The EU Comitology System in Theory and Practice: Keeping an Eye on the Commission?

Palgrave Studies in European Union Politics Edited by: Michelle Egan, American University, USA; Neill Nugent, Manchester Metropolitan University, UK and William Paterson OBE, University of Aston, UK. Editorial Board: Christopher Hill, University of Cambridge, UK; Simon Hix, London School of Economics, UK; Mark Pollack, Temple University, USA; Kalypso Nicolaïdis, University of Oxford, UK; Morten Egeberg, University of Oslo, Norway; Amy Verdun, University of Victoria, Canada; Claudio M. Radaelli, University of Exeter, UK and Frank Schimmelfennig, Swiss Federal Institute of Technology, Switzerland. Following on from the sustained success of the acclaimed European Union Series, which essentially publishes research-based textbooks, Palgrave Studies in European Union Politics publishes cutting-edge, research-driven monographs. The remit of the series is broadly defined, both in terms of subject and academic discipline. All topics of significance concerning the nature and operation of the EU potentially fall within the scope of the series. The series is multidisciplinary to reflect the growing importance of the EU as a political, economic and social phenomenon. Titles include: Ian Bache and Andrew Jordan (editors) THE EUROPEANIZATION OF BRITISH POLITICS Thierry Balzacq (editor) THE EXTERNAL DIMENSION OF EU JUSTICE AND HOME AFFAIRS Governance, Neighbours, Security Jens Blom-Hansen THE EU COMITOLOGY SYSTEM IN THEORY AND PRACTICE Keeping an Eye on the Commission? Falk Daviter POLICY FRAMING IN THE EUROPEAN UNION Renaud Dehousse (editor) THE ‘COMMUNITY METHOD’ Obstinate or Obsolete? Kenneth Dyson and Angelos Sepos (editors) WHICH EUROPE? The Politics of Differentiated Integration Michelle Egan, Neill Nugent and William E. Paterson (editors) RESEARCH AGENDAS IN EU STUDIES Stalking the Elephant Kevin Featherstone and Dimitris Papadimitriou THE LIMITS OF EUROPEANIZATION Reform Capacity and Policy Conflict in Greece Stefan Gänzle and Allen G. Sens (editors) THE CHANGING POLITICS OF EUROPEAN SECURITY Europe Alone? Eva Gross THE EUROPEANIZATION OF NATIONAL FOREIGN POLICY Continuity and Change in European Crisis Management Adrienne Héritier and Martin Rhodes (editors) NEW MODES OF GOVERNANCE IN EUROPE Governing in the Shadow of Hierarchy Wolfram Kaiser, Brigitte Leucht and Michael Gehler TRANSNATIONAL NETWORKS IN REGIONAL INTEGRATION Governing Europe 1945–83 Hussein Kassim and Handley Stevens AIR TRANSPORT AND THE EUROPEAN UNION Europeanization and its Limits

Robert Kissack PURSUING EFFECTIVE MULTILATERALISM The European Union, International Organizations and the Politics of Decision Making Katie Verlin Laatikainen and Karen E. Smith (editors) THE EUROPEAN UNION AND THE UNITED NATIONS Intersecting Multilateralisms Esra LaGro and Knud Erik Jørgensen (editors) TURKEY AND THE EUROPEAN UNION Prospects for a Difficult Encounter Karl-Oskar Lindgren and Thomas Persson PARTICIPATORY GOVERNANCE IN THE EU Enhancing or Endangering Democracy and Efficiency? Ingo Linsenmann, Christoph O. Meyer and Wolfgang T. Wessels (editors) ECONOMIC GOVERNMENT OF THE EU A Balance Sheet of New Modes of Policy Coordination Hartmut Mayer and Henri Vogt (editors) A RESPONSIBLE EUROPE? Ethical Foundations of EU External Affairs Philomena Murray (editor) EUROPE AND ASIA Regions in Flux Daniel Naurin and Helen Wallace (editors) UNVEILING THE COUNCIL OF THE EUROPEAN UNION Games Governments Play in Brussels David Phinnemore and Alex Warleigh-Lack REFLECTIONS ON EUROPEAN INTEGRATION 50 Years of the Treaty of Rome Sebastiaan Princen AGENDA-SETTING IN THE EUROPEAN UNION Carolyn Rowe REGIONAL REPRESENTATION IN THE EU Between Diplomacy and Interest Mediation Emmanuelle Schon-Quinlivan REFORMING THE EUROPEAN COMMISSION Roger Scully and Richard Wyn Jones (editors) EUROPE, REGIONS AND EUROPEAN REGIONALISM Asle Toje AFTER THE POST-COLD WAR The European Union as a Small Power Richard G. Whitman and Stefan Wolff (editors) THE EUROPEAN NEIGHBOURHOOD POLICY IN PERSPECTIVE Context, Implementation and Impact Richard G. Whitman (editor) NORMATIVE POWER EUROPE Empirical and Theoretical Perspectives

Palgrave Studies in European Union Politics Series Standing Order ISBN 978–1–4039–9511–7 (hardback) and ISBN 978–1–4039–9512–4 (paperback) (outside North America only) You can receive future titles in this series as they are published by placing a standing order. Please contact your bookseller or, in case of difficulty, write to us at the address below with your name and address, the title of the series and the ISBN quoted above. Customer Services Department, Macmillan Distribution Ltd, Houndmills, Basingstoke, Hampshire RG21 6XS, England

The EU Comitology System in Theory and Practice Keeping an Eye on the Commission? Jens Blom-Hansen Professor, Department of Political Science, Aarhus University, Denmark

© Jens Blom-Hansen 2011 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No portion of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, Saffron House, 6-10 Kirby Street, London EC1N 8TS. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The author has asserted his right to be identified as the author of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2011 by PALGRAVE MACMILLAN Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN 978–0–230–24142–8 This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Blom-Hansen, Jens. The EU comitology system in theory and practice : keeping an eye on the Commission? / Jens Blom-Hansen. p. cm. Includes bibliographical references and index. ISBN 978–0–230–24142–8 (alk. paper) 1. European Commission—Committees. 2. Decision making— European Union countries. 3. Delegated legislation—European Union countries. 4. Legislative oversight—European Union countries. I. Title. JN33.5.B56 2011 341.242 226—dc23 2011029567 10 9 8 7 6 5 4 3 2 1 20 19 18 17 16 15 14 13 12 11 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne

Contents

List of Figures

vi

List of Tables

vii

List of Boxes

ix

Foreword and Acknowledgements

x

1 Introduction

1

2 A Crackpot Concept

15

3 A Delegation Perspective on Comitology

34

4 The Origins of Comitology

53

5 What is at Stake? A Case Study of the 2006 Comitology Reform

72

6 Comitology Preferences in Daily Legislation

94

7 Comitology Across Policy Areas

118

8 The Comitology System in Daily Practice

143

9 Conclusion

176

Notes

193

Literature

197

Index

212

v

List of Figures

3.1 Modelling the choice of administrative regime for common policies as a pure coordination game nested within a ‘battle of the sexes’ game 3.2 Understanding the daily workings of the comitology system 6.1 Regulations and directives proposed and enacted under the 1999 comitology decision (no. of acts)

vi

39 50 104

List of Tables

2.1 2.2 2.3 2.4 2.5 3.1 5.1 5.2 6.1 6.2 6.3 6.4 6.5 7.1 7.2 7.3 7.4 7.5 7.6 8.1

Treaty regulation of the comitology system over time The list of comitology procedures over time An example of a comitology provision in a directive Standard rules of procedure for comitology committees (excerpts) The comitology system according to institutional affiliation in the Commission in 2008 Overview of hypotheses, data and chapters What is at stake? Manifestations of institutional interests in the 2006 reform of the comitology system Instruments to pursue institutional interests in the 2006 reform of the comitology system Comitology preferences, 1999–2006: Commission versus Council Comitology preferences, 1999–2006: Commission versus European Parliament Comitology preferences, 1999–2006: European Parliament versus Council Comparison across time of Commission–Council agreement on comitology procedures (per cent) Pure cases of inter-institutional disagreement on comitology under the 1999 comitology decision Comitology procedures across policy areas, 1999–2006 (per cent) Summary statistics on variables used in logistic regression models Logistic regression analysis of the choice of comitology over no comitology Probability of choosing comitology over no comitology for selected types of conflict and complexity Logistic regression analysis of the choice of stringency of comitology control Probability of choosing stringent comitology for selected types of conflict and complexity The Brussels–Copenhagen comitology connection vii

22 24 27 28 31 50 90 92 108 109 110 111 113 119 134 136 138 139 141 154

viii List of Tables

8.2 Descriptive statistics on all variables used in regression analyses in Chapter 8 8.3 Comitology committees scoring low and high values on the issue complexity measure 8.4 Regression analysis of the discretion of the Danish comitology representatives 8.5 Factor analysis of the behavioural dispositions of the Danish comitology representatives 8.6 Regression analysis of the behavioural dispositions of the Danish comitology representatives 8.7 Factor analysis of the Danish comitology representatives’ evaluation of the nature of meetings in their committee 8.8 Factor analysis of the Danish representatives’ evaluation of the Commission’s roles in the comitology committees 8.9 Regression analysis of the Commission’s behaviour in the comitology committees

157 160 161 164 164 168 170 172

List of Boxes

6.1 8.1

Data set on comitology in EU regulations and directives, 1999–2006 The Danish comitology survey

ix

104 149

Foreword and Acknowledgements

This book is the result of work conducted on the comitology system over several years. Papers that ended up as chapters or parts thereof have been presented at a number of occasions including the 2010 Annual Meeting of the American Political Science Association, the 2007 Annual Meeting of the Midwest Political Science Association, the 2007 General Conference of the European Consortium for Political Research, the 2008 Pan-European Conference on EU Politics organized by the ECPR’s Standing Group on the European Union and the 2008 Conference of the Nordic Political Science Association. I am thankful for comments and suggestions from participants at these meetings. For detailed feedback on various drafts, I am indebted to many people. I would especially like to mention Gijs Jan Brandsma, Claudio Radaelli, Kutsal Yesilkagit, Ulf Sverdrup, Åse Gornitzka and Kieran St Clair Bradley. My excellent colleagues in the public administration division at the Political Science Department here in Aarhus have also read and commented intensively on different parts of the book. Their efforts are much appreciated. I wish to thank Manuel Szapiro and Paolo Ponzano, both from the Commission’s General-Secretariat, Marta Arpio Santacruz from the Council’s Legal Service and Kieran St Clair Bradley from the European Parliament’s Legal Service. They have all taken time to let me ask questions and discuss ideas on the 2006 comitology reform. I also wish to thank the many civil servants in the ministries of the Danish government in Copenhagen who took time from their busy schedules to fill out the questionnaire that constitutes the data for Chapter 8. I owe Kutsal Yesilkagit special thanks for help in organizing my stay at the Utrecht School of Governance in the summer of 2006. I also owe Cathie Jo Martin many thanks for help in organizing my stay in Boston in the summer of 2009. These occasions gave highly valued breathing spaces to work on the project. I am grateful to the globalization project organized by the Faculty of Social Sciences at Aarhus University for funding my travel and datacollection expenses. Some of the material in Chapter 4 appeared in the Journal of European Public Policy, vol. 15(2), 2008, pp. 208–27, and parts of Chapter 5 in the European Law Journal, vol. 17(3), 2011, pp. 343–64. x

Foreword and Acknowledgements xi

I am grateful to these journals’ publishers for permission to reuse the materials here. I owe many thanks to Annette Andersen here in Aarhus for her superb secretarial skills and help in preparing the manuscript. Finally, thanks to Marina for always being there and taking on the job as my first critic. Jens Blom-Hansen, Aarhus

This page intentionally left blank

1 Introduction

Second-tier rule-making On 8 July 2010, officials from the Transport Ministries in the EU1 member state capitals travelled to Brussels. Their destination was the headquarters of the European Organisation for the Safety of Air Navigation (Eurocontrol) at Rue de la Fusée 96. They were to meet with the Commission2 to discuss its latest initiative under the single European sky programme launched around the turn of the millennium to improve regulation of the European airspace. On the agenda for their meeting was a proposal from the Commission to amend regulation no. 1794/2006, a Commission regulation establishing a common charging scheme for air navigation services in the airspace over Europe. The Commission now wanted to amend it by a new Commission regulation, but could not proceed with its own proposal unless the officials from the national Transport Ministries formally approved. The reason was that the officials constituted a comitology committee, the so-called single sky committee, which was established in 2004 to support the Commission in the implementation of the single European sky programme. In addition to member state representatives the committee consisted of observers from Iceland, Norway, Switzerland and Eurocontrol. Most of the members knew each other well because they had met regularly since 2004, four to five times every year. The meeting on 8 July 2010 was their 36th meeting. When the committee convened in the Europa conference room at Eurocontrol, the Commission, as usual, chaired the meeting. Having achieved the committee’s approval of the meeting’s agenda and the minutes of the committee’s previous meeting, the chairman turned to item 2a on the agenda, the amendment of the Commission regulation 1

2

The EU Comitology System in Theory and Practice

on charging schemes for air navigation services. The chairman asked for a formal vote on the proposal. Only member state representatives were allowed to take part in this formal exercise. Using the same voting system as the Council of Ministers each national delegation then formally gave a vote on behalf of its national government. This resulted in 273 weighted votes in favour of the Commission’s proposal, no votes against and 58 abstentions. The Commission was happy; it had obtained a qualified majority in favour of its proposal. The Commission could now proceed with its proposal. This was done in the autumn of 2010 when the College of Commissioners formally adopted the regulation. It was then duly published in the EU’s Official Journal on 17 December 2010 and entered into force three days later. The single European sky programme had taken another step forward.3 The position of the single sky committee illustrates the importance of the EU’s second-tier rule-making system. The first tier, where the Commission proposes legislation which is then approved, modified or rejected by the Council4 and the European Parliament, is familiar and visible. But underneath that top layer hundreds of second-order decisions are taken each month to complete or implement rules adopted by the Council and the Parliament. These decisions are taken by the Commission and formally adopted as Commission regulations, directives or decisions. However, in most cases the Commission must first submit its draft rules to a committee such as the single sky committee. The single sky committee is one out of 250–300 committees that monitor the Commission’s rule-making activities. Sometimes it is enough that the Commission consults with the committee; sometimes, like in the case of the single sky committee, it must obtain the formal approval of the committee. The purpose of this committee system is to allow the member states to monitor the Commission’s implementation of legislation adopted by the Council and the Parliament. The system is known as the EU comitology system. Although second-tier rule-making in the comitology system is less familiar and visible than activities at the first tier, it is important. Every year the comitology committees formally approve 1500–2500 implementation measures – Commission regulations, directives and decisions – which complete or implement primary regulation in all policy areas. Examples include rules on animal welfare, the organization of agricultural markets, energy networks, cross-border cooperation within higher education, harmonization of product requirements under the internal market programme, handling of waste to protect the environment, the operation of the EU structural funds, road transportation

Introduction 3

systems and many other things. There is far more rule-making at the second tier than at the well-known first tier, at least in a quantitative sense. This is why all EU actors take a keen interest in the comitology system and its institutional design. As aptly summarized by the weekly magazine European Voice (15 July 2010, p. 6) in a report on the Lisbon Treaty’s impact on the comitology system: What decision-making rights should be given to the [comitology committees], by whom and under what conditions, is a matter of great importance. This is not the Parliament and the Council squabbling over whether the Commission should do the housework. It is about who gets to redesign the house. This book is a journey into the universe of the comitology committees. It investigates their daily operation, their institutional set-up, the system’s development over time and its broader role in the EU system.

A closer look at the comitology system The body of EU law, the acquis communautaire, consists of several thousand legal acts, most of which are made by the Commission as delegated rule-making. Only a minority is made by the Council of Ministers and the European Parliament. In fact, according to Eur-Lex, the EU’s official online register of EU law, the Commission has issued 69 per cent of all EU legislation that is in force today.5 Although the Commission is thus an important rule-maker, it is not autonomous. As the story of the single sky committee showed, the Commission is in many areas monitored by committees of member state representatives, known as comitology committees. These committees are not involved in all Commission rules; but almost. There are no official accounts, but an investigation of all Commission rules made in the period 2004–08 – a total of 11,056 rules – shows that 55–65 per cent were checked by a comitology committee (Brandsma, 2010a, p. 33). In other words, comitology is a standard operation procedure for delegated rule-making in the EU system. The committees in the comitology system are gatekeepers. They cannot amend or reject Commission proposals, but may refer them to the Council if they disagree with them. There is a lot of variation in the exact working rules of committees. Until the Lisbon Treaty they were specified in so-called comitology decisions made by the Council, for example Council of Ministers (2006a). After the Lisbon Treaty they are made as

4

The EU Comitology System in Theory and Practice

regulations by the Council and the European Parliament. These working rules specify the decision rule in the committees to refer proposals to the Council, the timing of Council control, the default condition if the Council does not act and the degree of involvement of the European Parliament. The comitology system has changed considerably over time. It was established in the early 1960s when the Common Agricultural Policy was introduced. With the growth of Community legislation, the practice of establishing comitology committees also grew. It proved to be a practical solution for the member states to the problem of delegating decision-making powers to the Commission without losing control (Bergström, 2005). The system did not have any solid treaty foundation until 1987 when the Single European Act amended article 202 (ex-145). From then on the treaty recognized that the Council might ‘impose certain requirements’ upon the Commission when delegating implementing powers (Haibach, 2000). Following this change of the treaty the Council adopted its first comitology decision specifying the operating rules of the committees in the system (Council of Ministers, 1987). The rules were overhauled and simplified in 1999 when the Council adopted its second comitology decision (Council of Ministers, 1999a). But simplification was rolled back in 2006 when the 1999 decision was amended and the so-called regulatory procedure with scrutiny was introduced (Council of Ministers, 2006b). The Lisbon Treaty, which is introduced in more detail in the next chapter, paves the way for a new comitology regime; partly because it introduces a distinction between legislative, delegated and implementing acts and only specifically operates with a comitology system for implementing acts; partly because the comitology framework rules, which until now have been specified by the Council’s comitology decisions, in the future will be set in regulations enacted by the codecision procedure6 and thus give the European Parliament more influence on the rules (Craig, 2008; Ponzano, 2008; Hoffmann, 2009). The comitology system is still to a large extent a white spot on the EU map. Many aspects of the system remain under-researched, and there is a glaring imbalance between the impressive amount of research on first-tier EU decision-making by the Council and the Parliament and the limited research on second-tier EU rule-making by the Commission and the comitology committees. This book seeks to redress that imbalance and find answers to three fundamental, unresolved questions about the comitology system.

Introduction 5

Comitology: A unique system – but why? Parliamentary control of delegated rule-making by the executive is a problem known from all political systems, but solving it by a comitology system is unique to the EU. How and why did comitology start? The dominant answer in the literature is a functional one. The system was created in the early 1960s as a response to the Council’s need to delegate implementing powers to the Commission without losing control. With only slight variations this is the standard textbook explanation (Nugent, 2003, pp. 131–40; Pollack, 2003a, pp. 114–52; Hix, 2005, pp. 52–9). Even special studies focusing on the development of comitology over time devote surprisingly little attention to the origins of the system (Vos, 1997; Dogan, 1997; Haibach, 2000; Bergström, 2005, pp. 43–57). However, functionalist reasoning may be misleading. The functions of institutions may be quite different from the motives that led to their establishment. At the very least, functionalist reasoning about institutional origins should be a hypothesis, not a premise (Pierson, 2000). But even if the functional reason is correct, it overlooks the importance of choice. If the member states really needed to delegate tasks, but at the same time worried about losing national control, comitology represented an institutional solution. But other solutions could easily be imagined, for example solutions used in their own national systems to solve similar problems, so why was comitology chosen? This is the first important question that this book addresses.

Full of sound and fury, signifying – what? Almost all observers of comitology are struck by the inter-institutional rivalry about the system. Many use warlike metaphors to describe the political processes behind the development of the system. For instance, Franchino (2007, p. 283) talks of a ‘legislative battle on comitology’ in his investigation of the European Parliament’s endeavours to strengthen its position in the system. In a similar vein, Bergström’s (2005, p. 313) comprehensive treatment of comitology concludes that ‘the role of the European Parliament has been characterized by struggle . . . the tactic of the European Parliament has been to wear its opponents down by use and abuse of all means available until the point has been reached where they realize that it will be less costly to make concessions than to resist’. Likewise, Pollack (2003a, p. 120) speaks of ‘an ongoing political battle’ among the EU institutions when explaining the

6

The EU Comitology System in Theory and Practice

evolution of the comitology system. Bradley (2008, p. 850), a seasoned observer of parliamentary tactics in the comitology area, concludes on the 2006 reform of comitology that ‘Parliament had to use the full panoply of its institutional prerogatives, legislative, budgetary and jurisdictional . . . these are for the most part weapons of mass obstruction . . . yet the fact is that, without recourse to such tactics, it seems unlikely that the 2006 reforms would ever have got off the ground’. But what exactly is at stake? Despite the obvious intensity of the interinstitutional negotiations behind the comitology rules, we really know very little of the driving forces behind the system. What are the interests of the negotiating actors, their strategies and the instruments they have at hand to pursue their interests? Understanding the development of the comitology system over time is the second important question addressed by this book.

