interpretation of Article 85, 94-5 issue, 242 justiciability of Article 81(3), 243-5 national competition authorities, 237-8 national courts: application of Article 81, 232-5 function of national courts, 243 notifications, 108—9 objections, 27-8, 167-71 positive decisions, 171—2 practical considerations: accession, 109-10 change, dangers of, 104-5 fictitious advantages of reform, 103-4 legal certainty, 102—3 notifications, 108-9 thirty entities, enforcement before, 105-8 primacy of Community law, 238-40 relationship between Article 81(1) and Article 81(3), 96-8 rule of reason, 101—2 self-help, 100-2 subsidiarity, 96 Competition law: certainty, legal, 8-10, 19 approach to problem, 37 lack of, 47 United States, 25-7 see also, Legal certainty

Index

sphere of, 511 Competition policy, EC: coherent, 7-8 modernisation of EC antitrust policy, 52 reform: overall, 4 Decentralisation: allocation of cases, 355-6 background to proposed reforms, 458-9 benefits, 361-3 Commission's enforcement competencies, 358-9 enforcement, 380-1 forum shopping, 355-6, 444-6 French point of view, 335-51 inconsistent enforcement of Article 81, 363-5 increasing consistency of enforcement, 365-7 isolated application of EC competition law, 337-^8 modernisation of EC antitrust policy, 19-20, 207-21 generally, 207-10 multiple enforcement, 210—13 subsidiarity, 210-11 policy, 380-1 practical workings of system, 379-85 radical, 335-51 requirements for effective application, 353-5 risk assessment, 323-5 sacrifice of, 353-9 threat to consistency: Commission intervention, 557-8 generally, 551-2 judicial review, 559-63 national courts, 555 notification, 555-6 problem of consistency, 552 potential conflicts, 553 reserve power to prohibit, 558-9 territoriality, 553-4 uniform application of Community competition law, 356-8 see also, Coherence Depression cartels, 29 Dominance of Community principles, 309-10 Economic effects based approach, 16-17, 304 EEA Agreement, 477, 521-2 competencies, 523-4 competition law regime, 522—6 enforcement, 524-6 impact of modernisation on, 477-9 implications of modernisation for, 526-37 substantive law, 522 Efficiency, 501-2 enforcement, 7, 125-6, 247-50 modernisation of EC antitrust policy, 39—40 adjustment, 126 compliance, 124

Index

dimensions, 124 dynamic, new, 127—8 effect on efficiency of business, 126-7 enforcement, 125-6, 247-50 generally, 123 EFTA Court: one-stop shopping, 477 EFTA Surveillance Authority, 477 Enforcement: allocation of cases between national authorities, 71 arbitration, 13 Commission's role, 21—2, 43 criticisms of current system, 22—5 decentralisation, 28-30, 71-3 decentralisation, see Coherence; Decentralisation efficiency, 7, 125-6, 247-50 high market shares, firms with, 16-17 modernisation of EC antitrust policy, 30-2 allocation of cases, 71 coherence, 71—3 control of abuse, 223-4 decentralisation, 71-3 efficiency, 247-50 guidelines, 60-1 legal certainty, 71—3 legal exception system, 224-6 national courts, 73 see also, Modernisation of EC antitrust policy multiple, 125-6 national courts, 48-9 private, 479, 498, 504-5 problems with current system, 4-5 United States: cooperation, 430—5 developments, 429—41 generally, 429-30 inherent inconsistencies, 435-7 proposals for modification of dual enforcement regime, 438 European Agreements, 283 European Court of Justice: arbitration, 14 direct action, 515 interpretation of judgments, 305 intervention by, 515 length of time before judgment, 507 role, 44 use, 14 Exemptions: administrative nature, 301-2 block, see Block exemptions Expert witnesses, 509 Fines: legal certainty, 253 Forum shopping, 329-33, 454, 492, 506 decentralisation, 355-6, 444-6

(Al

hazards of, 355-6 modernisation of EC antitrust policy, 331-2, 544-5 mutually exclusive application of EC and national competition rules, 330-1 national courts, 454 one-stop shop, 332—3 risks, 453-4, 507-8, 511 Hard-core cartels: modernisation of EC antitrust policy, 69 Hard-core restrictions: legal certainty, 253 Horizontal agreements: modernisation of EC antitrust policy, 68-9 Individual exemptions, 13 national competition authorities, 142 Italian Banks, 465-71 Japan: modernisation, view of, 28-9, 197-205 Judiciary: ability, 619-21 application: Article 81(1), 487-9, 587-9 Article 81(3), 589-91 ascertaining law, 495—7 federalism, techniques of, 627-38 ability of national courts to decide competition cases, 628-31 background, 627-8 co-ordination, effective, 631-6 coherent and consistent substantive law, 636-8 issues for, 565-70 language, 498, 510 preliminary rulings, 591-2 procedure: competence of Commission, 584—5 general issues, 579-80 inconsistencies, 585-6 influence on application of Community law by Commission, 586 jurisdiction, 580-2 national courts'jurisdiction, 284—5 territorial jurisdiction, 583-4 types of civil lawsuits, 582 role, 50 specialisation, 575-6 support for national judges, 571-7 changes in Community's judicial architecture, 574-5 generally, 571-2 linkage of public and private enforcement, 576-7 specialisation, 575-6 task of national judge, 572-4 see also, National courts