The comitology system in practice: Bargaining or deliberation? Most comitology observers agree that the system’s numerous committees are established to control the Commission. Most observers also agree that the different decision procedures used by the committees constrain the Commission to different degrees. The accepted view is that, among the pre-Lisbon procedures, the regulatory procedure is more restrictive than the management procedure, which again is more restrictive than the advisory procedure. This hierarchical ordering has been confirmed by game theoretical analyses of the procedures (Steunenberg et al., 1996, 1997; Ballman et al., 2002). After the Lisbon Treaty the regulatory and management procedures have been merged into a new so-called examination procedure, while the advisory procedure is unchanged. The constraining effects of the two future procedures are obviously very different. However, the constraining effect of the comitology procedures is contingent upon certain, but often implicit, behavioural assumptions. It is an old insight from the institutional literature that outcomes are determined by the interaction of institutional rules and actor orientations. In Ostrom’s words (1986), you need both a model of the situation and a model of the actors to explain outcomes. Or, as Scharpf (1989; see also 1997, pp. 84–9) has put it, policy choices are determined by the combination of decision rules and decision styles. The underlying behavioural assumption in comitology analyses inspired by game theory or rational choice theory is that preferences are fixed and defined exogenously to

Introduction 7

the decision situation. In this view, the national representatives travel to the comitology meetings in Brussels to defend predefined national interests. However, considerable evidence suggests that this working assumption is, at best, only partially valid. The literature provides two images of the committees’ daily workings (Pollack, 2003b). The first image – associated with authors such as Joerges and Neyer (1997a, b), Dehousse (2003) and Wessels (1998) – is drawn from sociological institutionalism and constructivism. It suggests that comitology committees provide a forum in which experts meet and deliberate to find the best or most efficient solutions to common policy problems. According to this image, comitology is a technocratic version of deliberative democracy in which informal norms, deliberation and good arguments matter more than economic interests and formal voting rules. The second image – drawn from rational choice theory and associated with authors such as Steunenberg et al. (1996, 1997), Pollack (2003a, pp. 114–46), Ballman et al. (2002) and Franchino (2000a) – portrays comitology committees as miniature versions of the Council. They provide arenas for tough intergovernmental bargaining where the member states fight over secondary rules to implement EU legislation. The primary evidence to shed light on the two images is provided by case studies of individual committees. But apart from Joerges and Neyers’ (1997a, b) well-known study of the Standing Committee for Foodstuffs, case studies of comitology committees seem to find traces of both images in the daily workings of comitology committees (Bradley, 1998; Daemen and van Schendelen, 1998; Philip, 1998; Töller, 1998; Gehring, 1999) as do survey-based studies (Egeberg et al., 2003; Blom-Hansen and Brandsma, 2009). These findings are puzzling. Can it really be true that comitology committees operate as autonomous free-floating decision-making bodies? Are the national representatives in the comitology committees not instructed by their home government? Do the member states really struggle to establish control mechanisms and then forget about them once they are in place? The third and final important question addressed here is how the numerous comitology committees operate in day-to-day practice.

The book’s argument This book argues that the comitology system is best understood from a delegation perspective. In general, legislators value the ability to delegate power to the executive because this allows them time and

8

The EU Comitology System in Theory and Practice

energy to deal with more pressing business, helps them resolve difficult issues and provides them with information from specialists on complex matters. However, legislators cautiously decide when to delegate and which powers to give away because they face the possibility that the executive will make decisions that can go against their preferences. The comitology system was originally created to facilitate delegation from the member states to the Commission without losing control. The fact that the system is first and foremost an oversight mechanism has implications for all subsequent decisions on the system. The member states, the European Parliament and the Commission care so much about getting the minutest details of the comitology procedures right because they suspect that this may affect policy decisions down the line. This is the logic that drives the design of the system. The need to control the Commission varies across issue areas, and delegation can take place without comitology procedures, with permissive procedures or with strict procedures. But as the book will demonstrate, the pattern is not random. Predictable factors that are well-known from the delegation literature determine the choice of comitology procedures as well as the daily workings of the committee system and the discretion enjoyed by the individual national representatives. When the comitology system appears to be dominated by a deliberative decision style, and when national representatives appear to enjoy considerable discretion, it is not a coincidence, but because the member states have deliberately decided so. The delegation argument on the comitology system is explained in more detail in Chapter 3.

The contributions of this book In addition to the literature on comitology, the book contributes to the literature on EU committee governance, multi-level governance and rational delegation. EU committee governance. The first contribution is to the literature on committee governance in the EU. The starting point for this literature is the fact that committees are involved in all phases of the EU decisionmaking process. In the agenda-setting phase the Commission takes advice from expert committees of member state representatives and other actors. In the decision-making phase the Council’s and Coreper’s decisions are prepared in Council working groups of member states’ representatives. Finally, in the implementation phase the Commission is assisted by comitology committees. These three types of committees

Introduction 9

are not only present in the decision-making process, they are heavily involved. The treaty-based EU decision-makers – the Commission, the Council, the European Parliament – can be considered the tip of the iceberg, while the committees are the larger, submerged part of the iceberg. Since the three types of committees occupy distinct roles in the EU decision-making process, and since their formal competence varies, it is natural to study them in their own right. So just as some studies focus on comitology committees, some studies focus on Commission expert groups (Mahoney, 2004; Broscheid and Coen, 2007; Gornitzka and Sverdrup, 2008) and Council working groups (Beyers, 2005; Fouilleux et al., 2005; Häge, 2007; Naurin, 2009). However, a growing literature makes cross-cutting studies of these committees under the assumption that they are all examples of a broader EU committee system (Egeberg et al., 2003; Quaglia et al., 2008; Rhinard, 2002). Since the volume edited by Christiansen and Kirchner (2000), it has been common to refer to this broader system as ‘EU committee governance’. The study of this system has not least been advanced by a number of volumes where editors have collected analyses of selected aspects of the system (Pedler and Schaefer, 1996; van Schendelen, 1998; Joerges and Vos, 1999; Andenas and Türk, 2000; Christiansen and Larsson, 2007). Since the different types of committees are involved in different phases of the same policy process, they often deal with the same issues. There may also be considerable overlap at the personal level across the committees because they may all want to draw on the same type of expertise from the member states’ national governments. There is also considerable overlap between the political and democratic concerns they give rise to. Can committee members be held accountable? Are they socialized into a special supranational ethos? Are the committees too technocratic? Do they strike the right balance between professional solutions and democratic transparency? Are they bargaining forums or arenas for expert deliberation? Have they got the balance between input and output legitimacy right? These questions are relevant for all types of committees. This book investigates in depth several of them in the comitology area and thus contributes to understanding the broader phenomenon of EU committee governance. Multi-level governance. The book’s second contribution is to the growing body of studies of multi-level governance in the EU. The starting point for this literature is the fact that European integration is a process in which authority and policy-making influence are shared across multiple levels of government. According to this literature, a new type of

10

The EU Comitology System in Theory and Practice

decision-making system is emerging in which various territorial levels form intermeshing networks and require mutual cooperation in order to carry out joint tasks. Multi-level governance was first proposed by Marks (1992, 1996) and Hooghe (1996) as a useful concept to understand policy dynamics in the area of EU cohesion policy, but was soon advocated as a general model of EU policy-making (Marks et al., 1996). Multi-level governance has been studied in a number of EU policy areas. It draws attention to the complex, multi-layered nature of the EU political system and raises a number of questions (Bache, 2004) that are also relevant in the comitology area: How is multi-level governance different from hierarchical control? What is the relationship between formal and informal institutions? What does multi-level governance imply for the power, position and role of the nation state? To what extent does multi-level governance vary across policy areas? The multi-level governance model has been criticized for conceptual ambiguity, for empirical imprecision and for equating multi-level involvement with multi-level influence (Pollack, 1995; Bache, 1999; Peters and Pierre, 2004; Blom-Hansen, 2005; Piattoni, 2009). Refining the theory has therefore been a challenge. This has been taken up by Hooghe and Marks (2003) who, on the basis of a broad range of literature, distil two distinct types of multi-level governance. Type I describes general-purpose, permanent, non-intersecting and territorially nested jurisdictions, for instance a federal system. Type II governance describes task-specific, intersecting, flexible, fluid and semi-permanent arrangements that typically target functionally specific policy problems, for instance inter-regional commissions and intercity agencies. Hooghe and Marks consider both types of multi-level governance radical departures from the centralized state, but models that diffuse authority in contrasting ways. They represent alternative models of coordinating across levels and as such raise troubling questions. ‘How do these types coexist? What are their dynamic properties? How is democracy limited or enabled in each institutional setup?’ (Hooghe and Marks, 2003, p. 241). By focusing on comitology, this book investigates a prominent example of type II multi-level governance. In fact, in their original presentation of the multi-level governance model, Marks et al. (1996, p. 367) pointed to comitology as a clear-cut example of the intermeshing of levels in the EU, but it was a call for further research that was not heeded in the multi-level governance literature. In Chapters 5–7, this book investigates the dynamic properties of the comitology system, and

Introduction 11

in the conclusion it addresses the broader democratic aspects of the system. In this way the book contributes to understanding multi-level governance in the EU. Rational delegation. The third literature to which this book contributes is the rational delegation literature. The proper dividing line between the legislative and executive branch of government – or, between politics and administration – has occupied the minds of social scientists for as long as social science has existed. About 100 years ago Woodrow Wilson (1887) and Max Weber (1970) discussed the dilemma between the need for bureaucratic expertise and the dangers of bureaucratic power. Delegation and legislative control have never left the research agenda (cf. Ogul and Rockman, 1990) and have been revitalized as research questions in the past 10–15 years where they have increasingly been studied through the lenses of rational choice theory. This redirection of the delegation literature began as studies of the US Congress (Kiewiet and McCubbins, 1991; Epstein and O’Halloran, 1996), but it soon moved beyond this empirical setting and is today applied to a wide range of systems, including the EU (Franchino, 2007; Pollack, 2003a). The rational delegation literature will be introduced in more detail in Chapter 3. For the moment I want to make the point that this book not only draws upon this literature to understand the comitology system. By investigating a unique, but widely used, legislative oversight mechanism it also contributes to our understanding of legislative delegation in general. I elaborate this point in more detail in the concluding chapter. The comitology system allows the EU member states to monitor delegated decision-making. At the same time it qualifies delegated decision-making by adding expert knowledge to the system. As a side effect, this latter point may make comitology attractive from a democratic-normative perspective, a point that will also be discussed in the concluding chapter.

Outline of the book The objective of this book is to increase our understanding of the comitology system and add to the comitology literature. Chapter 2 reviews this literature and introduces the nuts and bolts of the comitology system. It pays special attention to the changes brought about by the Lisbon Treaty. Readers who are already familiar with the comitology system may want to skip this chapter. The more specific purpose of the book is to address three prominent, unresolved issues in the comitology literature. Why does the comitology

12

The EU Comitology System in Theory and Practice

system exist? What are the driving forces of the system’s development over time? How does the system work in daily practice? Chapter 3 presents the book’s theoretical perspective on these questions. It argues that they are best understood from a delegation perspective. Chapter 3 also derives a set of hypotheses to be investigated in the empirical chapters. Chapter 4 addresses the origins of the comitology system. It argues that, contrary to conventional wisdom, the exact choice of comitology can only be explained by the actions of the Commission. This conclusion is based on a case study of the first uses of comitology within three EU policy areas: agriculture, development aid and customs policy. The chapter relies on three types of documentary sources – the legal texts in the EU’s Official Journal, the Commission’s monthly Bulletin and the daily news coverage of EU affairs by the Brussels-based news agency, Agence Europe. Over the eight–nine years in the 1960s when comitology was introduced, this agency produced more than 20,000 pages of detailed coverage of the EU’s daily affairs. Chapter 5 is the first of three chapters that investigate the driving forces of the comitology system. The three chapters analyse the two primary instruments to influence the system’s design under a given treaty regime: first, the Council’s so-called comitology decisions, which specify the range of decision procedures that a given comitology committee can be equipped with, and, second, daily EU legislation which decides whether or not to install a comitology committee as a watchdog over delegation of specific rule-making powers to the Commission. Chapter 5 focuses on the first instrument, the Council’s comitology decisions. They specify the procedures which regulate how the committees may refer proposals to the Council, the timing of Council control, the default condition if the Council does not act and the degree of involvement of the European Parliament. In short, the Council’s comitology decisions provide a laundry list of institutional checks that may, or may not, be installed in any given delegation situation. The logic behind the comitology decisions is analysed by a case study of the 2006 comitology reform where the latest comitology procedure before the Lisbon Treaty – the regulatory procedure with scrutiny – was introduced. The case study makes it possible to flesh out in detail what institutional interests and instruments mean in the area of comitology. It demonstrates that the Council’s comitology decisions are the result of a two-dimensional constitutional struggle. The first dimension concerns the relative supervisory position of the two legislative actors, the

Introduction 13

Council and the European Parliament. The second dimension concerns the relationship between the legislative and the executive branch of the EU system. Chapter 6 focuses on the second instrument to influence the design of the comitology system, namely the exact specification of comitology provisions in individual legislative acts. That is, it focuses on daily legislative practice under the framework specified by the Council’s comitology decisions. The aim is to determine the specific institutional comitology preferences of the Commission, the Council and the European Parliament. This is done by analysing all directives and regulations enacted under the 1999 comitology decision, a total of 686 legislative acts. The chapter shows that the handling of comitology issues in daily legislative practice has become routinized over time. Compared with the period before the turn of the millennium there is much more agreement on the choice of comitology procedures among the legislative actors. It would be misleading to interpret this as a sign of convergence of preferences over time. It is more likely due to a growing degree of correct anticipation by the legislative actors of each other’s preferences. The relatively few cases of open disagreement exhibit a systematic pattern of institutional preferences that corroborates Chapter 5’s findings on the Council’s comitology decision. Overall, the two chapters suggest that the inter-institutional battle over comitology is not fought over daily legislative matters, but over the general rules specified by the comitology decisions. Chapter 7 investigates daily legislative practice in further detail. It seeks to explain the EU legislators’ exact choice of comitology procedure within the rules specified by the Council’s comitology decisions. Like Chapter 6, it investigates all directives and regulations enacted under the 1999 comitology decision. It demonstrates that the EU legislators’ choice of when and how to delegate powers to the Commission is determined by factors that are well-known from the rational delegation literature. In areas characterized by high issue complexity and in areas of institutional conflict, the Council and the Parliament install stricter comitology procedures to control the Commission. Chapter 8 addresses the daily workings of the numerous comitology committees. Most scholars agree that the committees are established to control the Commission, but observers of the committees’ daily practice report that the committees do not always act as controllers of their home government’s interests. The chapter offers a more systematic approach to this phenomenon in a population study of all

14

The EU Comitology System in Theory and Practice

comitology committees. The data, collected via a questionnaire to all comitology representatives from the author’s home country, Denmark, make it possible to provide systematic answers to questions such as how member states control their comitology representatives, the extent to which committee members deliberate or bargain and the role of the Commission which chairs and supports all committees. Chapter 9 concludes. It reviews the book’s main empirical findings in light of the existing literature on comitology and delegation. It demonstrates how the findings complement, revisit and sometimes challenge existing studies. The chapter ends by evaluating the normative implications of the findings with regard to the EU’s democratic legitimacy and the changes introduced by the Lisbon Treaty. Does the comitology system contribute to the EU’s so-called democratic deficit? This is an often-heard criticism, but while the system may be lacking in traditional input legitimacy, its ready access to unmatched expertise may mean that it contributes to output legitimacy.

2 A Crackpot Concept

‘If you meet someone who says he knows all about comitology, run a mile’, the magazine European Voice (28 May 1998) warned its readers several years ago. This system is ‘astoundingly arcane’, it explained. Ten years on, the Financial Times (14 July 2009) saw little change. The comitology system is ‘impenetrable even to most Brussels-watchers’, it reported. However, according to The Economist (19 April 2001), comitology has some potential, at least within complex regulatory fields such as European securities regulation. Although sceptical about the system’s democratic credentials this magazine, not normally known as an admirer of Brussels bureaucracy, acknowledged that comitology may be a ‘less crackpot concept than its name implies’. These mixed outside evaluations are not without resonance inside the EU system. Taking over after the disgraceful dismissal of the Santer Commission in 1999 the new Commission President Romano Prodi launched an ambitious institutional reform programme. According to the famous White Paper European Governance (Commission, 2001a), the EU’s main problem was that ‘many people are losing confidence in a poorly understood and complex system’ (ibid., p. 3). Prodi wanted to increase transparency, accountability and decision-making effectiveness. One of his reform targets was the comitology system. ‘It must . . . be clearer who is responsible for policy execution’ (ibid., p. 31), the White Paper declared and went on to question the need for the most complicated procedural aspects of comitology. Their abandonment would ‘make decision-making simpler, faster and easier to understand’ (ibid., p. 31). Five years later, in the summer of 2006, the comitology system was finally changed but the reform was not as envisaged by the Commission. The most complicated procedures were not abolished. On the contrary, a new and even more complicated procedure, the so-called regulatory 15

16

The EU Comitology System in Theory and Practice

procedure with scrutiny, was added to the system’s already labyrinthine rules. The reason was that the European Parliament, long an ardent critic of comitology, seized the moment to gain more influence over executive rule-making, and the new procedure was the means to this end. The Parliament’s rapporteur admitted that ‘comitology is confusing and easily subject to misinterpretation’. But rectifying this problem was, after all, of secondary importance. A policy window had opened and the Parliament suddenly had a chance to gain more influence over comitology. As the Parliament saw it, this was a ‘golden opportunity, not to be wasted’ (European Parliament, 2003a, p. 21). The Parliament got its way because the member states in the Council of Ministers felt it increasingly unjustifiable to keep the Parliament out of delegated rule-making. The growing use of the codecision procedure had turned the Parliament into an almost full-blown co-legislator, but in the area of delegated rulemaking its position was still second to the Council. This state of affairs was increasingly difficult to defend. The Lisbon Treaty finally established the European Parliament as a full co-designer of the comitology system, triggering a new round of inter-institutional infighting over the system. As this account indicates, comitology is a part of constitutional politics in the EU. The system is one of the determinants of the balance of power between the Commission, the Council and the European Parliament, and that is why they take such an active interest in comitology. The system is the result of half a century’s political haggling, and that is why it is so Byzantine. However, the criticisms of comitology’s bewildering complexity are less valid today than they used to be. The system is still shrouded in secrecy in many respects. For instance, it is very difficult to find out who the member states’ official representatives in the many comitology committees are and how they receive instructions from their home government before formal votes are taken. But two things have changed over the last 10–15 years. The first is that the system is now much more open than in the past. Since 2000 the Commission has published an annual report on the comitology system that contains information on the number and types of committees, their meeting frequency and the number of adopted opinions and implementing measures (e.g. Commission, 2009a). In 2003, the Commission established an online comitology register that provides detailed information on agendas for meetings in comitology committees, attendance lists, summary records, vote results and draft measures (http://ec.europa.eu/transparency/regcomitology/ index.cfm). Although these sources may not be perfect in every respect (Brandsma et al., 2008), they still represent a remarkable improvement in transparency.

A Crackpot Concept

17

The second thing that has changed is scholarly attention. After having neglected comitology for 30–40 years since its introduction in the 1960s (but see Bertram, 1967–68; Schindler, 1971), social scientists finally took an interest in the system around the turn of the millennium. Although the body of works on executive rule-making by the comitology method still pales in comparison with studies of ordinary EU legislative rulemaking, the last 10–15 years have witnessed a dramatic increase in studies of the comitology system. White spots remain, but comitology is no longer completely unmapped territory in the EU system. The purpose of this chapter is to review the literature and introduce the nuts and bolts of the comitology system. The following sections describe the history of the system, the changes introduced by the Lisbon Treaty, the system’s institutional set-up and its incidence in daily politics.

The history of comitology Although the comitology system was not part of the original treaties, it is almost as old as the EU. The Rome Treaty included a vague provision on delegation of implementation power to the Commission (article 155), but did not specify how and when the Council should delegate powers, or whether any control mechanisms were to be installed. However, soon after the treaty entered into force the need for delegating powers to the Commission became evident, not least due to the introduction of the Common Agricultural Policy, which required daily administration at the collective level. The member states delegated this task to the Commission but required it to obtain approval from management committees when adopting implementation measures (Demmke et al., 1996; Haibach, 2000; Bergström, 2005, pp. 38–78). The use of committees of member state representatives to ‘assist’ the Commission in delegated decision-making grew over the decades, and implementation committees soon amounted to several hundred (Falke, 1996, 2000). The weak treaty foundation of the comitology system was rectified by the Single European Act in 1987. The EC Treaty (article 202, ex-145) from then on required the Council to delegate powers to the Commission, but also allowed the Council to attach conditions to the delegated powers. However, the new treaty provision required the Council to establish these conditions in advance in the form of framework rules for the use of delegated powers. This was done by the Council’s first comitology decision in 1987 (Council of Ministers, 1987), which basically confirmed the committee procedures that had grown out of practice over the years. The Commission’s delegated powers would thus be subject to review by advisory, management and regulatory

18

The EU Comitology System in Theory and Practice

committees and a safeguard procedure (Vos, 1997). The Council’s decision was a disappointment to the European Parliament whose wish for supervisory control rights had been ignored. The Parliament challenged the comitology decision before the European Court of Justice, but the case was declared inadmissible. However, inter-institutional rivalries continued, not least after the Maastricht Treaty which introduced the codecision procedure. As a co-legislator the Parliament demanded the same rights as the Council in the control of the Commission’s delegated rule-making (Bradley, 1992, 1997). The inter-institutional skirmishes led to a new comitology decision in 1999 (Council of Ministers, 1999a), which simplified, but upheld, the advisory, management, regulatory and safeguard procedures, and allowed the Parliament some control of ultra vires behaviour by the Commission (Haibach, 1999; Bergström, 2005, pp. 209–85). However, the 1999 comitology decision did not put a stop to inter-institutional rivalries, not least due to the European Parliament which still found its weak position in the area of delegated rule-making unsatisfactory. And with the growth of the codecision procedure it was now in a better position to pursue its long-held ambition of more control of delegated decision-making. After protracted negotiations a third comitology decision was made in 2006 (technically an amendment to the 1999 decision), introducing a new version of the regulatory procedure – a ‘regulatory procedure with scrutiny’ – that put the Parliament on the long-sought equal footing with the Council in the control of the Commission’s delegated decision-making (Christiansen and Vaccari, 2006; Schusterschitz and Kotz, 2007). This was the last change before the Lisbon Treaty.

The Lisbon Treaty The Lisbon Treaty, which entered into force in December 2009, has important implications for the comitology system. The treaty repeals the EC Treaty’s article 202, the former legal basis of the comitology system, and introduces a new hierarchy of legal acts. Legislative acts are adopted by the Council and Parliament, while executive acts are adopted by the Commission, or in some instances by the Council, as either delegated acts or implementation acts (articles 289–291 TFEU). The Lisbon Treaty thus represents a new approach to delegated rule-making in the EU (Craig, 2008; Ponzano, 2008; Hoffmann, 2009). Delegated and implementation acts are both subjected to parliamentary control, but through different mechanisms. Delegated acts are Commission non-legislative rules of general application that

A Crackpot Concept

19

supplement or amend non-essential elements of a legislative act. Before the Lisbon Treaty these Commission rules were typically subjected to comitology control by the regulatory procedure with scrutiny. The Lisbon Treaty specifies two future parliamentary control mechanisms here, a right of revocation and a right of opposition, but not a comitology system. In contrast to delegated acts, implementation acts are to be controlled by a comitology system. The Lisbon Treaty explicitly states that the Commission’s exercise of implementing powers is to be controlled by the member states. The treaty also stipulates that framework rules on the operation of the future comitology system must be made in advance. But in contrast to the pre-Lisbon system the framework rules are to be made in regulations enacted under the codecision procedure, not in Council decisions. Following the entry into force of the Lisbon Treaty, the Commission proposed new systems for parliamentary control of the two new types of executive rules: delegated acts and implementation acts. After protracted inter-institutional negotiations the systems were agreed upon in late 2010 and formally introduced in 2011 (Brandsma and Blom-Hansen, 2011). Regarding delegated acts, the new parliamentary control mechanisms introduced by the Lisbon Treaty do not, in principle, need any secondary rules to become operative. However, the EU institutions agreed from the outset that a common understanding on their use in daily legislation would be desirable. Consequently, the Commission (2009b) made a proposal specifying a general framework for the operation of the new control mechanisms. The proposal was negotiated with the Council and the European Parliament, and a common understanding was reached in late 2010 that will function as an inter-institutional agreement on the practical arrangements of delegation of powers under article 290 TFEU (European Parliament, Council and Commission, 2011). The contested points in these negotiations were threefold: the extent of Commission consultation of member state experts in the preparation of new acts; the duration of delegated powers; and the exact time limits for expressing parliamentary opposition to draft acts. The Commission took a minimalist approach to these issues and valued executive autonomy, while the Council and the Parliament took a maximalist approach and valued effective control positions. The final compromise, as stated in the common understanding, was held in broad and relatively abstract terms. Consultation of member state experts must be ‘appropriate’. Powers may be delegated for either an unlimited or a limited period to be

20

The EU Comitology System in Theory and Practice

automatically extended unless the legislators oppose. The time limit for expressing parliamentary opposition to draft acts is set at two months, which can be extended by another two months. In sum, the three institutions’ common understanding leaves much to be decided in subsequent legislative practice. Only broad parameters have been set for the exact operation of the new parliamentary control mechanisms under article 290 TFEU. Regarding implementation acts, the Lisbon Treaty requires that framework rules on the comitology system are made before the system can become operative. This requirement is identical to the pre-Lisbon situation with the important change that the framework rules will no longer take the form of a Council decision adopted under the consultation procedure, but of a regulation adopted under the codecision procedure. The Commission (2010b) presented its proposal for new framework rules in early 2010. If adopted, it would lead to a much simpler comitology system and more executive autonomy. The Commission envisaged only two future comitology procedures. The regulatory procedure with scrutiny would be abolished since acts adopted under this procedure would, in the future, be made as delegated acts controlled by the mechanisms under article 290 TFEU. The advisory procedure would be upheld and upgraded to the general comitology procedure. The management and regulatory procedures would be merged into a new so-called examination procedure modelled on the old management procedure. Furthermore, the Commission made the radical proposal of abolishing the principle of referring cases to the Council when committees deliver unfavourable opinions. Instead, the Commission was to be entitled to submit an amended proposal to the same committee. Finally, the Commission proposed that the common commercial policy be included in the comitology system. This would mean that the EU’s external trade protection measures, including the numerous acts introducing anti-dumping or countervailing duties on third countries trying to export goods to the EU area, would be subjected to comitology control in the future. The Commission’s comitology proposal met considerable opposition in both the Council and the European Parliament. The Council was especially concerned about the weak position of the member states under the proposed examination procedure and the abolishment of the system of referring controversial cases to the Council. The Parliament worried about its right of scrutiny under the 1999 comitology decision, which was not included in the Commission’s proposal.

A Crackpot Concept

21

The negotiations ended with a compromise in late 2010, and the new post-Lisbon comitology system was subsequently introduced by regulation 182/2011 (European Parliament and Council, 2011). The system of referring cases to the Council when committees deliver negative opinions is abolished, as proposed by the Commission. But instead a so-called appeal committee of member state representatives is introduced. In the case of highly controversial meetings the member states may let ministers represent them in this committee. Furthermore, the Commission’s proposed new examination procedure is upheld, but variants are introduced that in essence bring the old regulatory procedure back in. For example, in selected cases of no opinion from the committees the Commission must refer its draft implementing act to the new appeal committee. Finally, the inclusion of the EU’s common commercial policy under the comitology system turned out to be a thorny issue. In the end, a special arrangement was agreed upon for this policy area under the new comitology regulation. In sum, although some important details have been changed, the post-Lisbon comitology system is closely modelled on the pre-Lisbon system. The regulatory and management procedures live on as variants of the new examination procedure, and the advisory procedure is not changed at all. Criteria for choosing procedures in daily legislation remain non-binding, the Commission continues to preside over committee meetings, and the European Parliament’s right of scrutiny is intact. The new system still includes an appellate body, although this function is no longer carried out by the Council, but by the new appeal committee to which member states may decide to appoint ministers. Acts previously decided under the regulatory procedure with scrutiny are taken out of the comitology system and will in the future be enacted as delegated acts. New types of acts are added to the system, namely the numerous trade protection measures adopted under the common commercial policy. In short, the post-Lisbon comitology reform represents an incremental, but hotly contested, change of the system which leaves the comitology system more complicated and labyrinthine than ever. In this sense, the reform resembles the 1987, 1999 and 2006 comitology reforms.

The institutional set-up of comitology The numerous comitology committees operate under rules at four levels: treaty rules, framework rules, rules in individual legislative acts and working practices in the individual committees.

22

The EU Comitology System in Theory and Practice

Table 2.1 Treaty regulation of the comitology system over time The Rome Treaty (1957), article 155: In order to ensure the proper functioning and development of the common market, the Commission shall . . . exercise the powers conferred on it by the Council for the implementation of the rules laid down by the latter. Single European Act (1986), article 10: Article 145 of the EEC Treaty shall be supplemented by the following provision: [the Council shall . . . ] confer on the Commission . . . powers for the implementation of the rules which the Council lays down. The Council may impose certain requirements in respect of the exercise of these powers. . . . The procedures referred to above must be consonant with principles and rules to be laid down in advance by the Council, acting unanimously on a proposal from the Commission and after obtaining the opinion of the European Parliament. Lisbon Treaty (2007), article 2, point 236 (new article 249 C): . . . 2. Where uniform conditions for implementing legally binding Union acts are needed, those acts shall confer implementing powers on the Commission . . . 3. For the purposes of paragraph 2, the European Parliament and the Council, acting by means of regulations in accordance with the ordinary legislative procedure, shall lay down in advance the rules and general principles concerning mechanisms for control by Member States of the Commission’s exercise of implementing powers. c European Union, http://eur-lex.europa.eu/’. Source: Eur-Lex.‘

Level 1: The treaty. The first level is the treaty, which specifies how rules on the operation of the comitology system are made. Over the course of history there have been three different treaty provisions, cf. Table 2.1. The first was the vague delegation provision in the Rome Treaty’s article 155. The Single European Act replaced this with the new article 202 TEC (ex-145). Both provisions privileged the Council as the designer of the comitology system although the Single European Act specified that framework rules on the comitology system must be made by the consultation procedure and thus left a role for all three legislative actors. This state of affairs was changed by the Lisbon Treaty which stipulates that in the future comitology rules will be made by the Council and the Parliament acting together under the codecision procedure (article 291 TFEU). The treaty provisions are important because, as argued by Bergström et al. (2007), they specify the actors’ bargaining powers when negotiating rules on the comitology system. And the historical record shows that the actors strategically use their negotiation resources to try to change the comitology rules according to their preferences. For example, it was

A Crackpot Concept

23

no accident that the member states were privileged in the comitology system until the Single European Act. The comitology system until then simply ‘reflected the superior bargaining power of the Council in negotiating the institutional set-up of the procedures with the Commission and the Parliament’ (Bergström et al., 2007, p. 351). Level 2: Framework rules. The second level of rules consists of framework rules on the comitology system. That is, rules on committee procedures, time limits, the involvement of the Council and the European Parliament. The framework rules constitute a list of committee types that can be used in daily legislation. Until the Lisbon Treaty the framework rules were made as Council decisions – in 1987, 1999 and 2006 (Council of Ministers, 1987, 1999a, 2006b). The Lisbon Treaty now specifies that framework rules must be made as regulations under the codecision procedure. In legal terms the framework rules occupy a position under the treaty, but over ordinary legislation in the sense that they are binding on legislative acts that confer implementing powers (Schaefer and Türk, 2007, p. 184). They specify how the individual committees may refer proposals to the Council, the timing of Council control, the default condition if the Council does not act and the degree of involvement of the European Parliament. The exact nature of the various comitology procedures has changed over time, but until the Lisbon Treaty they were always referred to as the advisory, management, regulatory and safeguard procedures, and they constrained the Commission to varying degrees. As game theoretic analyses confirm, the safeguard procedure was the strictest, the regulatory procedure was stricter than the management procedure, which again was stricter than the advisory procedure (Steunenberg et al., 1996, 1997; Ballman et al., 2002). Following the Lisbon Treaty, a compromise has been reached that the four comitology procedures are simplified into two new procedures: an advisory procedure that is identical to the previous advisory procedure and a new examination procedure. The latter procedure is introduced in variants that replace the old management and regulatory procedures (European Parliament and Council, 2011). The specific details of the different procedures and their development over time are summarized in Table 2.2. Traditionally, the position of the European Parliament in the comitology system has been weak. Under the Council’s 1987 decision it played no role at all. The 1999 decision granted the Parliament some insight and supervisory powers. It was now to receive all relevant information and to exercise some control of ultra vires behaviour by the Commission. If the Parliament found that the Commission’s

24 Table 2.2 The list of comitology procedures over time 1987–1999 (Council’s 1987 decision)

1999–2006 (Council’s 1999 decision)

2006–2011 (Council’s 2006 decision)

2011–? (EP/Council’s 2011 regulation)

Advisory procedure

The Commission must take ‘the utmost’ account of the committee’s opinion, but may enact the measure regardless of that opinion.

No change

No change

No change

Management procedure/ examination procedure

The Commission can enact the measure unless the committee opposes by QMV∗ in which case the measure is referred to the Council and either:

Variants removed. The Commission can enact the measure unless the committee opposes by QMV∗ in which case the measure is referred to the Council. The Commission may then enact the measure, but the Council may oppose by QMV∗ within three months.

No change

Renamed examination procedure (primary variant) and slightly changed. The Commission can enact the measure unless the committee opposes by QMV∗ in which case the measure is either referred to the appeal committee, which may oppose by QMV∗ , or an amended version resubmitted to the committee.

Variants removed. The Commission can enact the measure only if the

New variant introduced: The regulatory procedure with

Renamed examination procedure (secondary variants) and slightly

Variant A: the Commission may enact the measure, but the Council may oppose by QMV∗ Variant B: the Commission must defer enactment until the Council reacts by QMV∗ or fails to react. Regulatory procedure/ examination procedure

The Commission can enact the measure only if the committee supports by

25 QMV∗ . Otherwise the measure is referred to the Council and either: Variant A: the Council may oppose by QMV∗

committee supports by QMV∗ . Otherwise the measure is referred to the Council which may oppose by QMV∗ .

scrutiny. Irrespective of the opinion of the committee, the Commission’s measure is always referred to the Council and the European Parliament. If they object, the measure is blocked.

changed. The Commission can enact the measure only if the committee supports by QMV∗ . Otherwise the measure is either referred to the appeal committee, which may oppose by QMV∗ , or an amended version resubmitted to the committee.

Variants removed. Any committee member may refer the Commission’s measure to the Council which may oppose by QMV∗ within a prescribed period. No reaction may mean that the measure is revoked.

No change

Abolished

Variant B: the Council may oppose by simple majority.

Safeguard procedure

Any committee member may refer the Commission’s measure to the Council which: Variant A: may oppose by QMV∗ within a prescribed period. No reaction allows the Commission to enact the measure Variant B: may oppose by QMV∗ within a prescribed period. No reaction means the measure is revoked.

Note: Until 1987 there were no framework rules on the procedures to be followed by the comitology committees. Instead, this was decided on a case-by-case basis. ∗ Qualified majority voting. Source: Council of Ministers (1987, 1999a, 2006b); European Parliament and Council (2011).

26

The EU Comitology System in Theory and Practice

proposal exceeded its delegated powers, it might adopt a resolution. The Commission had to take this resolution into account, but was nonetheless allowed to continue with the procedure (Lintner and Vaccari, 2007). The 2006 decision was widely seen as a victory for the European Parliament because it gained veto rights over delegated decision-making in the EU for the first time. Under the new regulatory procedure with scrutiny both the Council and the European Parliament could block the Commission’s proposals (Bradley, 2008). In the post-Lisbon system the European Parliament has full supervisory powers over delegated acts under article 290 TFEU and continues to hold a right of scrutiny in relation to acts adopted by comitology procedures. The framework rules specify different types of committee procedures, but provide few guidelines on when to use which type. The 1987 comitology decision provided no guidelines at all, but the 1999 decision stated in article 2 that the management procedure ‘should’ be used within the areas of the common agricultural and fisheries policies and in relation to programmes with substantial budgetary implications. The regulatory procedure ‘should’ be used in relation to measure of general scope within the areas of health or safety of humans, animals or plants. The advisory procedure is to be used ‘in any case in which it is considered to be the most appropriate’. However, as the wording signals, these guidelines were informal. This is even stressed in the fifth consideration of the 1999 decision which explicitly states that they are of a ‘non-binding nature’, a stipulation that the European Court of Justice has sanctioned (Türk, 2007, p. 233). The 2006 comitology decision provided a more binding guideline for the use of the new regulatory procedure with scrutiny which, according to the decision’s recital 7a, must be used for implementing measures ‘of general scope’. The post-Lisbon system introduced more guidelines, although still of a non-binding nature. According to article 2 in the 2011 comitology regulation (European Parliament and Council, 2011), the new examination procedure applies ‘in particular’ to implementation acts of general scope, acts with substantial implications or acts relating to the following policy areas: the common agricultural and fisheries policies, the environment, the health of humans, animals or plants, the common commercial policy and tax policies. For other types of implementation acts the advisory procedure is to be used ‘as a general rule’. In sum, the legislators have in practice considerable freedom to decide the choice of comitology procedure on an ad hoc basis in daily legislation. Within the menu provided by the pre-Lisbon comitology decisions and the postLisbon comitology regulation, the choice of committee type is relatively free (see also Bergström, 2005, pp. 272–4).

A Crackpot Concept

27

Table 2.3 An example of a comitology provision in a directive Directive 2002/49/EC of the European Parliament and of the Council of 25 June 2002 relating to the assessment and management of environmental noise∗ Provision delegating power to the Commission (article 12): ‘The Commission shall adapt Annex 1, point 3, Annex II and Annex III to technical and scientific progress in accordance with the procedure provided for in Article 13(2).’ Provision specifying the comitology procedure (article 13(2)): ‘Where reference is made to this paragraph, Article 5 [the regulatory comitology procedure] and 7 of Decision 1999/468/EC [the Council’s 1999 comitology decision] shall apply.’ ∗

Published in OJ L 189, 18 July 2002, pp. 12–25.

There is broad agreement in the literature that the framework rules on the comitology system are the object of intense inter-institutional rivalry (Franchino, 2000b; Pollack, 2003a, pp. 114–46; Bergström, 2005, pp. 285–308; Bradley, 2008), but there is less agreement on what exactly is at stake at this level. Chapter 5 is my contribution to this debate. Level 3: Daily legislative practice. The comitology framework rules specify a list of committee types. However, the decision on which committee type to install in a given delegation situation is taken on a case-by-case basis in daily legislative practice. This is the third level of rules. Whenever the Council and the European Parliament decide to insert a delegation provision in a legislative act, they may decide to add a comitology provision too. In the example in Table 2.3, the act, in this case a directive, contains an article delegating power to the Commission on the condition that a certain comitology procedure is used. The exact choice of comitology procedure is then specified in another article by a reference to the framework rules, in this case the Council’s 1999 comitology decision. This is standard legislative practice. The literature does not agree on how exactly the comitology framework rules are used in daily legislative practice. The question is why comitology clauses are sometimes inserted into legislative acts and sometimes not, and why comitology clauses are sometimes strict and sometimes permissive. A number of scholars argue that comitology is installed in concrete situations to give the Commission access to expertise in the member states’ national administration (Falke, 1996, 2000; van Schendelen, 1996). Other scholars argue that comitology is installed to control the Commission, and that strict procedures are used when control needs are high (Dogan, 1997, 2000; Franchino, 2000a,

28

The EU Comitology System in Theory and Practice

2007, pp. 273–86; Pollack 2003a, pp. 140–4). Chapters 6 and 7 are my contribution to this debate. Level 4: Daily working practice. The final level of rules is the formal and informal working practices within the individual committees. The Commission chairs all committees and provides secretarial assistance. Each committee adopts its own rules of procedure on how to convene meetings, draw up agendas, send out material to committee members, represent absent committee members, admit third parties to the meetings and so on. In practice, the Commission’s (2001c) standard rules of procedure are followed to a great extent. See Table 2.4 for a condensed version of these rules. The individual committee members are normally policy specialists from the national central administrations. If they are not able to attend, Table 2.4 Standard rules of procedure for comitology committees (excerpts) Article 1: Convening a meeting 1.

A committee meeting is convened-by the Chairman . . .

Article 2: Agenda 1. 2.

The Chairman shall draw up the agenda and submit it to the committee. The agenda shall make a distinction between: (a) proposed measures about which the committee is asked to give an opinion . . . (b) other issues . . .

Article 3: Documentation to be sent to committee members 1.

The Chairman shall send the invitation to the meeting, the agenda and proposed measures about which the committee’s opinion is required and any other working documents to the committee members . . . no later than 14 calendar days before the date of the meeting.

Article 4: Informing the European Parliament 1. 2.

The Commission shall send the agenda . . . to the European Parliament for information, The Commission shall send the overall result of voting . . . to the European Parliament . . .

Article 5: Opinion of the committee 1.