648

Language, 498, 510 Legal certainty, 545-7 clarification of rules, 256-7 competition authorities, 254-6 decentralisation, 143 direct applicability of Article 81(3), 254 fines, 253 forum: generally, 141 national competition authorities, 142-3 national courts, 143-4 shopping, 143 withdrawal of block exemption, 142 hard-core restrictions, 253 importance, 252-3 meaning, 129 modernisation of EC antitrust policy, 129-44 national competition authorities, 142-3 national courts, 143—4, 256 opinions, 257-8 origins, 129 outcome, predicatability of: generally, 130 increasing, proposals for, 137—40 permission, per se rules of, 132-4 prohibition, per se rules of, 130-2 rules of reason, 134-7 present system, under, 253—4 requirements, 252-3 Liberalisation, 297-8, 299-300 Market share thresholds: modernisation of EC antitrust policy, 64 Modernisation of EC antitrust policy, 613-16 abolition of notification system, 68-71 advantages, 261 advisors, 548-50 approach, 16 Article 81 and, 64-5 assessment, 64-5, 259-67 associated countries 387-97 see also, Associated countries block exemptions: basis, 59-60 market share thresholds, 59-60 Bunderskartellamt's opinion on, 67—73 burden of proof, 305-7, 374-6 burden shifted to business, 57-9 business community, 548-50 case law, previous, 59 comfort letters, 63—4 Commission's role, 21 compatibility with EC Treaty, see Compatibility competition analysis, 55-6 competition policy, 52 consistent application of Community competition rules, 61—3 constitutional problems raised by, 505 decentralisation, 19-20,207-21

Index

co-operation, 217-21 competition law, 211-13 generally, 207-10 multiple enforcement, 210-16 subsidiarity, 210-11 United States and, 216-17 directly applicable exemption system, 56-57, 263 ECJ, 264 economic effects based approach, 16-17 effect, 18 efficiency, 39-40, 123-8 adjustment, 126 compliance, 124 dimensions, 124 dynamic, new, 127-8 effect on efficiency of business, 126-7 enforcement, 125-6,247-50 generally, 123 enforcement, 30-2 allocation of cases, 71 coherence, 71—3 control of abuse, 223-4 efficiency, 247-50 guidelines, 60-1 legal certainty, 71-3 legal exception system, 224-6 national courts, 73 "families of reform", 20-2 forum shopping, 331-2, 544-5 generally, 241—2 German criticism, 35-6, 40-2, 45 guidelines for enforcement, 60—1 hard-core cartels, 69 horizontal agreements, 68-9 horizontal co-operation agreements, 69-71 implementation, background of, 59 introduction of White Paper, 55 issues raised by, 64-5 Italian Banks, 465-71 Japanese view of, 197-205 legal certainty, 129-44 legal exception created, whether, 376-9 market share thresholds, 64 nature of authorisation, 198-202 nature of control, 202-4 network effect, 14 non-infringement decisions, 35-6, 250-2 notifications, 68-71 objections to, 262-3 objectives, 39, 73, 259, 370-4 options for reform, 56-7 positive decisions, 63-4 practical effect, 24-5 present position, 3-4 process, 260-1 Regulation 17/62, 3-4 repercussions, 384-5 significance for countries outside EU, 197-9 structure of Article 81, 123

Index

subsidiarity, 210-11 summary of proposals, 63, 73,453 uniformity, 25<M>0 United States experience, 216-17 National competition authorities, 301 adoption of decision by Commission, 412 allocation of cases between, 19, 71,410-11, 500,514-15,518 application of EC competition law, 444-6 block exemptions, 142 co-operation with Commission, 15 cohabitation concerning application of Article 81(3), 459-65 compatibility, 237—8 consultation of Commission on prohibition decisions, 411 double barrier, 303 Eastern Europe, 301 forum shopping, 454 harmonisation of structure, powers and procedures, 382^1 individual exemptions, 142 inducement to apply EU law, 13 legal certainty, 142-3 minimum institutional standards, 298-9 national courts, relationship with, 486-7 network of competition authorities, 307-8 risks of conflicts: national courts and NCAs, between, 449-52 NCAs, between, 446-9 role, 315-18 using, 483-4 withdrawal of case by Commission, 411—12 see also, Coherence National courts: administrative functions, 484-6, 489-1, 492-3,516-18 appeals, 485-6 application of EC competition law by, 446, 493-5 Article 81(3), application of, 539-43, 593-612,617-25 ability to judges, 619-21 administrative decentralisation, 597-600 CAMPSA doctrine, 608-9 changes, 617-19 generally, 593 harmonisation, 623-5 inconsistency problem, 621-3 legal exception system, 609-12 problems, 600-7 role of national judges, 593-7 solutions to problems, 607-12

649

specialisation, 607-8 ascertaining law, 495-7 compatibility: application of Article 81, 232—5 function of national courts, 243 forum shopping, 454, 492 jurisdiction, 584-5 legal certainty, 143-4, 256 national competition authorities: relationship with, 186—7 risks of conflicts between, 452 parallel proceedings, 484 role, 51-2, 319-22 specialisation, 491-2, 513-14 specialised courts, 498-9, 502-4 see also. Judiciary Notifications, 13,22 abolition proposals, 68-71 bureaucracy, 56 compatibility, 108-9 modernisation of EC antitrust policy, 68-71 tactics, 293 Opinions, 14-15,22 legal certainty, 257-8 Positive decisions, 63-4 Regulation 17/62: modernisation of EC antitrust policy, 3-4 Soft harmonisation, 15-16 Subsidiarity: compatibility, 96 modernisation of EC antitrust policy, 210-11 United States: Chevron doctrine, 515 coherence, 285-8 cooperation, 430-5 developments, 429—41 enforcement: cooperation, 430-5 developments, 429-41 generally, 429-30 inherent inconsistencies, 435-7 proposals for modification of dual enforcement regime, 438 legal certainty, 24-6 multiple enforcement, 216-17 recent developments, 429-41 White Paper on Modernisation, see Modernisation of EC antitrust policy