When the advisory procedure leads to a vote, the outcome of the vote shall be decided by a simple majority of the committee members. When the committee’s opinion is required under the management or regulatory procedure, this is determined by means of a majority vote, as provided for in Article 205(2) of the Treaty. . . .

A Crackpot Concept

29

Article 6: Representation and quorum 1.

Each Member State delegation is considered to be one committee member. . . .

Article 7: Working groups 1.

The committee may create working groups . . .

Article 8: Admission of third parties 1.

The Chairman may decide to invite experts to talk on particular matters . . .

Article 9: Written procedure 1.

If necessary and justified, the committee’s opinion can be obtained by a written procedure. . . .

Article 10: Secretarial support The Commission shall provide secretarial support for the committee . . . Article 11: Minutes and summary report of the meeting 1.

The minutes of each meeting shall be drawn up under the auspices of the Chairman. . . .

Article 12: Attendance list 1.

At each meeting, the Chairman shall draw up an attendance list . . .

Article 13: Correspondence 1.

Correspondence relating to the committee shall be addressed to the Commission . . .

Article 14: Transparency 1.

The principles and conditions concerning public access to the committee’s documents shall be the same as those applying to Commission documents . . .

Source: Commission (2001c).

they may be replaced by persons from their permanent representation in Brussels, or arrangements are made so that representatives from other member states can vote on their behalf. The committees meet with highly varying frequency. Some meet almost weekly, some only once or twice a year. The number of participants in a given meeting can be quite high since member states are allowed to bring more than one person. The increasing number of participants, especially after the latest enlargement, appears to have given meetings a more formal character, less room for informal discussions and a greater reliance on English as the daily working language (Schaefer and Türk, 2007; Piedrafita, 2009; Alfé et al., 2009).

30

The EU Comitology System in Theory and Practice

The exact nature of the day-to-day practice of the numerous committees is a matter of debate in the literature (cf. Pollack, 2003b). Some scholars, most famously Joerges and Neyer (1997a, b), argue that comitology committees function as deliberative arenas for policy experts, while other scholars portray them as intergovernmental bargaining sessions where the member states fight over secondary rules to implement EU legislation (Steunenberg et al., 1996, 1997; Franchino, 2000b; Ballman et al., 2002; Pollack, 2003a, pp. 114–46). Chapter 8 is my contribution to this debate.

The incidence of the comitology system As noted in Chapter 1, approximately two-thirds of all rules in the EU are made by the Commission as delegated rule-making, and comitology committees control approximately half of all Commission rules. Against this background it is not surprising that the comitology system is found in all policy areas where the EU holds regulatory authority. The spread of the system across policy areas has been investigated by Dogan (1997, 2000) and Falke (1996, 2000). More up-to-date accounts are provided by the Commission in its annual comitology reports. Some key figures on the pre-Lisbon system from the Commission’s 2008 report are listed in Table 2.5. The table shows that the five comitology procedures were used to different degrees, although many committees operated according to multiple procedures. The regulatory procedure was the most frequently used single procedure. The management procedure was used to a somewhat lesser extent, while the advisory procedure and the regulatory procedure with scrutiny were used rarely. The safeguard procedure has always been a rarely used procedure and was in 2008 no longer used at all. However, there were important variations across policy areas. For instance, the management procedure was frequently used in agriculture, while the regulatory procedure was popular in the environmental area. The last three columns of Table 2.5 report summary measures of the spread of comitology across policy areas. These quantitative measures should be used with caution. First, they disregard the importance of the various committees relative to the total regulatory activity of the Commission within the given area. Second, the exact number of comitology committees is a subject of some controversy since there is no agreed upon definition of a committee. Some committees are divided into permanent subcommittees, and some committees are ‘dead’ in the sense that they meet rarely or never, so providing an exact number is difficult.

Table 2.5

The comitology system according to institutional affiliation in the Commission in 2008 No. of committees according to decision procedure

Enterprise and Industry Employment, Social Affairs and Equal Opportunities Agriculture and Rural Development Transport and Energy Environment Research Information Society and Media Fisheries and Maritime Affairs Internal Market Regional Policy

Several procedures

Total no. of committees

Total no. of meetings

Total no. of adopted implementation measures

Advisory

Management

Regulatory

Regulatory with scrutiny

Safeguard

6

3

15

0

0

10

34

57

307

2

0

2

0

0

1

5

4

20

0

19

5

0

0

4

28

226

439

3

3

9

0

0

21

36

60

63

0 0 0

2 2 4

21 0 0

0 0 0

0 0 0

13 4 5

35 6 9

57 62 28

85 168 79

0

1

0

0

0

3

4

15

39

1

1

7

2

0

3

14

42

16

0

0

0

0

0

1

1

9

0

31

32

Table 2.5 (Continued) No. of committees according to decision procedure

Taxation and Customs Union Education and Culture Health and Consumers Justice, Liberty and Security External Relations Trade Enlargement EuropeAid Humanitarian Aid Statistics Budget Anti-Fraud Office Information Technology Total

Total no. of meetings

Total no. of adopted implementation measures

Advisory

Management

Regulatory

Regulatory with scrutiny

Safeguard

1

3

5

0

0

1

10

114

55

1

1

0

0

0

5

7

14

69

2

0

8

0

0

7

17

123

311

4

1

4

2

0

13

24

37

106

0

2

0

0

0

1

3

4

0

3 0 0 0

3 3 7 0

3 0 0 0

0 0 0 0

0 0 0 0

2 0 1 1

12 3 8 1

24 6 38 6

12 69 110 47

0 1 0

3 0 0

2 1 1

0 0 0

0 0 0

4 0 0

9 2 1

16 5 3

20 6 0

0

1

0

0

0

0

1

3

1

24

59

83

4

0

100

270

953

2,022

Source: Commission (2009a).

Several procedures

Total no. of committees

A Crackpot Concept

33

But with these reservations in mind Table 2.5 indicates that comitology, although used in all policy areas, is most prevalent in a few policy areas. Agriculture is still the area where comitology is used most, but other areas are catching up: environment, enterprise and industry, health and consumers and research. This pattern reflects the widening of the EU’s functions and the extent of delegation of decision-making authority to the Commission.

Conclusion: A standard operating procedure The comitology system has for long led a quiet life in the minds of EU observers. Recent years have seen a growing interest in the system, but in many respects it remains a white spot on the EU map. This state of affairs is quite remarkable because the comitology system has developed into a standard operating procedure within delegated decision-making in the EU. The system involves hundreds of committees and thousands of civil servants in Brussels and in member state capitals. It is used in all areas where the EU holds regulatory authority, and where power is delegated to the Commission. It is a relatively complex system comprising four distinct institutional levels. This is the result of a historical process that is almost as old as the EU. It is time that the comitology system receives as much scholarly attention as the rest of the EU system. This book seeks to contribute to that endeavour.

3 A Delegation Perspective on Comitology

‘Precision, speed, unambiguity, knowledge of the files, continuity, discretion, unity, strict subordination, reduction of friction and of material and personal costs.’ This is what Max Weber (1970, p. 214) saw, about 100 years ago, as the advantages of political delegation of powers to the bureaucracy. Delegation would increase the capacity of the government immensely. ‘The bureaucracy compares with other organizations exactly as does the machine with the non-mechanical modes of production’ (ibid., p. 214). However, while praising the technical potential of delegation of powers to the bureaucracy Weber was also deeply worried about the political influence of the bureaucracy. ‘The power position of a fully developed bureaucracy is always overtowering. The “political master” finds himself in the position of the dilettante who stands opposite the “expert”’ (ibid., p. 232). At around the same time, across the Atlantic, Woodrow Wilson had similar thoughts. He also saw great potential in delegating decisionmaking power to the administrative branch of government. The reason was that legislators need the technical insight of trained administrators. ‘There is scarcely a single duty of government which was once simple which is not now complex’ (Wilson, 1887, p. 200). However, like Weber, he also feared the power of the executive branch: I know that a corps of civil servants prepared by a special schooling and drilled, after appointment, into a perfected organization, with appropriate hierarchy and characteristic discipline, seems to a great many very thoughtful persons to contain elements which might combine to make an offensive official class, – a distinct, semicorporate body with sympathies divorced from those of a progressive, free-spirited people (ibid., p. 216). 34

A Delegation Perspective on Comitology

35

In short, to both Weber and Wilson, the challenge for legislators was to delegate without unduly empowering the executive. In the following century this challenge occupied the minds of social scientists on both sides of the Atlantic, although maybe not to the extent wished by Weber and Wilson (Rockman, 1984; Ogul and Rockman, 1990). However, the last 10–15 years have seen a renewed interest in delegation studies. This has coincided with the rise of rational choice theory, and delegation is increasingly studied as a rational institutional choice made by utility-maximizing legislators. This approach to delegation began with studies of the US Congress (Kiewiet and McCubbins, 1991; Epstein and O’Halloran, 1999), but the literature soon included delegation studies of other legislative–executive settings such as the US states (Huber and Shipan, 2002, pp. 139–71; Volden, 2002), Western parliamentary systems (Huber and Shipan, 2002, pp. 171–210; Strøm et al., 2003) and the EU (Kelemen, 2002; Pollack, 2003a; Franchino, 2004, 2007). This literature points to several reasons why legislators delegate to the executive (for reviews see Huber and Shipan, 2000, 2006; Bendor et al., 2001; Voigt and Salzberger, 2002; Strøm, 2003). First and foremost, legislators need the technical insight of the executive branch to formulate detailed policies, but other considerations may also be important. Delegation may be a tool to reduce the workload of legislators, to shift responsibility for unpopular decisions (‘blame avoidance’) or to protect policies against reversals in the future. Delegation may also be a means to avoid legislative cycling, solve coordination problems or establish credible commitments. In short, delegation may reduce decision-making costs for the legislature. However, the literature also points to several dangers of delegating powers to the executive. First and foremost, delegation means empowerment, and power may be used for unintended purposes. Delegatees may choose to drift from the delegators’ preferred policies, so there is a risk of agency loss. But other dangers may also be relevant. Delegation increases the costs of reversing policies, reduces legislative flexibility and may create legitimacy problems. Delegation also creates problems of coordinating and monitoring the work of the delegatee. Obviously, delegation of powers involves benefits as well as costs for the legislators. Since delegation involves both benefits and costs, calculating the right amount of discretion to grant the executive is not straightforward. However, the amount of agency loss can be minimized by institutions (Lupia, 2003; Huber and Shipan, 2006). Legislators may install ex ante institutions to influence the executive’s incentives and possible actions. Moe

36

The EU Comitology System in Theory and Practice

(1990a, b) argues that legislators may pre-empt future policy reversals by delegating and thus shielding their policy from the influence of a different future legislative coalition. In this way delegation may freeze politics in time. Moe refers to this as the politics of structural choice. McCubbins et al. (1987, 1989) argue that legislatures can design administrative procedures to control the executive. By controlling procedures, legislators can ‘stack the deck’ in favour of particular constituents and thereby channel executive decisions towards the outcome favoured by those who are intended to benefit from the policy. Legislators can also install ex post institutions to monitor the actions of the executive. According to McCubbins and Schwartz (1984), they can choose between ‘police patrols’ and ‘fire alarms’. Control by police patrols is centralized, systematic, regular and costly control, for example, by an audit office, while fire alarm control is decentralized, less systematic and less costly since it relies on third parties. For instance, interest organizations may be given privileged access to the executive rule-making process to facilitate monitoring, or they may simply be relied upon to speak up when their members’ interests are threatened by executive rules or actions. In this sense they may act much like concerned citizens pulling fire alarms to alert fire fighters about dangerous activities. In the following sections I seek answers in the delegation literature to the three questions that this book addresses: Why does the comitology system exist? What are the driving forces of the system’s development? How does the system work in daily practice? Chapter 1 introduced the questions, and this chapter develops a set of hypotheses. The rest of the book investigates the hypotheses empirically.

The origins of the comitology system The puzzle about the origins of comitology is not why the system was installed. Parliamentary control of delegated rule-making by the executive is known in all democratic systems. In the early 1960s, the member states faced an obvious need to delegate implementing powers to the Commission in order to start the common policies on agriculture, the internal market and competition, custom policies, external trade policies and so on. At the same time the member states worried about losing national control, so there was also an obvious need to monitor the Commission. It is not puzzling that a control mechanism of some sort was introduced. There is agreement in the literature that the comitology system was created in the 1960s as a response to the EU member states’

A Delegation Perspective on Comitology

37

dilemma of delegating rule-making powers to the Commission without losing control (Demmke et al., 1996; Dogan, 1997; Vos, 1997; Haibach, 2000; Bergström, 2005, pp. 43–57). Rather, the puzzle is why the member states settled on a comitology system to solve their control problem. The nature of the problem was familiar to them from their national systems, but they still found a solution that none of them knew beforehand. Why was comitology chosen as the solution when the member states were familiar with other solutions from their own national systems? The functional reasons offered in the literature can explain why a control system was installed, but not why the member states made the exact choice of a comitology system. As argued by Pierson (2000), functional reasoning tends to overlook the importance of choice. Comitology represented an institutional solution to the member states’ problem, but why was this solution chosen among the many potential ones? My answer starts with the fact that the instalment of a control mechanism constituted a collective action problem for the member states, and to solve it they delegated agenda-setting powers to the Commission. The comitology system was then proposed by the Commission, and the member states agreed because there was no obvious alternative proposal that could defeat the Commission’s proposal. This explanation represents a new perspective on the origins of the comitology system. Contrary to conventional wisdom, it implies that the exact choice of comitology can only be explained by the actions of the Commission, not the member states. In the following pages this explanation is spelled out in more detail, and testable hypotheses are derived. Comitology was introduced in the early 1960s as part of the Common Agricultural Policy (CAP), which represented a complicated coordination problem. The stakes were huge. It is an example of what Moravcsik (1998) calls a major turning point in EU history. From his investigation of these turning points we know that they are often subjected to hard intergovernmental bargaining because the distributional consequences are clear and the decisions are made in information-rich environments under intense societal pressure. These situations can be modelled as coordination games with clear distributional consequences – ‘battle of the sexes’ games. However, the administration of common policies is analytically distinct from their policy contents and should be understood as a separate negotiation issue. Although administration is seldom neutral, deciding administrative regimes does not lend itself easily to hard intergovernmental bargaining. The implications of different administrative

38

The EU Comitology System in Theory and Practice

solutions are less calculable, and the preferences of domestic groups are typically less intense. Negotiating the administration of the common policy should thus be easier since the distributional consequences of various administrative arrangements are less certain. This situation resembles a pure coordination game known from, for example, traffic rules. Since the negotiations on the contents of common policies and their administration are often linked, the situation can be modelled as a pure coordination game nested inside a ‘battle of the sexes’ game. The game can be illustrated in the two-player version of Figure 3.1. The pay-offs to the actors of coordinating their policies are larger than those of not coordinating, but the choice of cooperative strategy has distributional consequences making actor A prefer policy 1 and actor B policy 2. Given the choice of how to coordinate their policies, the actors must choose an administrative set-up. In the stylized version of the game in Figure 3.1 the choice has no distributional consequences, which leaves the actors without any preference ordering of the administrative alternatives. The game shows that both administrative solutions are stable solutions, but it does not explain why or how the actors would settle on one of them. Predicting outcomes in coordination games is generally difficult because several solutions may serve as equilibrium. However, in pure coordination games the actors are relatively disinterested players. In this situation it is rational to delegate the task of proposing solutions to an agent in order to reduce decision-making costs. When deciding the framework rules on the Common Agricultural Policy, this is exactly what the six member states did. In the Rome Treaty they required the Commission to submit proposals on the implementation of this policy. The Commission was thus entrusted with the role of a formal agendasetter, which may be a powerful role, but it depends on the decision rule in the legislative arena. If the agenda-setter’s proposal is easier to adopt than amend, then it may effectively structure and limit the choices available to the legislators (McKelvey, 1976). In the case of the Common Agricultural Policy, the Rome Treaty specified that proposals required unanimous consent in the Council, which effectively stripped the Commission of formal agenda-setting power. This fact may be taken as an indication of the member states’ reluctance to delegate wide-ranging powers to the Commission. But it did not render the Commission completely powerless because, even if the conditions for formal agenda-setting powers are absent, the agenda-setter may still have causal influence on legislative decisions. As argued by Pollack (2003a, pp. 47–56), agenda-setting may take place informally when an

Player B Policy 1

Policy 2

Administration 1

Administration 2

Administration 1

Administration 2

Administration 1

2–1

0–0

0–0

0–0

Administration 2

0–0

2–1

0–0

0–0

Administration 1

0–0

0–0

1–2

0–0

Administration 2

0–0

0–0

0–0

1–2

Policy 1 Player A Policy 2

Figure 3.1 Modelling the choice of administrative regime for common policies as a pure coordination game nested within a ‘battle of the sexes’ game Note: The pay-offs to player A appear first.

39

40

The EU Comitology System in Theory and Practice

actor can set the agenda for a group of legislators, not through formal powers, but through the ability to define issues and present proposals that can rally consensus. One particular way this may happen is laid out by Garrett and Weingast (1993), who build on Schelling’s (1963) idea of focal points. Schelling argued that coordination problems with multiple solutions can be solved by means of such focal points: It often seems that a cynic could have predicted the outcome on the basis of some ‘obvious’ focus for agreement, some strong suggestion contained in the situation itself, without much regard to the merits of the case, the arguments to be made, or the pressure to be applied during the bargaining. The ‘obvious’ place to compromise frequently seems to win by some kind of default, as though there is simply no rationale for settling anywhere else. (Schelling, 1963, pp. 69–70) According to Schelling (1963, p. 70) focal points have ‘intrinsic magnetism’ and enjoy ‘prominence, uniqueness, simplicity, precedent, or some rationale that makes them qualitatively different from the continuum of possible alternatives’. Building on Schelling’s thoughts, Garrett and Weingast (1993) develop the idea of constructed focal points. They argue that in the absence of a natural or pre-existing focal point, an institution can create one and may even be specifically entrusted with this task. In other words, constructing focal points is one way to exercise informal agenda-setting power. It may seem plausible that constructed focal points can explain why actors settle on one among many potential solutions when they have no particular preferences among the potential outcomes. However, Garrett and Weingast argue that constructed focal points may also have an impact in coordination games with distributional consequences. But so may power. The relative influence of focal points and power is likely to vary with the structure of the strategic setting. The smaller the distributional asymmetries among potential cooperative equilibria, the smaller the differences in actors’ power resources; and the less certain the consequences of agreements, the more important focal points will be. Deciding the administrative set-up of common policies is an example of a coordination game with potential, but also uncertain, distributional consequences. It is plausible that focal points may help solve the game. Based on the theory of constructed focal points, three hypotheses on the origins of comitology can be made. We already know that comitology was first used within the Common Agricultural Policy in

A Delegation Perspective on Comitology

41

the early 1960s. What we do not know is why. The starting point then is the difficult negotiations on the creation of the Common Agricultural Policy. Since administration is seldom neutral, the member states were likely to favour supranational administration to different degrees. The first and basic hypothesis consequently states: Hypothesis 1: The member states had divergent preferences about the administration of the Common Agricultural Policy. Second, compared with the substantial policy issues at stake (target prices, import levies, export subsidies, etc.) the distributional consequences of the administration of the Common Agricultural Policy were low. Although the member states were not completely disinterested players, their preferences were not intense. Consequently, negotiating the administrative set-up had sufficient resemblance to a pure coordination problem to be solvable by way of a carefully constructed focal point. Since this task was deliberately delegated to the Commission, the second hypothesis is: Hypothesis 2: The Commission’s proposal for an administrative set-up for the Common Agricultural Policy represented a solution on which intergovernmental bargaining converged. While hypothesis 1 is not necessarily contrary to functionalist explanations of comitology, hypothesis 2 is where functionalist explanations and the focal point explanation diverge. Functionalist explanations are silent on how the decision-makers arrived at their institutional solution. In contrast, the focal point explanation is able to explain the choice of one among many potential solutions. Third, if the Commission’s proposal functioned as a focal point, this will also help explain why this administrative model had staying power. A focal point is one among many potential equilibria. When one is chosen, it is a stable solution, and the original choice of administrative model for the Common Agricultural Policy was therefore difficult to change. However, the idea of focal points can help explain not only the stability of comitology within agriculture, but also its spread to other policy areas. The strength of a focal point can be increased by its sheer use. As noted by Schelling (1963, pp. 67–8): ‘Precedent seems to exercise an influence that greatly exceeds its logical importance . . . often it seems that there is simply no heart left in the bargaining when it takes place

42

The EU Comitology System in Theory and Practice

under the shadow of some dramatic and conspicuous precedent.’ The third hypothesis is an extension of hypothesis 2: Hypothesis 3: The choice of administrative model for the Common Agricultural Policy represented a solution on which intergovernmental bargaining converged in other policy areas where supranational administration and national control had to be reconciled.

The driving forces of the comitology system There is broad agreement in the literature that the Council, the Parliament and the Commission take a keen interest in the comitology system, and that it is the object of intense inter-institutional rivalry (Franchino, 2000a; Pollack, 2003a, pp. 114–46; Bergström, 2005, pp. 285–308; Bergström et al., 2007; Bradley, 2008). But what exactly is at stake? My argument is that the design of the comitology system is a game of control positions. No one in the EU system knows for certain how powers delegated to the Commission are going to be used, or what the future precisely entails, but everybody knows that there will be policy decisions to make down the line. When specifying the details of the various comitology procedures, EU legislators cannot specify the contents of these decisions, but they can make sure that they will be in a position to influence them. They know that the structures they choose may well influence the content, direction and effectiveness of future policies. In this sense, administrative procedures are political weapons. They are structural means to pursue political interests. The political choice of administrative procedures is referred to as the politics of structural choice (Moe, 1990a, b) or deck stacking (McCubbins et al., 1987, 1989). It is a messy affair, and instead of technically rational procedures, it results in complex, restrictive, even bizarre, administrative arrangements. The reason is threefold. First, actors cannot afford to prioritize efficiency. Since the future is uncertain, they need to secure control positions in future decision-making situations. Administrative efficiency comes second, if at all. Second, decision-making in a democratic system often involves compromising, which means that the losing side is granted concessions in the design of administrative procedures. Since it is opposed to the policy decisions of the winning side, the losing side does not want efficient administrative procedures, but procedures that make it difficult to implement the winning side’s

A Delegation Perspective on Comitology

43

policies. Third, some degree of autonomy at the level of the bureaucracy is unavoidable because it is impossible to completely foresee all future contingencies within a policy field in legislative acts. Since the bureaucracy may use its autonomy contrary to the interests of the legislative coalition, this coalition cares not only about administrative efficiency, but also about overseeing the administration. In sum, administrative procedures are designed by actors who may care about administrative efficiency, but who primarily care about controlling future policy decisions, who want to subvert administrative performance and who also want to monitor the daily workings of the administration. When designing the comitology procedures, EU legislators specify the degree of control they will have over the use of delegated powers in the future. They do not know what the future will bring, but the comitology rules can be specified to ensure that future decisions will be enacted through procedures which maximize their control over those decisions. The deck may be stacked in favour of some actors and against others. Consequently, the individual EU legislators will press for comitology rules that ensure efficient institutional control positions for themselves and inefficient control positions for their opponents. If the making of comitology rules is an example of the politics of structural choice or deck stacking, the individual EU legislators seek to use these rules to increase their own control over delegated decisionmaking. This means that the legislators will have different preferences over the comitology rules, typically of the following nature. First, the member states in the Council acknowledge the need to delegate decision-making power to the Commission, but they are wary of the Commission and want to be in a position to monitor it and intervene in individual cases. The comitology system is a convenient control instrument since it is staffed by its own stand-ins. In this sense the member states can afford to delegate and to adopt a broad definition of ‘executive decision-making’. Second, the European Parliament is in a very different position. Since the distinction between legislative and executive decision-making is blurred, the European Parliament may fear that the growth of the cumbersome codecision procedure may induce the Council to transfer decisions to the comitology system because the European Parliament is conveniently absent from this system. The comitology system is thus a threat to the European Parliament’s position in the EU institutional system. From the Parliament’s perspective, comitology interferes with its fundamental right to exercise political supervision over the Commission

44

The EU Comitology System in Theory and Practice

and to participate in the legislative process. If seeking to protect its control, the Parliament’s interest is to leave as much real decision-making as possible in the legislative arena, to keep executive decision-making to purely technical matters and to achieve real supervising powers over the Commission by either gaining access to the comitology system or by replacing the comitology system with other monitoring devices. Third, the Commission is also likely to be sceptical about the comitology system, but for different reasons. It survives by being above national interests and acting like a neutral arbiter and entrepreneur on behalf of all member states. In this sense interference from the Council and the member states in its daily affairs is dangerous. Furthermore, it does not want to be bogged down by nightmarish administrative procedures that drag decisions out for years. The Commission’s interest is to protect its autonomy, and in this respect the comitology system represents a threat. However, to some extent the Commission is likely to have mixed feelings about the comitology system. Aware of the member states’ reluctance to delegate decision-making powers, it may be willing to accept comitology procedures in order to increase delegation. These behavioural logics lead to the following set of hypotheses on the EU legislators’ comitology preferences: Hypothesis 4: The Council favours strict comitology procedures. Hypothesis 5: The Commission favours permissive or no comitology procedures. Hypothesis 6: The European Parliament favours: a. narrowing delegation to technical matters b. permissive or no comitology procedures, or alternatively c. access to the comitology system. These hypotheses are likely to hold as general statements, but they paint a broad picture. They do not take into account that the incentive to delegate and to control delegation varies across policy areas. To understand the variation in the use of comitology committees I turn to the part of the rational delegation literature that for a number of years has studied how legislators calculate the right amount of discretion to grant the executive. This literature has led to many insights, but two appear especially robust. The first is that the legislature delegates less discretion when there is conflict either within the legislature or between the legislature

A Delegation Perspective on Comitology

45

and the executive. As to conflicting preferences within the legislature, many scholars, for instance Volden (2002, p. 191) and Strøm (2003, p. 58), point to collective action problems as the root of delegation. Legislators may delegate to avoid endless cycling of policy proposals, to overcome stalemate or to solve coordination problems. For example, McCubbins and Page (1987, p. 418) argue that conflicts among legislators make it difficult to specify the tasks to be delegated to the executive because the exclusion of controversial issues can lead to the breakdown of the legislative coalition. The mandate to the executive is therefore likely to be rather broad when the legislature is internally divided. But increased delegation due to conflict within the legislature creates an incentive for procedural oversight. As noted by McCubbins and Page (1987, p. 118): ‘With increased conflict there are increased political risks and therewith increased incentive to direct the choice of regulation by the agency, through the imposition of extensive procedural requirements, away from potentially costly alternatives.’ In short, a positive relationship between conflict within the legislature and the stringency of procedural controls is to be expected. As to conflicting preferences between the legislature and the executive, many scholars, for instance Epstein and O’Halloran (1994, 1996, 1999, pp. 77–9) and Franchino (2007, p. 56), argue that the legislature delegates less discretion if its policy goals and preferences diverge from those of the executive branch. This is known as the ally principle (Bendor et al., 2001, p. 236; Huber and Shipan, 2006, p. 260). When legislators have reason to suspect that the executive will make policy decisions they dislike, they will not delegate broad discretion. Everything else being equal, seen from the perspective of the legislators, preference divergence, or conflict, with the executive branch increases the risk of agency drift, and hence reduces the value of unlimited delegation. One way to limit delegation is to tighten procedural controls. As noted by Epstein and O’Halloran (1996, p. 274), ‘as bureaucrats’ preferences diverge from those of legislators, . . . Congress will rationally place tighter constraints on the use of delegated authority through restrictive administrative procedures’. In short, a positive relationship between legislative–executive conflict and the stringency of procedural controls is to be expected. The second insight from the rational choice delegation literature is that the legislature delegates more authority to the executive in complex issue areas. As discussed by Bawn (1995), legislators often face a trade-off between control and expertise, and this is likely to affect

46

The EU Comitology System in Theory and Practice

the authority granted to the executive. Increasing executive authority makes it possible for the executive to incorporate knowledge of policy consequences into its decisions. Seen from the legislators’ perspective, growing complexity means that the gains from the executive’s expertise will gradually outweigh the losses from potential agency drift. In other words, the optimal level of executive authority increases with growing issue complexity. However, when executive authority grows, the risk of agency loss increases. In this situation McCubbins et al. (1989) argue that legislators can constrain the agency’s policies through procedural controls. This will allow them to capture the benefits of agency flexibility while minimizing the risk of non-compliance. Similarly, McCubbins and Page (1987, p. 417) find that complexity forces Congress to delegate regulatory authority, but that ‘with increased uncertainty congressmen prefer to prescribe an increasingly extensive array of regulatory procedures for agency decision making’. In short, a positive relationship between complexity and the stringency of procedural controls is to be expected. In sum, it is a common expectation of the delegation literature that procedural control of delegated powers increases with the degree of conflicting preferences either within the legislature or between the legislature and the executive, and with the complexity of the issue area. If the comitology committees function as procedural control mechanisms of delegated power in the EU system, the delegation literature would expect the following hypotheses to be important explanations of the variation of comitology control across policy areas: Hypothesis 7: Growing conflict within the Council and the European Parliament or between the Council, the European Parliament and the Commission leads to stricter comitology procedures. Hypothesis 8: Growing issue complexity leads to stricter comitology procedures.

The comitology system in daily practice There is agreement in the literature that comitology committees are installed by the EU legislators to control the Commission. There is less agreement on how the committees function in day-to-day practice. Observers of the day-to-day operations in various comitology committees report that the national representatives do not act as controllers of their home government’s interests, but rather as policy experts seeking professionally sound solutions to common policy problems. They

A Delegation Perspective on Comitology

47

appear unconstrained by their home government, or at least operating within wide margins of discretion to solve problems. Meetings in the comitology committees seem to be consensual deliberations between well-intentioned experts, not bargaining sessions between guardians of national interests (Joerges and Neyer, 1997a, b; Wessels, 1998; Dehousse, 2003; Savino, 2009). These findings represent a paradox. Why would the member states struggle to establish control mechanisms and then forget about them once they are in place? My argument is that in order to begin to understand this problem we need to shift our focus from Brussels to national capitals. The national representatives in the various comitology committees are not parliamentarians or government ministers, but civil servants from ministerial departments and agencies. To understand how they behave in Brussels, we need to understand their instructions. How are these civil servants controlled by their ministers? Do they receive instructions before votes are taken in the comitology committees? To what extent is their work in the committees monitored by their home government? In other words, do the seemingly autonomous actions by the national representatives indicate out-of-control bureaucrats or obedient servants faithfully operating within well-understood limits of their mandate? The national government faces a delegation problem. For reasons of a lack of time and insight the ministers cannot sit on the many comitology committees themselves. They need to delegate this task. But how much autonomy should they allow their representatives in the hundreds of comitology committees? To answer this question I again turn to the rational delegation literature which, as explained above, argues that the exact degree of control of delegated powers depends on two key factors. First, the more complex the issue area to be regulated, the greater the need for expert insight to decide regulatory measures. In this situation principals need to allow agents more discretion. However, when bureaucratic discretion grows, the risk of agency drift increases, creating an incentive to offset increased discretion with tighter procedural controls. Second, institutional conflict is likely to lead to stricter procedural control of the executive. According to the ally principle, if there is conflict between the preferences of the principal and the agent, this is likely to make the principal distrustful of the agent and create an incentive to impose more control on the agent. In situations of conflict, delegators tend to delegate less discretion and impose tighter procedural controls on the delegatees.

48

The EU Comitology System in Theory and Practice

In sum, based on the delegation literature the following hypotheses on the national control of comitology representatives are expected to hold: Hypothesis 9: Growing issue complexity leads to increased discretion for the national representatives in the comitology committees. Hypothesis 10: Growing conflict between the national government and the national representatives leads to reduced discretion for the representatives in the comitology committees. These hypotheses imply that the autonomy of comitology representatives is carefully calculated by national politicians according to predictable factors. Some representatives do indeed enjoy considerable discretion – not by coincidence, but because their political masters have deliberately decided so. Having understood the representatives’ instructions, we can shift our focus back to Brussels and study the meetings of the comitology committees. My argument is that representatives armed with clear instructions are not likely to engage in deliberation but rather to defend their negotiation mandate by bargaining. In contrast, representatives who are allowed discretion can afford to deliberate. In other words, it is likely that there is a link between the discretion of the representatives and their behaviour in the committees. This leads to the following hypothesis: Hypothesis 11: Reducing/increasing the national representatives’ discretion leads to more/less bargaining behaviour and less/more deliberative behaviour by the representatives in the comitology committee. The behaviour of the individual national representatives determines the interaction style in the committees. If committees are dominated by deliberating or bargaining representatives, the interaction style will be, respectively, deliberative or bargaining. Committees with representatives exhibiting different types of behaviour are likely to have less clear-cut interaction styles. In other words, the interaction styles in the committees do not develop by accident, but by design. They are the aggregate result of the representatives’ individual instructions. Formulated as a hypothesis this argument is: Hypothesis 12: Committees dominated by deliberating or bargaining representatives are characterized by, respectively, a deliberative or bargaining interaction style.

A Delegation Perspective on Comitology

49

Understanding interaction styles in the comitology committees helps us understand the effect of the comitology system. Does it work as a control mechanism? Investigating the constraining effect of formal comitology procedures is not sufficient to answer this question. It is an old insight from the institutional literature that outcomes are determined by the combined effect of institutional rules and actor orientations (Ostrom, 1986). If actors are not necessarily self-interested bargainers, they may transcend the constraints of institutional rules. They may obtain socially optimal outcomes although caught in prisoner’s dilemmas or other conflicts between individual and collective interests. As argued by Scharpf (1989, p. 168; see also 1997, pp. 84–9) in a discussion of confrontational, bargaining and problem-solving styles of decision-making, ‘if solidaristic goals and common cognitive orientations can be generated and maintained among participants, decision rules, and institutional arrangement generally, have much less of an influence on policy choices’. Scharpf (1989, p. 167) refers to this phenomenon as ‘the power of common orientations’. Here, I want to focus on a more immediate impact of actor orientations than policy outcomes, namely the constraining effect of comitology on the Commission. The comitology procedures are installed to control the Commission but, if the institutional insight is correct, actor orientations also matter. In committees characterized by a deliberative interaction style we should expect the Commission to have more manoeuvrability, irrespective of formal comitology procedure. Hence the final hypothesis on the daily operation of the comitology system is this: Hypothesis 13: The more comitology committees characterized by a deliberative interaction style, the more active the Commission. In sum, to understand the daily operations of the comitology system and the puzzling phenomenon of committee autonomy, we need to shift from the supranational to the national level and start by inspecting the instructions of the national representatives. Only if they are allowed discretion is it reasonable to expect autonomy and deliberation at the supranational level. But the supranational interaction style is important, because it may alleviate the constraining effect of strict comitology procedures. The overall rationale of the comitology system – to control the Commission – may thus to some extent be neutralized by the behaviour of the national representatives. But if that is the case, it is because the member states have decided so. The argument is summarized in Figure 3.2.

50

The EU Comitology System in Theory and Practice

Issue

Discretion of

Behaviour of

Interaction

The

complexity

the national

the national

style in the

Commission’s

and

comitology

comitology

comitology

role in the

institutional

representative

representative

committee

comitology

conflict

Figure 3.2

committee

Understanding the daily workings of the comitology system

Conclusion This chapter has applied a delegation perspective on the comitology system and used this literature to derive empirical expectations on the three questions addressed in the book: Why does the comitology system exist? What are the driving forces of the system’s development? How does the system work in daily practice? The overall argument is that these questions represent delegation challenges for the EU member states. They seek to reap the gains of cooperation and delegation, which involves finding a balance between delegating powers to the Commission and not losing control over EU policies. The comitology system is the instrument. The hypotheses derived from the delegation literature will be subjected to empirical analysis in the following chapters. Table 3.1 outlines the empirical analyses and shows where in the book they are presented and the type of data used to investigate the hypotheses. The data will be presented in detail in the relevant chapters.

Table 3.1 Overview of hypotheses, data and chapters Hypotheses

Methods and data

Chapter

The origins of comitology: Hypothesis 1: The member states had divergent preferences about the administration of the Common Agricultural Policy. Hypothesis 2: The Commission’s proposal for an administrative set-up for the Common Agricultural Policy represented a solution on which intergovernmental bargaining converged.

Case study of the first uses of comitology within agriculture, development aid and customs policy in the 1960s

4

51 Hypothesis 3: The choice of administrative model for the Common Agricultural Policy represented a solution on which intergovernmental bargaining converged in other policy areas where supranational administration and national control had to be reconciled. The driving forces of comitology: Hypothesis 4: The Council favours strict comitology procedures. Hypothesis 5: The Commission favours permissive or no comitology procedures. Hypothesis 6: The European Parliament favours: a. b.

c.

narrowing delegation to technical matters permissive or no comitology procedures, or alternatively access to the comitology system.

Case study of the 2006 comitology reform (introduction of the regulatory procedure with scrutiny)

5–7

Quantitative analysis of all directives and regulations enacted under the 1999 comitology decision (686 legislative acts)

Hypothesis 7: Growing conflict within the Council and the European Parliament or between the Council, the European Parliament and the Commission leads to stricter comitology procedures. Hypothesis 8: Growing issue complexity leads to stricter comitology procedures. Comitology in daily practice: Hypothesis 9: Growing issue complexity leads to increased discretion for the national representatives in the comitology committees.

Quantitative analysis of survey data from questionnaire to all Danish comitology representatives

8

52

The EU Comitology System in Theory and Practice

Table 3.1 (Continued) Hypotheses

Methods and data

Chapter

Hypothesis 10: Growing conflict between the national government and the national representatives leads to reduced discretion for the representatives in the comitology committees. Hypothesis 11: Reducing/ increasing the national representatives’ discretion leads to more/less bargaining behaviour and less/more deliberative behaviour by the representatives in the comitology committee. Hypothesis 12: Committees dominated by deliberating or bargaining representatives are characterized by, respectively, a deliberative or bargaining interaction style. Hypothesis 13: The more comitology committees characterized by a deliberative interaction style, the more active the Commission.

The questions addressed in this book have troubled observers and analysts of the comitology system for many years, and the delegation perspective is not the only approach they have used. The questions have been analysed from many other vantage points over the years and the main contenders have already been introduced: functionalist reasoning to understand the origins of comitology, and deliberative politics to understand the daily workings of the numerous comitology committees. To gauge their relative explanatory power, they will be described in more detail and confronted with the delegation perspective in the empirical chapters. In the concluding chapter, I make a final evaluation of the theoretical lessons of my analyses.

4 The Origins of Comitology

The establishment of the Common Agricultural Policy (CAP) in the early 1960s required extensive and detailed technical regulation which the Council of Ministers could not carry out alone. It simply lacked the time and resources to respond to the needs of day-to-day management. To get an impression of the administrative challenge, it is worth quoting at length from the Commission’s own announcement of its new ‘considerable managerial responsibilities’ on 30 July 1962, the day the Common Agricultural Policy came into force: Customs duties, quantitative restrictions and minimum prices on cereal, pork, eggs and poultry prices, which till now have operated at the member states frontiers, have been replaced by levies fixed by Community rules . . . The application of the cereal regulation will give the EEC Commission the most work; each day [Commissioner for Agriculture] M. Mansholt’s departments will have to register at least 2,000 facts on changes in the world prices. In a few hours they will have to assimilate this information which the Commission will use as a base in establishing the day’s caf price for the various types of cereal . . . The Commission must also fix daily the caf price which is the basis for the amount of the refund granted to the Community exporters to enable them to sell their products on the world market. In addition the Commission must fix daily the levy to be paid by Community dealers on certain cereal imports . . . M. Mansholt’s departments will have to deal with some 2,000 facts per day, and 6,000 once a week. (Agence Europe, 30 July 1962) It was evident to everyone that the Council needed to delegate the daily administration of the Common Agricultural Policy. However, the member states were reluctant to grant the Commission unlimited 53

54

The EU Comitology System in Theory and Practice

implementation power because they feared that they would lose control over an important and sensitive policy area. There is widespread agreement in the literature that this dilemma is the background for the comitology system (Bertram, 1967–68; Schindler, 1971; Demmke et al., 1996; Vos, 1997; Bergström, 2005, pp. 43–57). The puzzle is not that some sort of control system was created to handle this dilemma. Parliamentary control of delegated rule-making by the executive is, after all, a well-known problem in all political systems. All member states had found solutions in their own national systems, but comitology was not among them. This system is unique to the EU. How and why did comitology start? The dominant functional reasoning in the literature overlooks the importance of choice. Why was comitology chosen as the solution to the member states’ dilemma among many potential solutions? This question will be investigated in this chapter. In Chapter 3, I argued that when deciding the Common Agricultural Policy in the early 1960s, the member states faced a collective action problem and solved it by delegating agenda-setting functions to the Commission. This argument implies that, contrary to conventional wisdom, the choice of comitology can be explained by the actions of the Commission, not the member states. The logic of the argument is that negotiating the administrative set-up of the Common Agricultural Policy had sufficient resemblance to a pure coordination problem to be solvable by a carefully constructed focal point. Such a point is one among many potential solutions to a coordination problem. When one solution is chosen, it is stable, so this would explain why the comitology system was installed in other policy areas later in the 1960s when the need arose to reconcile delegation and control. Based on this argument three hypotheses were developed: Hypothesis 1: The member states had divergent preferences about the administration of the Common Agricultural Policy. Hypothesis 2: The Commission’s proposal for an administrative set-up for the Common Agricultural Policy represented a solution on which intergovernmental bargaining converged. Hypothesis 3: The choice of administrative model for the Common Agricultural Policy represented a solution on which intergovernmental bargaining converged in other policy areas where supranational administration and national control had to be reconciled. This chapter subjects the three hypotheses to an empirical investigation. I start with a brief introduction to the birth of the Common Agricultural

The Origins of Comitology 55

Policy. This section underscores the premise of my argument, namely that the distributional consequences of the administration of the Common Agricultural Policy were small compared with the huge stakes involved in this policy’s substantial contents. I then move on to a presentation of the data and methods to investigate the three hypotheses. The remainder of the chapter is devoted to empirical analysis of the creation of comitology in the agricultural area and of the spread of comitology to other areas. Two areas, development aid to Africa and the common customs union, are chosen for closer inspection. The chapter concludes with a discussion of the relative merits of my explanation based on the delegation literature and the dominant functional explanation found in the comitology literature.

Coordinating agricultural policies in the early 1960s: Huge stakes The adoption of the Common Agricultural Policy in 1962 was a landmark decision in the history of the EU. Including agriculture in the common market had been decided from the outset, and it was given special treatment in the Rome Treaty. Harmonizing agricultural policies was seen as indispensable for the economic integration process. But at the same time agriculture constituted a special problem. The other major integration steps contained in the Rome Treaty – establishing a customs union, freedom of movement for factors of production and free competition among member states – represented measures of negative integration, that is, they removed obstacles to free and undistorted competition (cf. Scharpf, 1999, pp. 43–83). In the post-treaty phase, integration could be extended in these areas without much political attention through Commission interventions against infringements of treaty obligations, and through decisions by the European Court of Justice. In contrast, establishing a Common Agricultural Policy represented positive integration, that is, the reconstruction of a regulatory system at the supranational level. This requires agreement by national governments in the Council, and in contrast to negative integration it is subjected to the difficulties of consensual intergovernmental decision-making in the post-treaty phase. In addition to this institutional challenge, agreement on a Common Agricultural Policy was made difficult by a number of structural and political factors at the national level. First, the importance of agriculture to the national economy varied considerably among the six member states in terms of share of exports and imports, contribution to GDP,

56

The EU Comitology System in Theory and Practice

share of labour force working in agriculture, agricultural price levels and farmer income levels. Germany, Belgium and Luxembourg were large net importers of agricultural products. The Netherlands was the only significant net exporter, but France was the chief agricultural producer among the member states, accounting for almost 40 per cent of all EU production. Italy, however, had the largest proportion of the population in agriculture, but the lowest agricultural productivity and income level in the EU. Due to these differences, agricultural interests differed widely among the six member states (PEP, 1958, pp. 4–9; 1959, pp. 3–18). Second, the member states all had well-established national agricultural support schemes before the Rome Treaty, including structural reform support aiming at amalgamating holdings, land development programmes, educational and technical advice systems and credit and investment assistance. Furthermore, the member states all had price intervention schemes in the form of fixed, minimum or target prices, state purchases to stabilize prices, export subsidies and regulation of imports via quotas and tariffs. The actual use of the various schemes, support levels and administrative arrangements varied widely among the member states, however. For instance, France had a range of single marketing agencies that bought agricultural products at governmentfixed prices and administered price supports and state purchases. Belgium left support schemes to a single Agricultural Fund whose task was to administer credits and subsidies in support of a system of target prices. In Germany, a set of import and stabilization boards regulated market supply within agriculture (PEP, 1958, pp. 9–13; 1959, pp. 3–18). Merging these separate national systems of agricultural support into a single supranational policy was no small challenge. Finally, agriculture represented one of the best organized economic sectors in all the member states. Agricultural interest groups had achieved stable relationships with government agencies and welldeveloped lines of access to political power in all six countries. Government representatives negotiating the supranational policy were thus subject to intense pressure from producer groups intent on protecting their national prerogatives (PEP, 1961, pp. 159–65; Lindberg, 1963, pp. 225–32, 261–6). In sum, the stakes were huge. Despite agreement on the benefits of coordinating policies, the distributional consequences of different solutions were enormous. It is no surprise that finding the solution involved tough intergovernmental negotiations. The story of the difficult birth of the Common Agricultural Policy has been told elsewhere;1 here it suffices to note that it entailed a common organization of

The Origins of Comitology 57

the agricultural markets of the member states based on the following concepts (introduced to varying degrees in different product areas): • A target price system on the Community market • Community purchases to secure the target prices • Levies on imports to the Community area to bridge the gap between the EEC and world market prices • Export subsidies to Community producers exporting to the world market. The organization of the daily administration of this policy regime was not unimportant. Administration is seldom completely neutral, but compared with the substantial issues at stake, the distributional consequences of alternative administrative arrangements were minor. The decision on administration approximated a pure coordination problem that could be solved by a focal point. The remainder of the chapter investigates whether the origins of comitology can be explained in this way, but first a look at the methods and data.2

Methods and data To study institutional origins is to conduct an investigation of events that transpired in a particular period and setting. The analysis is necessarily more ideographic than nomothetic. However, the following analysis is carefully designed to be of more general value. First, the analysis is a narrative, but an analytic one (Bates et al., 1998). It pays close attention to the specifics of time, sequence and context, but is based on explicit lines of deductive reasoning. I seek to capture the essence of the origins of comitology by modelling the situation as a coordination game solved by a focal point. If confirmed, this means that while the specific game played in the early 1960s is not portable, certain elements of it may be. I return to this question in the conclusion. Second, the analysis is based on a careful selection of cases. When studying the spread of comitology, I move outside the realms of agriculture. If the original choice of comitology did indeed function as a focal point for later choices, this is more likely to be the case for the various product areas that were to be covered by the Common Agricultural Policy during the 1960s – cereals, pork, beef, dairy products, poultry, eggs, vegetables, oils, wines and so on – and less likely to be the case for entirely different areas with different actors. In order to conduct a strong test of

58

The EU Comitology System in Theory and Practice

hypothesis 3, I choose the latter option and focus on development aid to Africa and the common customs union. The analysis seeks to uncover the sequence of events and the motivations and arguments of the actors. It is based on documentary data occupying a middle ground between first and secondary sources. I rely on three types of official documentary sources – the legal texts in the EU’s Official Journal, the Commission’s monthly Bulletin3 and the daily news coverage of EU affairs by the Brussels-based news agency Agence Europe, which publishes 5–15 pages of EU news every day. Over the eight–nine years in the 1960s analysed, this amounts to more than 20,000 pages of detailed coverage of the EU’s daily affairs.

The creation of comitology in the agricultural area The Rome Treaty gave the Commission two years – until the end of 1959 – to propose a Common Agricultural Policy for the six member states (Lindberg, 1963, pp. 219–83). According to the treaty, the common organization of agricultural markets within the Community could take one of three forms depending on the specific product concerned: • A new market organization at the European level • Compulsory coordination of the member states’ own national market organizations • Common rules of competition Assuming that the Commission wanted to maximize its control, it must have preferred the first solution. At the same time, the Commission knew that agriculture was an area of national concern and the member states were therefore more likely to favour the second or third solution. Somehow a balance had to be struck between these positions, and the Commission had the privilege of being the first mover. The Commission’s proposal came in three main rounds. The first in November 1959 was a preliminary draft proposal presented somewhat prematurely to respect the treaty’s two-year deadline. The second proposal in June 1960 was the Commission’s final draft proposal. One year later, in the summer of 1961, the Commission presented a formalized proposal, a set of draft regulations. These were decided upon when the Council reached its final compromise in January 1962. In the first and second round (November 1959–June 1960), the Commission presented its main proposal. The treaty’s first and most radical

The Origins of Comitology 59

possibility – new market organizations at the European level – was proposed for cereals, dairy products and sugar. The reason was that in these areas the member states’ own policies differed the most which, according to the Commission, made a mere coordination of national systems impossible. The second possibility was proposed for beef, veal, pork, poultry and eggs; and the third for fruit, vegetables and wine. New market organizations at the European level also meant supranational administration. The Commission originally proposed a small set of so-called European offices for cereals, dairy products and sugar, respectively. But in the June 1960 proposal European offices were also proposed for beef, veal, pork, poultry and eggs, although with less extensive functions. These offices would all be supranational administrative entities under the Commission, financed by the agricultural import levies. They would administer the systems of target and intervention prices, make support purchases on the internal market and provide external protection by means of variable levies and import licences. According to the Commission’s proposal, the offices would be assisted by product-specific consultative committees. In the November 1959, proposal the Commission suggested creating a committee with equal representation from the Commission, the member states, the agricultural producers, industry, trade and the agricultural workers. Initially the committee was only suggested in the sugar area, but in June 1960 this proposal was adjusted. First, the Commission expanded the proposal to include consultative committees for all main products. Second, the Commission now envisaged two types of advisory committees: consultative committees representing private agricultural interests and so-called directors’ committees consisting of directors of the national bodies responsible for the various product markets. The latter were the first steps on the path to the comitology system (EEC-Commission, 1959; 1960, section 229–59; 1961a, pp. 126–38; EEC-Bulletin, 5/1960, pp. 38–46; PEP, 1961). At this stage, the member states did not pay much attention to the Commission’s idea of directors’ committees. Their prime concern was the policy contents of the Common Agricultural Policy, not least the price policies. For instance, German agricultural prices were the highest among the six member states and, not surprisingly, German farmers and the German government were against the idea of a common – and lower – price level (PEP, 1961, pp. 159–65). But administrative worries were soon aired. France made it clear that, while generally positive towards the Common Agricultural Policy, it was sceptical about its institutional aspects. It accepted that European market organizations

60

The EU Comitology System in Theory and Practice

might be necessary, but wanted the Council to be in control. Germany was outright hostile and rejected the idea of supranational bodies. Belgium, Luxembourg and Italy were also sceptical. Even the Netherlands, the warmest supporter of the Common Agricultural Policy and supranational institutions, was initially sceptical (Agence Europe, 10 Oct. 1960, pp. 2–3). To properly deal with the Commission’s Common Agricultural Policy proposal, the member states created the Special Committee on Agriculture to prepare the Council’s decisions. Its work began in September 1960, and for the next 18 months it would be the member states’ central arena for deciding the Common Agricultural Policy. However, negotiations soon proved to be very difficult. Frustrated at the lack of progress, the Commission tried to push the Common Agricultural Policy project forward. In November 1960, it challenged the member states to take a stand on the institutional aspects of the Common Agricultural Policy, especially the proposed directors’ committees. This got a ‘fairly cool reception’ from the Special Committee on Agriculture, which found it ‘far too early’ to raise this question (Agence Europe, 26 Nov. 1960, p. 2; see also 29 Nov. 1960, p. 1; 1 Dec. 1960, p. 2; 7 Dec. 1960, p. 2; EEC-Bulletin, 10/1960, pp. 47–8). The member states were clearly not yet ready to discuss the Common Agricultural Policy’s precise administrative set-up. In the spring of 1961 the Commission then worked hard to turn its general Common Agricultural Policy proposal into a set of concrete draft rules – regulations, directives and recommendations (EEC-Commission, 1962, pp. 142–3; EEC-Bulletin, 2/1961, pp. 31–2; 3/1961, pp. 41–3; 4/1961, pp. 46–7; 5/1961, p. 38). During this process, the member states accepted the idea of common European market organizations, and not merely a coordination of national market organizations (Agence Europe, 6 April 1961, p. 4, 7 April 1961, pp. 2–3). The Commission then directed its efforts to policy substance. In the following months it discussed substantial matters such as the agricultural price system and the exact nature of the levy and export subsidy system with member states’ representatives. These were thorny issues and no agreement was reached before the draft regulations had to be presented (Agence Europe, 1 July 1961, pp. 3–4; 5 Oct. 1961, p. 4; 9 Oct. 1961, pp. 3–4; 10 Oct. 1961, pp. 2–3; 24 Oct. 1962, pp. 3–4; 31 Oct. 1961, pp. 3–4; 22 Nov. 1961, p. 5). In the summer of 1961, the Commission then presented draft regulations for grain, pork, eggs, poultry, wine, fruit and vegetables to the Council. A few months later proposals for common policies (but not

The Origins of Comitology 61

concrete draft regulations) in the markets for rice, oils and fats followed. The Commission still suggested advisory directors’ committees, the forerunners of comitology, in the various product markets, but it was a consultative system without any legal foundation. The concrete draft regulations contained no provisions specifying its exact nature; only that consultation somehow had to take place (see e.g. EEC-Commission, 1961b, Article 5). Similar non-binding provisions were announced in the rice, fats and oils markets (Agence Europe, 2 Aug. 1961, p. 3; EEC-Commission, 1961c). In the Council and its Special Committee on Agriculture, matters did not really move forward until November 1961 when Germany showed willingness to compromise (Agence Europe, 30 Nov. 1961, p. 2) and the member states could finally discuss national control of the supranational administration of the Common Agricultural Policy. France presented a concrete proposal. Although realizing that this would introduce organs not provided for in the Rome Treaty, France suggested that the powers the member states had to give up be delegated not to the Commission, but to specially appointed committees composed of representatives of the Commission and the member states and equipped with their own administrative secretariats. The French proposal was a new version of the Commission’s proposal of directors’ committees. The proposal was discussed by the Council’s Special Committee for Agriculture. Most member states were in favour of some system of control by committees, but opinions diverged on how these committees should make decisions. France, Belgium and Italy favoured qualified majority voting. Germany, sceptical as usual, preferred unanimity. The Netherlands was against the whole idea and ‘concerned that the Commission should be handing over part of its responsibilities to intergovernmental organisations, and consider that this might be a dangerous precedent for the common policy in all sectors of the Community’s economic life’ (Agence Europe, 30 Nov. 1961, p. 3). No agreement was reached in the Special Committee. The French proposal was then discussed in the Council in December 1961. Working now under intense time pressure since failure to agree on the Common Agricultural Policy before the new year would prevent the Rome Treaty from passing on to the second stage, most member states approved. The Netherlands was still against and wanted ‘the Commission to be the sole body to take the day-to-day decisions necessary for the application of the common policy’ (Agence Europe, 13 Dec. 1961, p. 2; see also Agence Europe, 14 Dec. 1961, p. 1). Before Christmas, the Council agreed in principle on the establishment of the committees

62

The EU Comitology System in Theory and Practice

proposed by France, but disagreement about the committees’ decision method continued – Germany favoured unanimity, the other member states some kind of majority voting (Agence Europe, 20 Dec. 1961, p. 2). On 18 December 1961, Commission President Walter Hallstein made a compromise proposal to the Council: For day-to-day decisions necessary to the functioning of common market organisations, the Commission informs the administrative committees of its intentions. If the latter do not approve of these, or do not take a decision within a given period, the decisions of the Commission are put into force. If the committees decide by a qualified majority of 12 votes against the EEC Commission’s draft decisions, the latter can refer to the Council of Ministers, which has one month in which to arbitrate. It decides by a qualified majority on the Commission’s draft decisions. If the Council does not take a decision within the prescribed period, the EEC Commission’s decisions are applied without modification. (Agence Europe, 22 Dec. 1961, p. 2) Note that this proposal modified the Commission’s original proposal of directors’ committees in one respect only: their formal competence. They were no longer merely advisory, but would have real decision-making power. However, Hallstein’s proposal made independent secretariats for the committees superfluous. Hallstein wanted the Commission to have this function. Although the Netherlands was still against and wanted the committees to have a purely advisory function, it is clear in hindsight that Hallstein’s proposal was the decisive compromise. Comitology was born. The Dutch reservations were withdrawn between Christmas and New Year (Agence Europe, 2 Jan. 1962, p. 4). This did not mean that the Common Agricultural Policy was enacted. Disagreement on the policy contents continued and was not settled until 14 January 1962. But the decision to create the comitology system was taken. The committees were soon officially called ‘management committees’ instead of ‘directors’ committees’ (as in the Commission’s original proposal) or ‘administrative committees’ (as in the French proposal). Establishing committees with formal powers may seem like a radical departure from the Commission’s original proposal of consultative directors’ committees. Certainly that is how it was perceived at the time. For instance, the European Parliament (1962), in a resolution that would become typical of future inter-institutional negotiations on the comitology system, protested vigorously and denounced the Council’s

The Origins of Comitology 63

decision which ‘would curtail the functions of the executive Commission’ and ‘would rob the Commission of the powers it holds in conformity with the spirit of the Treaty’. However, given that the member states’ control problem could be solved in various ways, it is striking how close the end result was to the Commission’s original proposal. Parliamentary control of delegated legislation made by the executive is a problem all member states knew from their national systems, but this is not where the idea of the comitology system came from. Few national systems have any specific control instruments of delegated legislation (although parliamentary committees may have oversight responsibilities). At the national level controlling delegated legislation is considered part of the general problem of parliamentary control of the government. This means that it relies on general parliamentary control instruments such as the ministerial responsibility doctrine, parliamentary questions, committee investigations, funding, audit reports, ombudsman systems and so on (Peters, 1995, pp. 289–337). None of these well-known institutional solutions, or for that matter other models, were ever discussed by the member states. Since the EU is not a parliamentary system, but a type of separation-of-powers system, it might have been relevant to build on institutional solutions found in these systems. For example, Bignami (1999; see also Hix, 2000) argues that important lessons on how to control delegated decision-making can be learnt from the federal system in the USA, in particular the provisions in the US Administrative Procedure Act from 1946 on notice, comment and judicial review. However, there is no record that the US experience was ever taken into account in the decision to establish the agricultural management committees. The member states focused on a solution unknown from their own national systems and other systems. The reason is that this idea was presented to them by an external actor: the Commission. The effect was that the complex institutional question was reduced to one of more or less comitology. To conclude this section, I return to hypothesis 1 and 2. The first hypothesis held that the member states had divergent preferences about the Common Agricultural Policy’s administration. The above historical account makes this hypothesis difficult to refute. Preferences varied from Germany’s hostility to the Netherlands’ support. The second hypothesis held that the Commission’s proposal for an administrative set-up for the Common Agricultural Policy represented a solution on which intergovernmental bargaining converged. This

64

The EU Comitology System in Theory and Practice

hypothesis is also difficult to refute. Initially the member states were against the Commission’s suggested directors’ committees, but at the same time they found it difficult to propose something else. Instead of debating a range of alternative institutional solutions to their control problem – for example, well-known solutions to similar problems in their own national systems – the member states discussed various degrees of comitology. The end result was an administrative system strikingly close to the Commission’s original proposal. It is also striking how quickly the solution was found. The negotiations on the Common Agricultural Policy lasted more than two years, November 1959–January 1962, but as the historical account makes clear, most of this time was spent negotiating the difficult policy contents of the Common Agricultural Policy. Negotiations on its administration did not really start until November 1961; it was settled in less than two months. In short, it seems that the Commission succeeded in constructing a focal point for the negotiations on the Common Agricultural Policy’s administration.

The spread of comitology to other areas Committees with monitoring powers comparable to the agricultural management committees were established in other policy areas during the 1960s (Bertram, 1967–68, p. 249; Bergström, 2005, pp. 78–88). Was this due to the precedent of the agricultural management committees in 1961–62? This is the third and final hypothesis to be investigated. If the management committees did indeed function as a focal point in this process, we must look for references to them in negotiations leading to similar structures in other areas. A first, but powerful hint that this might be the case is found in statements made by the French government during the Community’s ‘empty-chair crisis’ in 1965–66. The crisis was not solved until January 1966 when extraordinary meetings were held in Luxembourg to discuss a wide range of institutional aspects of the Community, including delegation of implementing powers to the Commission. France, more worried than ever about losing national control, reminded the other five member states of the following: In certain cases the Commission can obtain authority from the Council to put into effect the rule which the latter lays down. This delegation of powers must not imply that the tasks entrusted to the Commission will then be outside the purview of the Council. True, in certain sectors such as agriculture, the Council can intervene at executive

The Origins of Comitology 65

level through its representatives on the Management Committees. However, . . . the Commission is endeavouring to replace the Management Committees by simple advisory committees. (French foreign minister Couve de Murville, 1966; emphasis added) This statement, made during the most profound crisis the Community had yet experienced, is powerful evidence that the agricultural management committees established four years earlier had left a lasting impression as a possible solution to the general dilemma of national control and supranational administration. In agriculture, management committees were soon used whenever the member states wanted to monitor the Commission’s rule-making powers. In a reply to a question from the European Parliament in 1968 the Commission reported that nine management committees had been established covering the areas of cereal, pork, beef and veal, dairy products, poultry and eggs, fruit and vegetables, oils and fats, wines and sugar. Their meeting frequency varied; the most active, the management committee for cereals, met 53 times from March 1967 to April 1968 (Agence Europe, 25 June 1968, p. 7). Management committees were soon used in other agricultural areas, such as financing, information and research (EEC-Bulletin, 11/1965, p. 21; Bertram, 1967–68, p. 249). What about other policy areas? In the following paragraphs I take a closer look at development aid to Africa and the customs area where the member states soon needed to delegate rule-making powers to the Commission as well, but were sceptical about losing national control.

Development aid to Africa The Rome Treaty obligated the Community to give associated status, including favourable trade conditions and development aid, to a number of overseas countries and territories with special, for example colonial, relations with the member states. The treaty provision was implemented in a five-year convention associating a number of African states with the Community and in the creation of the European Development Fund, which financed development projects in the associated states. Since then, the convention has been renewed every five years: the Yaoundé Conventions (1964–75), the Lomé Conventions (1975–2000) and the Cotonou Agreement (2000–). The first renewal would be negotiated in 1962. The Community envisaged a considerable expansion of development aid and a firmer institutional foundation of the convention. The question soon arose

66

The EU Comitology System in Theory and Practice

of how to reconcile national control and supranational administration of the increased development aid. Under the first convention, projects to be financed by the European Development Fund were proposed by the Commission and approved by the Council. Increasing the amounts available for aid meant a rapidly growing number of projects to approve. In order to reduce its workload, the Council decided that further powers had to be delegated to the Commission. To study the problem, the member states established a working group of national experts under the Committee of Permanent Representatives (Coreper). In September 1962, the group suggested that the Commission remain responsible for proposing projects to be financed by the European Development Fund, but that an approval system similar to the management committees in the agricultural area should be established. In a reaction that would become typical for future inter-institutional negotiations about the comitology system the Commission vehemently opposed this proposal: This procedure is not at all to the Commission’s tastes. . . . it takes away from the Commission part of the powers which it has under the present association regime. . . . the national experts’ plan alters the relationships between the Commission and the Council, for it sets between them a body to which the Council will, in a certain sense, delegate part of its powers. But the Commission does not recognise any partner other than the Council, and believes that the proposed procedure would threaten the institutional balance set up by the Treaty. (Agence Europe, 27 Sept. 1962, p. 4; see also 22 Oct. 1962, p. 2) But the Council liked its experts’ proposal, and it took care to rectify a point which had proven unclear in the daily workings of the agricultural management committees: the committees, established only a few months earlier, gave their opinion on the Commission’s draft proposals by qualified majority. However, if the committees voted against a Commission proposal, but without a majority large enough to constitute a qualified majority, the implications were unclear. Could the Commission implement its draft rule since the committee had not given an opinion, or should the Council intervene? This problem, which had already occurred in several committees, was discussed during the summer of 1962, but any firm conclusions were not reached. The Commission maintained that it was legally entitled to implement its proposed measures if no qualified majority was raised against them, but also said that in practice it would be flexible (see Agence Europe, 19 July 1962, p. 3; 24 July 1962, p. 1; 27 July 1962, p. 4). The Council

The Origins of Comitology 67

wanted no confusion in development aid. If the new committee did not approve the Commission’s proposals by a qualified majority, the Commission had to withdraw the proposals or seek arbitration by the Council. Only in the case of a formal approval by a qualified majority was the Commission to implement its proposal. The issue was hotly debated between the Council and the Commission during the autumn and winter of 1962, but in December 1962 the Commission gave in and accepted the strengthened version of the procedures used by the agricultural management committees (Agence Europe, 27 Nov. 1962, p. 2; 17 Dec. 1962, pp. 5–6; 18 Dec. 1962, pp. 1–2). The remaining obstacles to the renewal of the convention were cleared away in the spring of 1963, and at the Council’s 100th session on 1–2 April 1963 all issues were declared settled (EEC-Bulletin, 4/1963, pp. 53–6). The new procedure for approving development projects was an internal agreement between the six member states (Agence Europe, 6 April 1963, pp. 2–3) and not a formal part of the convention, which was finally signed in Yaoundé, Cameroon in July 1963. The internal agreement spelled out that the Commission had to withdraw its proposal or contact the Council not only if a qualified majority were raised against it in the new committee, but also if the committee failed to approve the proposal by a qualified majority (Council of Ministers, 1964, Article 11). With these rules, the committee – soon to be known as the ‘EDF committee’ because it approved financing by the European Development Fund – was the first example of what was later known as a regulatory committee.

The common customs union According to the Rome Treaty, a customs union was to be gradually established between the member states over 12 years, abolishing all national customs duties and establishing a common external tariff in relation to third countries. Although this process went far quicker than expected and the customs union was established already on 1 July 1968, one and a half years earlier than the deadline specified by the treaty, the creation of a common customs regime represented quite a challenge. The problem was twofold. First, the customs union did not entail the creation of a common European customs service as the common rules would be administered by the existing national customs services. Second, the member states already had national rules on a host of common technical issues, such as rules on classification of goods, storage of imported goods, valuation of goods, transit rules, rules on duty-free areas, duty-free entry rules, granting of customs exemptions, systems

68

The EU Comitology System in Theory and Practice

of customs payment, rules on formulas and customs documents, collections of statistics and trade recordings, rules on the reexportation of imported goods, and many other issues. In the absence of common rules and a uniform administration, third countries would have an incentive to export to the member state with the most lenient rules. This would lead to distortions of trade and customs revenue. In short, the creation of a customs union in reality meant much more than the abolishment of national customs duties and the establishment of a common external tariff. It also meant the harmonization of customs rules and administrative practice. It was a massive task. The Council would not be able to set all rules and obviously some rule-making competence had to be delegated to the Commission. The problem of national control and supranational administration popped up again. The problem was particularly pressing in two areas. First, a common definition of the origin of goods had to be established. Since tariffs, quotas, exchange controls, embargoes, retaliatory duties and so on differ according to country of origin, a precise and common determination of the origin of imported goods is necessary for any customs regime. However, there was no agreed upon international definition and all member states used different definitions. The Commission proposed common rules in December 1964 (EEC-Commission, 1965a, b). Second, common rules on the valuation of goods for customs purposes were necessary. Again, if goods were not valuated in similar manners in all member states, third countries would have an incentive to export to the member state with the most lenient rules because a lower value would mean a lower customs duty under the common customs tariff. Although all the member states had signed the international Brussels Convention from 1950 on the valuation of goods, common rules and practice were not ensured. The convention contained optional clauses which were implemented in different manners in the six member states. In addition, the national courts in the member states interpreted the Brussels Convention in different ways. The Commission proposed common rules in November 1967 (Commission, 1968). Both proposals entailed delegation of implementing powers to the Commission to specify more detailed rules. To assist it, the Commission proposed that an Origin of Goods Committee and a Valuation Committee be established. They were to be composed of member state representatives and to have an advisory function. The member states did not seriously consider the Commission’s proposals until the spring of 1968. This created some time pressure because by then the member states had already agreed to establish the customs

The Origins of Comitology 69

union and to harmonize customs rules by 1 July 1968 (see EEC-Bulletin, 7/1966, p. 12; 9–10/1967, pp. 25–9). The Commission was frustrated (see Rey, 1968, pp. 9–10), but by May 1968 things were still unsettled, for example the powers of the two committees to assist the Commission in the definition of origin and valuation of goods (Agence Europe, 6 May 1968, pp. 4–5). But at the Council meeting on 30–31 May 1968 a decision was reached and soon formalized into regulations (Agence Europe, 30 May 1968, pp. 8–9; EC-Bulletin, 7/1968, pp. 12, 46; Council of Ministers, 1968a, Articles 12–14; 1968b, Articles 15–17). Reluctant to delegate powers to the Commission, the member states decided to give the committees a status comparable to that of the agricultural management committees, but with stronger powers. Like the management committees, the two customs committees would approve Commission proposals by qualified majority. However, unlike the agricultural management committees, but like the European Development Fund committee, the Commission would only be able to implement its proposal in case of a formal approval by a qualified majority. Failing this, the Commission had to withdraw its proposals or seek arbitration by the Council. Like the European Development Fund committee, the customs committees were thus examples of what was later known as regulatory committees. However, unlike the European Development Fund committee, a further procedural clause was decided, namely that if the Council had not decided on a proposal within three months, the Commission could put the proposal into force. This was an example of a procedure soon to be known as the ‘safety net’ (‘filet’). The inspiration from the agricultural management committees was evident to everyone. As noted by Agence Europe (30 May 1968, p. 9): Once again the definition of procedures involves replying to the problems of . . . how responsibility is divided between the European Commission and the Committees made up of representatives of the national administrations. . . . Although in a different form these are the same problems as those concerning the Management Committees for agricultural products.

The spread of comitology to other areas: Summing up the evidence The third hypothesis stated that the original choice of administrative model for the common agricultural policy represented a solution on which intergovernmental bargaining converged in other policy areas

70

The EU Comitology System in Theory and Practice

in which supranational administration and national control had to be reconciled. In development aid to Africa and the customs union the member states chose institutional solutions that were strikingly close to the agricultural management committees, and they argued for this choice in terms of differences and similarities to the management committees. Alternative institutional solutions, for example solutions known from the member states’ national systems, were never discussed. In short, it seems that the agricultural management committees functioned as a focal point in negotiations on similar problems in other policy areas.

Conclusion The dominant explanation of the origins of the EU comitology system is a functional one. Committees are created to enable the Council to delegate powers to the Commission without losing national control. The evidence from this chapter’s historical analysis of the introduction of the agricultural management committees in the early 1960s confirms this explanation. The system was indeed created by the Council to monitor the Commission. But although the functional explanation is correct on motives, it cannot explain why comitology was chosen among the many potential solutions to the Council’s delegation problem. The functional explanation focuses on the member states as the decisive actors. The focal point explanation focuses on the Commission. Informal agenda-setting power was delegated to the Commission, and the Commission, not the member states, proposed committees as a solution. To the member states, caught in a coordination problem, this represented an obvious solution. It was a focal point, and a powerful one. It simplified matters and reduced the complex control problem to one of more or less comitology. It functioned as a precedent not only in agriculture, but also in other policy areas. When the same problem of supranational administration and national control arose in relation to development aid to Africa and the customs area, the member states found solutions similar to the management committees and never seriously considered alternative solutions, for example well-known solutions to similar problems in their own national systems. The chapter thus offers an explanation not only of the original choice of comitology, but also of its spread to other policy areas. The chapter illustrates the value of actively investigating institutional origins and the limitations of functionalist reasoning. Functional motives may be decisive, but this should be treated as a hypothesis,

The Origins of Comitology 71

not a premise. Investigating institutional origins enables the researcher to pay proper attention to choice. Institutional design always involves choice, but functionalist reasoning makes it difficult to explain the exact institutional choice when there are several solutions to a given functional need. The analysis is a narrative of events in a particular period and setting, but it is an analytic narrative based on deductive reasoning. The member states’ choice of the comitology system was modelled as a coordination problem solved by a constructed focal point. The analysis thus confirms the importance of focal points and informal agenda-setting for solving coordination problems. The analysis also confirms a central point from the delegation literature, namely that delegation to a central decisionmaker may solve collective action problems among legislators.

5 What is at Stake? A Case Study of the 2006 Comitology Reform

It is no secret that the EU institutions do not agree on the design of the comitology system. In 2006, when an inter-institutional compromise was about to be made on a revision of the 1999 comitology decision, Richard Corbett, the European Parliament’s comitology negotiator, gave his fellow Parliament members this explanation of ‘the problem’: It may be recalled that the problem was created with the development in the 1960s and the 1970s of a system for delegating implementing powers to the Commission whereby the latter is empowered to act in conjunction with a committee of national representatives. . . . such committees could block the Commission’s decision on an implementing measure . . . Parliament criticized the fact that only a committee, and not the Parliament, had such a right . . . Parliament took the view that codecision acts . . . implied that they should both be involved in defining the procedure for exercising delegated powers and that they should have equal rights regarding retrieval or callback. Council, however, argued . . . for Council (alone) to define the system for implementing powers. (European Parliament, 2006a, p. 6) It is not only participants in the EU decision-making process who talk about conflicts over the comitology system, but outside observers are also struck by the inter-institutional rivalry about the system (Franchino, 2000a; Pollack, 2003a, pp. 114–46; Bergström, 2005, pp. 285–308; Bergström et al., 2007; Bradley, 2008). Yet the question remains why there is so much conflict. What exactly is at stake? This chapter addresses this question and is the first of three chapters to investigate the driving forces of the comitology system. 72

What is at Stake? A Case Study of the 2006 Comitology Reform

73

In Chapter 3, I made the argument that the design of the comitology system is a game of control positions. No one in the EU system knows for certain how delegated powers to the Commission are going to be used or what the future precisely entails. But everybody knows that there will be decisions to make down the line. When specifying the details of the various comitology procedures, the EU institutions cannot precisely specify the contents of these decisions, but they can make sure that they will be in a position to influence them. This is the politics of structural choice (Moe, 1990a, b) or ‘deck stacking’ (McCubbins et al., 1987, 1989). It means that the individual EU legislators will press for comitology rules that ensure efficient institutional control positions for themselves and inefficient control positions for their opponents. This again means that the legislators will have different preferences over the comitology rules. Based on this argument the following hypotheses were developed: Hypothesis 4: The Council favours strict comitology procedures. Hypothesis 5: The Commission favours permissive or no comitology procedures. Hypothesis 6: The European Parliament favours: a. narrowing delegation to technical matters b. permissive or no comitology procedures, or alternatively c. access to the comitology system This chapter provides the first of two investigations of these hypotheses. While Chapter 6 focuses on daily legislation, this chapter focuses on the level between daily legislation and the treaty, namely the framework rules specifying a list of comitology procedures the legislators can choose from in daily legislative practice. Rules on these procedures were introduced in 1987 in the wake of the Single European Act. The background was that there were more than 30 variants of comitology procedures in operation by the 1980s. The ambition in the intergovernmental conference leading to the Single European Act was to cut down the range of procedures and establish an exhaustive list. Originally this list was to be set out in an annex to the treaty, but in the end, the decision was made to make the list by ordinary legislation using the consultation procedure (Bergström, 2005, pp. 185–92). The list was then made by the Council in its first comitology decision (Council of Ministers, 1987), which is under the treaty level, but over the level of daily legislation in the sense that the list specifies an exhaustive number of comitology procedures (Schaefer and Türk, 2007, p. 184; Türk, 2009, pp. 80–1).

74

The EU Comitology System in Theory and Practice

The framework rules have been changed three times since the Council’s first comitology decision in 1987. The first time was in 1999 when several variants of the basic procedures were removed. The second time was in 2006 when a new procedure was added to the list, the so-called regulatory procedure with scrutiny. The third time was in 2011 when the Council’s comitology decision was replaced by a comitology regulation made by the European Parliament and the Council in the wake of the Lisbon Treaty. The 2011 reform introduced a new examination procedure modelled on the management and regulatory procedures (Brandsma and Blom-Hansen, 2011). This chapter provides a case study of the 2006 reform of the Council’s comitology decision. This makes it possible to study the EU actors’ comitology preferences in a setting where the basic rules of the system are up for grabs. If the hypotheses are true, they should hold in this situation since it represents an obvious opportunity for all actors to improve their control positions over delegated decision-making. In other words, a reform of the Council’s comitology decision is a most likely case. If the hypotheses do not hold here, they are unlikely to hold elsewhere. The case study format not only allows an investigation of institutional preferences, it also makes it possible to put some flesh on what institutional interests and instruments mean in practice in the area of comitology.1 The chapter is organized as follows. I first explain methods and data applied to investigate the 2006 comitology reform. The empirical analysis investigates the four partially related processes that led to the reform: the Lamfalussy reform in financial regulation; the controversial use of comitology in GMOs, food safety and the environment; the failed Constitutional Treaty; and the revision of the 1999 comitology decision. The chapter concludes with a discussion of the lessons in relation to preferences, interests and instruments in the comitology area.

Methods and data The purpose of the case study is to gain an understanding of the driving forces of the development of the comitology system. Preferences, interests and instruments are identified by studying the process leading to the reform. Pressure for reform had built up over several years in two unrelated areas: regulation of financial services; and food safety and environmental issues. These pressures were first channelled into the process that led to the proposal for a Constitutional Treaty for the EU. When this proposal failed because of the referenda in the Netherlands

What is at Stake? A Case Study of the 2006 Comitology Reform

75

and France, the pressures were channelled into the process that led to the revision of the 1999 comitology decision. Studying these four processes is an exercise in process tracing (George and Bennett, 2005, pp. 205–33) as I seek to uncover the chain of events in the four processes that led to the final outcome. The four processes do not constitute independent cases to be compared; on the contrary, the goal is to study how they are causally related to each other and to identify the preferences, interests and instruments of the actors during the process. The EU is a fertile hunting ground for the process tracer because of the considerable paper trail left behind by the decision process. There is a wealth of official EU documentary data to study how events are linked in the policy process. However, there is one exception. The deliberations of the member states in the Council of Ministers are notoriously difficult to uncover. At best, official documentary data consist of brief press releases or minutes of Council meetings. To study the negotiations and compromises in the Council I turn from EU documents to national documents in one member state, Denmark. The Danish Parliament probably has the most powerful EU committee of any member state (Kassim, 2000; Bergman et al., 2003, pp. 173–7) and it is standard operating procedure for the Danish government to inform the Parliament’s EU committee on all EU issues, including comitology issues. This information is made public on the Parliament’s website (www.ft.dk).

Process 1: The Lamfalussy reform in the area of financial regulation Despite the deadline specified by the Single European Act, the internal market for financial services – banking, insurance and investment services – was not in place by the end of 1992. Differences in national regulation proved too difficult to reconcile, but the introduction of the euro led to increased pressure for reform. In 1998, the European Council invited the Commission to produce a plan to improve the single market in financial services. One year later the plan was ready. It proposed a large number of legal initiatives, but also complained of ‘painstakingly slow’ legislation and proposed ways to ‘accelerate the legislative process for financial services’ (Commission, 1999a, pp. 14–15). To deal with this challenge the Council established a Committee of Wise Men chaired by Alexandre Lamfalussy, former president of the European Monetary Institute.

76

The EU Comitology System in Theory and Practice

The Lamfalussy committee proposed a two-level legislative approach to financial regulation. At the first level, regulatory framework principles were to be decided by the normal EU legislative procedure (i.e., codecision). At the second level, more detailed technical measures were to be decided through a comitology procedure involving two new committees: the EU Securities Committee and the EU Securities Regulators Committee. This approach presupposed a distinction between policy formulation and implementation, or, in the EU jargon, essential and non-essential measures. The Lamfalussy committee did little to clarify the line between the two types of legislative measures. But since the committee’s emphasis was on regulatory flexibility and fast-track legislation, a reasonable expectation was that much future regulation of financial markets would take place at level two if the committee’s proposal were adopted (Committee of Wise Men, 2001; see also Avgerinos, 2002; de Visscher et al., 2008). The Lamfalussy proposal was warmly welcomed by the Commission (2001b). The member states were also positive, but reluctant to delegate far-reaching powers to the Commission since, after the 1999 comitology decision, a qualified majority in the Council of Ministers was required to block a Commission proposal referred to it under the regulatory comitology procedure. Especially Germany was sceptical and worried that the Commission might use its newly gained powers under the 1999 decision to push through legislation opposed by a substantial minority of member states. For instance, Germany feared that the Commission might favour London over Frankfurt as a centre for securities trading (Pollack, 2003a, pp. 140–4). However, the German objections were withdrawn when the Commission reiterated its statement regarding the 1999 comitology decision that it would ‘avoid going against predominant views’ in the Council. At the 2001 Stockholm summit, the European Council endorsed the Lamfalussy proposal. The European Parliament was more sceptical. Fearing that the Lamfalussy proposal would cause financial regulation to slide from the codecision procedure into the comitology system outside the Parliament’s control, it demanded that level one regulation contained not only framework principles, but also implementation measures. Furthermore, it wanted a call-back right in case delegation did not go as expected (European Parliament, 2001). After one year of inter-institutional negotiations a compromise was reached. It did not entail a call-back right for the European Parliament, but specified, first, that a sunset clause would be inserted into all Lamfalussy legislative acts adopted at level one. This clause would limit

What is at Stake? A Case Study of the 2006 Comitology Reform

77

delegation to four years. Second, the compromise gave the European Parliament three months to react to the Commission’s draft implementing measures. Third, the European Parliament was granted a more comprehensive supply of comitology documents than under the 1999 comitology decision. Both the European Parliament (2002, point 21) and the Commission (2002a) stressed that this compromise was temporary and only effective until the comitology system would be more thoroughly reformed by the next treaty revision, which was already set in motion by the European Council at the Laeken summit in 2001 (cf. below). The further reform pressures generated by this compromise were threefold. First, the European Parliament soon showed that the use of sunset clauses to limit delegation was meant seriously. To the annoyance of both the Council and the Commission, sunset clauses were now systematically inserted into financial directives.2 Second, the Council of Ministers’ new voting rules under the 1999 comitology decision on issues referred to it under the regulatory procedure gave pause to some member states. It was no longer easy to stop the Commission since this now required a qualified majority. Third, the Lamfalussy area illustrated the difficulties of drawing a clear line between policy formulation and implementation. It demonstrated a huge grey zone between these two concepts and thus increased the salience of the European Parliament’s weak status in the implementation arena. As a final note, it may be added that the European Parliament once again learnt that resistance in the area of comitology may pay off in terms of influence. In terms of institutional preferences, the Lamfalussy reform showed that the European Parliament is highly sceptical about the comitology system. It considers it a threat to its institutional position in the EU system, and it is willing to go to considerable lengths to defend this position. As for the member states, the reform demonstrates that they use comitology as an instrument to control the Commission. Finally, concerning the Commission the findings are mixed. It was in favour of the Lamfalussy reform including its strict comitology procedures, but they were probably accepted as a price worth paying for more delegated decision-making.

Process 2: Comitology in sensitive areas: GMOs, food safety and the environment In the areas of genetically modified organisms (GMOs), food safety and the environment, EU regulation has generated significant public debate

78

The EU Comitology System in Theory and Practice

and raised concern about health and environmental risks. The areas are politically sensitive in the member states; sometimes so sensitive that they prefer to opt out of the Community to avoid EU regulation. Since the control of the Commission in many of these areas rests on comitology procedures, actors in these areas are often sceptical about reforms of the comitology system that increase the autonomy of the Commission. The most prominent example of such a sensitive area is the use of GMOs in food and feed. EU regulation has been gradually expanded in order to deal with safety and legitimacy issues. In 1990, the first GMO directive introduced a uniform application procedure according to which permission to market GMOs must be sought with a competent food authority in one of the member states. This authority’s decision is then forwarded to the Commission and the food authorities in the other member states. If objections are raised, the matter is settled by the Commission acting under the regulatory comitology procedure (Skogstad, 2006). In principle, this system allows the member states to control the Commission’s approval of individual GMOs, but it presupposes that the member states are able to form an opinion by qualified majority, either in the comitology committee or in the Council. The comitology rules state that if a regulatory committee cannot form an opinion, the matter is referred to the Council. If the Council fails to form an opinion within three months, decisional power reverts to the Commission, which is free to adopt the draft rule. This deadlocked situation has often occurred. There are several examples of the Commission approving GMOs against the will of the comitology committee and the Council, but where the majority against the Commission’s proposal in these organs is not large enough to constitute a qualified majority (for a vivid example see Bradley, 1998; see also Pollack and Shaffer, 2008). This situation was unsatisfactory to a number of member states who faced strong public resistance to GMOs. Their reaction was to unilaterally defect from the EU regime: they invoked the safeguard clause in the GMO directive. An individual member state can restrict or prohibit the use or sale of a given GMO if it fears that it constitutes a risk to human health or the environment. This is meant to be a provisional measure, but in reality member states have been allowed to keep safeguard clauses in operation for years. The Commission might have used the European Court of Justice to enforce its decisions, but has refrained from doing so, presumably fearing the anti-GMO opinion in the member states (Skogstad, 2003).

What is at Stake? A Case Study of the 2006 Comitology Reform

79

In 1999, tension grew as the 1987 comitology decision was up for renewal. The Commission wanted to simplify the existing two regulatory procedures (IIIA and IIIB) and merge them into one new procedure. According to the old IIIB procedure, the Council could block a Commission proposal by simple majority. As per the Commission’s proposal this would be replaced by qualified majority which would make it considerably more difficult to stop the Commission (Haibach, 1999). The Council was divided. But in the end, with an eye to the coming enlargement and after the Commission made an official statement that it would ‘avoid going against any predominant position’ in the Council when making proposals for implementing measures in ‘particularly sensitive sectors’ (Commission, 1999b), consensus rallied around the Commission’s proposal. However, this statement did not reassure the member states facing strong negative public opinions on GMOs, and with the negotiation of the new 1999 comitology decision they contemplated the introduction of a moratorium on the authorization and marketing of GMOs. It was formally established by a declaration by the Danish, Greek, French, Italian and Luxembourg delegations at the Council meeting of environmental ministers on 24–25 June 1999 that the moratorium would remain in force until rules ensuring labelling and traceability of GMOs were adopted. The Commission then proposed reforms that introduced three sets of regulatory changes. First, a pre-marketing safety assessment by food authorities in the member states and/or the newly created European Food Safety Authority; second, a ‘one-stop’ authorization procedure to achieve access to the internal market involving the Commission and the member states through a regulatory comitology procedure (the Standing Committee on the Food Chain and Animal Health); and, third, a labelling provision that the use of GMOs in food and feed is clearly stated on the product (Tsioumani, 2004; Buonanno, 2006). While these regulatory initiatives may have had a soothing effect on public sentiments, they brought the EU on a collision course with external trading partners, notably the United States, which wants to export genetically modified crops and claims a right to do so under the World Trade Organization (WTO) agreements (Boisson de Chazournes and Mbengue, 2004; Morgan and Goh, 2004). The member states reacted by lifting the moratorium in June 2005, when the Council refrained from blocking the Commission’s proposal to authorize the agricultural company Monsanto’s MON 863, a corn variety genetically modified to improve its resistance to insects. Lurking

80

The EU Comitology System in Theory and Practice

in the background was also the WTO, where the United States and other agricultural exporters had raised complaints that the EU moratorium constituted a trade barrier. WTO proceedings began in 2003 and ended, somewhat post festum, with a ruling in 2006 that the moratorium was illegal under WTO rules. However, this did not stop unilateral non-compliance by individual member states via the GMO safeguard clause. At the very same Council meeting in June 2005 that failed to block Monsanto’s GMO corn, the Council also permitted Austria, France, Luxembourg, Germany and Greece to continue their national bans of GMOs authorized prior to the moratorium. Although invoking the safeguard clause is supposed to be provisional, these national bans had by then been effective for five to eight years. GMOs represent an extreme case of the workings of comitology in sensitive areas. But it is not the only area where comitology rules fail to give member states satisfactory control of the Commission. The environmental area offers other, although less extreme, examples (Demmke, 2000; Flynn, 2000), such as the implementation of the Restriction of Hazardous Substances (RoHS) directive, which prohibits the use of hazardous chemical substances in new electrical and electronic equipment. However, the directive allows certain exemptions to this general rule, for example if the substances are used in medical equipment or if the equipment was put on the market before 1 July 2006. Exemptions are granted by the Commission acting under the regulatory comitology procedure (the Waste Committee). In several instances, the Commission has proposed exemptions which fail to obtain a qualified majority in the comitology committee, and are then referred to the Council, which also fails to master a qualified majority for or against. Decisional power then returns to the Commission, which adopts the exemption (Maxianova and Rusche, 2006) – that is, the same outcome as in the deadlocked situation of GMOs. Several member states have openly protested against this working of comitology. In 2005, five member states – Denmark, Sweden, Belgium, Finland and Portugal – sent a joint letter to the Commission complaining about the planned exemption of the ban on the use of decaBDE (a brominated flame retardant). These member states recognized that the exemption could not be blocked under the comitology procedure because a qualified majority could not be established, but they challenged the Commission’s use of the comitology procedure because, in their view, the exemption would be too far-reaching (Ministers of five EU countries and Norway, 2005).

What is at Stake? A Case Study of the 2006 Comitology Reform

81

The five member states even challenged the Commission before the European Court of Justice (ECJ Case C-295/06). In sum, in the sensitive areas of GMOs, food safety and the environment the member states have learnt the tough lesson that the comitology system cannot effectively control the Commission in areas where they are divided and where a qualified majority can be mastered neither in the comitology committee nor in the Council. In this deadlocked situation decision-making power reverts to the Commission and despite its promise in relation to the 1999 comitology decision not to go against ‘any predominant position’ in the Council in ‘sensitive sectors’ (Commission, 1999b), the Commission has demonstrated that it is ready to use this power in practice. In terms of institutional preferences, the workings of comitology in the areas of GMOs, food safety and the environment are most revealing about the Commission and the member states. As to the Commission, we learn that, although it is not unresponsive to the member states’ wishes, it is in practice willing to step into the void created by deadlocks in the comitology system. Concerning the member states, we once again see that they regard the comitology procedures as a control mechanism, and in this area they have learnt that comitology does not always work as an effective control. A number of member states developed a strong scepticism towards any suggestions of loosening the Council’s control mechanisms or increasing the autonomy of the Commission, at least in sensitive areas.

Process 3: The European Convention and the Constitutional Treaty At the Laeken summit meeting in 2001, the member states decided to convene a European Convention to prepare the treaty revision envisaged at Nice in 2000. The Convention began its work by establishing a number of working groups. The complicated task of simplifying the EU’s instruments and procedures landed with working group IX. This group dealt, inter alia, with the distinction between legislative and executive law-making and its reform proposal was far-reaching (European Convention, 2002; see also Bergström, 2005, pp. 335–51). It advocated a three-tiered hierarchy of acts. First, legislative acts were to be laws adopted by the Council and European Parliament directly on the basis of the treaty. Second, delegated acts should be adopted by the Commission and flesh out the details of a legislative act. This

82

The EU Comitology System in Theory and Practice

type of act would also lay down control mechanisms for the legislators of which the working group envisaged three types: call-back rights, a period of tacit approval (the delegated act will only enter into force if the legislators have not objected within a pre-specified period) and time limitations on delegation. The working group thus did not envisage any comitology mechanism for delegated acts. Third, implementing acts were to be adopted by the Commission to implement the former two types of acts, for which the working group suggested upkeeping comitology, although it speculated that the regulatory procedure might be abolished. The proposal was radical because the most controversial matters dealt with under comitology were taken out of this system and placed under equal supervision by the Council and the European Parliament as delegated acts. The working group’s proposal was accepted almost intact by the Convention. In its final draft for a Constitutional Treaty, the three-tiered hierarchy of legal acts was codified in articles 33–36 with only few changes to the working group’s proposal. Most notably, the option of inserting time limits for delegation (sunset clauses) was removed. The Convention’s proposal had a structuring effect on the ensuing intergovernmental conference, which began in October 2003. There is no sign that the Convention’s envisaged hierarchy of legal acts and the consequences for the comitology system played any large role in the negotiations among the member states. The only indication that the issue came up at all is a declaration adopted by the intergovernmental conference which aimed at preserving the Lamfalussy system of committees in the financial area (IGC, 2004). The intergovernmental conference was brought to a close within a year with only minor changes to the Convention’s draft treaty. The new Constitutional Treaty was duly signed in Rome in October 2004 but as is well-known, the ratification process soon ran into problems. In both France and the Netherlands voters turned the new treaty down in national referenda, which put an effective stop to the treaty revision process. The failed treaty represented a victory for the European Parliament. Its long-sought equal status with the Council in delegated decision-making was almost achieved. It was given the same control mechanisms as the Council in relation to the new delegated acts, not least the call-back right it had wanted in the financial area. The reform pressures generated by the treaty revision were threefold. First, the European Parliament’s status was high on the agenda. It was not likely to give up the position and control mechanisms granted in the Constitutional Treaty. It had

What is at Stake? A Case Study of the 2006 Comitology Reform

83

once again learnt that reforms in the comitology area can only go one way, namely towards strengthening the Parliament. The European Parliament was likely to take pains to ensure that the provisions in the Constitutional Treaty would function as a reference point in any future comitology reform process. Second and closely related, the issues of sunset clauses lived on. This control mechanism had been discarded by the Convention and the Constitutional Treaty, but given the treaty’s fate the European Parliament was free again to use them. Third and finally, the Constitutional Treaty took a new approach to the age-old distinction between politics and administration, which had also hampered the Lamfalussy negotiations. The new three-tiered hierarchy of acts represented a new drawing line, and one which left more of the grey zone within the influence of the European Parliament. In terms of institutional preferences, the process leading to the Constitutional Treaty is most revealing about the European Parliament and the Commission. For the Parliament, the process confirms once again that it is deeply concerned about its position in delegated decision-making. It therefore had no hesitations about rolling back the comitology system and replacing it with other control mechanisms that strengthened its position. As to the Commission we learn that it cares about its autonomy. It favours delegation in the EU system and a clearer separation of legislative and executive powers. These measures would, if implemented, solidify the position of the Commission.

Process 4: The revision of the 1999 comitology decision The first step towards changing the 1999 comitology decision was taken in 2001 when the Commission (2001a) presented its White Paper on European governance. The Commission argued for clarifying executive responsibility by reconsidering the comitology system. It suggested that management and regulatory committees might be abolished (which would leave only advisory committees) and that the Council and the European Parliament ought to have an equal role in the supervision of the Commission. It foresaw both a short-term solution involving a change of the 1999 comitology decision and a long-term solution involving a change of the treaty’s framework. The Commission’s next move was a proposal to amend the 1999 comitology decision (Commission, 2002b; see also Bergström, 2005, pp. 326–35). This was to be a limited reform pending the more fundamental comitology reform anticipated in the treaty revision process. Although presented as an endeavour to put the European Parliament

84

The EU Comitology System in Theory and Practice

and the Council on an equal footing as legislative supervisors, the proposal strengthened the Commission’s position. The idea was to create two comitology systems – a new system for matters under codecision, and another (the existing) for other matters. The new system comprised several new features. First, the regulatory procedure was to be amended to comprise two distinct phases. In the first (executive) phase the Commission would present draft rules to the committee as under the existing system. In the second (supervisory) phase the draft would be forwarded to the Council and the European Parliament. If they objected to the draft, the Commission must withdraw it, present a new legislative proposal under the codecision procedure or adopt the draft ‘possibly’ amending it according to the objections of the European Parliament and the Council (Commission, 2002b, article 5a). While this new supervisory phase put the Council and the European Parliament on the promised equal footing, it also strengthened the Commission. It was no longer bound to follow the legislators’ objections – only to ‘possibly’ take them into account. Second, the regulatory procedure would be used in relation to measures designed to ‘widely implement the essential aspects of the basic instrument’ (Commission, 2002b, article 2a). This room for application was narrower than under the 1999 decision. Given the fact that the Commission also proposed abandoning the management procedure for matters under codecision, the proposal also strengthened the Commission since it left only the advisory procedure for all remaining matters. Finally, the Commission proposed more binding rules for the legislators’ choice of comitology procedure than under the 1999 decision. This would also strengthen the Commission since the Council would then find it more difficult to insert comitology clauses wherever convenient. In the European Parliament the proposal was assigned to the Committee on Constitutional Affairs. The committee’s first report, ready in April 2003, was highly positive, but raised four objections (European Parliament, 2003a). First and most importantly, it found it unacceptable that the Commission would be entitled to adopt draft measures even in case of disagreement with the legislature. This meant that the Commission’s option to only ‘possibly’ amend drafts in case of objections from the Council or the European Parliament had to be changed. Second, the committee wanted to uphold the Parliament’s existing callback right in relation to the scope of matters falling under the advisory procedure. This right would be abandoned as a consequence of the Commission’s new ‘supervisory’ phase which only comprised the regulatory procedure. The choice between the regulatory and advisory

What is at Stake? A Case Study of the 2006 Comitology Reform

85

procedure would, according to the Commission proposal, depend on the proposed measure having either an individual or a more general scope. As the committee observed, this distinction ‘can be subject to different interpretation’ (European Parliament, 2003a, p. 10). The problem of distinguishing between policy formulation and implementation popped up yet again. Third, the committee wanted all time limits for the Parliament to consider Commission proposals extended from one (as proposed by the Commission) to three months. Finally, the committee wanted access to more comitology documents than the Commission suggested. In sum, the committee welcomed the Commission’s proposal to give the Council and the European Parliament equal status as legislative supervisors, but it was not willing to give up anything in return. The report by the Committee on Constitutional Affairs was discussed by the European Parliament in May 2003. A representative from the Commission, Commissioner Margot Wallström, participated in the debate. She made it clear that the Commission would be willing to discuss many of the objections raised by the committee, but she was not willing to make it obligatory for the Commission to follow the objections of the Council or the Parliament to draft measures. Arguing in terms strikingly similar to those used by the Commission 40 years earlier – cited in the previous chapter – when the Council installed comitology in the area of development aid to Africa, she explained that the Parliament’s proposal would ‘tie the hands of the Commission and prevent it from exercising its responsibility as the executive’. She also reminded the assembly that ‘the Commission needs Parliament’s support in negotiating the strengthening of the European Parliament’s prerogatives with the Council’ (Wallström, 2003). Given the Commissioner’s partially positive attitude, the European Parliament sent the matter back to the Committee on Constitutional Affairs to pursue further negotiations with the Commission. Two months later the committee had a second report ready (European Parliament, 2003b). The rapporteur’s negotiations with the Commission had produced a compromise, which stipulated that in case objections were raised by the Council or the European Parliament, the Commission might adopt the measure ‘taking account’ of these objections. However, the exact meaning of this compromise was unclear. When the European Parliament voted on the report in September 2003, the rapporteur Richard Corbett (2003) explained that ‘if Parliament objects to an implementing measure adopted under comitology, then the Commission must either withdraw it, amend it or take account of Parliament’s

86

The EU Comitology System in Theory and Practice

objection in some other way . . . ’. He stressed that this result might not be ideal, but should be considered a provisional compromise until the draft Constitutional Treaty just completed by the European Convention would come into force. The Commission (2004) presented its interpretation of the compromise in April 2004 when it presented an amended proposal to amend the 1999 comitology decision. As promised, the Commission had accepted many of the European Parliament’s objections. But on the important question of the consequences of objections from the Parliament or the Council, the Commission’s interpretation differed somewhat from that of the Parliament’s rapporteur. The Commission maintained that it was to be entitled to adopt its draft measures without changes even in case of objections from the legislators. The only concession from the Commission was that it was to inform the legislators of its intended actions. The Commission (2004, p. 3) explained its reasons in blunt terms: ‘The Commission must be able to apply its executive responsibility autonomously and therefore cannot be bound by the legislature’s position.’ Meanwhile the member states had kept a low profile and not taken any official position on the Commission’s proposal. They stalled it in anticipation of, first, the European Convention’s treaty proposal and, later, the ratification of the Constitutional Treaty. However, when the Constitutional Treaty failed in France and the Netherlands in 2005, the European Parliament put pressure on the Council to open negotiations. This was done by a range of well-known means (cf. Bradley, 1997). First, by using its budgetary authority it withheld funding of comitology committees. After the first quarter of 2006 the Parliament’s budget committee only released means for committee meetings gradually, awaiting progress in the comitology negotiations (Danish Foreign Ministry, 2006, p. 56; European Parliament, 2006a, p. 8). Second, it continued its practice of inserting sunset clauses to limit delegation in new legislative acts. But it extended this technique to encompass threats of not renewing existing sunset clauses that were about to expire (Danish Foreign Ministry, 2006, p. 55). Third, it blocked a number of internal market directives underway in the legislative process (Danish Foreign Ministry, 2005, p. 7; Schusterschitz and Kotz, 2007, p. 76). The tactics worked. In the autumn of 2005, the Council finally opened negotiations with the European Parliament. According to information from the Danish government, there was ‘widespread scepticism about weakening the Council’s control mechanisms’ among the member states, but ‘a certain understanding for increasing the European Parliament’s control position’ (Danish Foreign Ministry, 2005, p. 7). The

What is at Stake? A Case Study of the 2006 Comitology Reform

87

Council consequently decided to disregard the Commission’s proposal and aim for a limited reform that only comprised the elements necessary for pleasing the Parliament. The final compromise consisted of an amended comitology decision (Council of Ministers, 2006b), a joint inter-institutional statement on the new decision – the so-called ceasefire declaration (European Parliament, Council and Commission, 2006) – and a set of statements from the Commission (2006a). The compromise included several novel features. First, it introduced the new regulatory procedure with scrutiny, which adds a fourth procedure to the existing advisory, management and regulatory procedures. The new procedure applies only to acts adopted under the codecision procedure and consists of an executive and a supervisory phase. In the executive phase the Commission submits its draft measure to the relevant committee, which states its opinion as usual. In the following supervisory phase the draft measure is submitted to the Council and the European Parliament, irrespective of the committee’s opinion. If the committee’s opinion is positive, both the European Parliament and the Council can object to the draft, in which case the Commission must abandon it. If the committee gives a negative or no opinion, the draft is submitted to the Council for decision and simultaneously sent to the Parliament for information. If the Council objects to the draft within two months, the Commission must abandon it. If the Council does not object, the Parliament then has four months to oppose the draft in which case the Commission must abandon it. This means that the Council and the Parliament can both revoke draft measures, although they are not on a completely equal footing. In any case, the Parliament finally gained the call-back right it had demanded for years, a result it considered an ‘almost revolutionary achievement’ (European Parliament, 2006b, p. 8). Second, the compromise took issue with the grey zone between policy formulation and implementation, or legislative and executive matters. The new procedure was to be applied to matters closer to the legislative end of this spectrum. The new comitology decision specifies that ‘measures of general scope’ must be subjected to the new procedure. This adds a compulsory element to the choice of procedure that did not exist under the 1999 decision, which only specified criteria that should ‘guide’ the choice of comitology procedure. Third, the European Parliament’s new call-back right came at a price. In the joint inter-institutional statement it agreed to end the use of sunset clauses, which had become a major irritant to both the Council and the Commission.

88

The EU Comitology System in Theory and Practice

Fourth, the introduction of the new procedure naturally raised questions about its possible retrospective application to existing legislative acts. Here, the compromise was fine-tuned. The joint inter-institutional statement specified 25 existing directives and regulations which should be subjected to the new procedure. The list included a number of acts in the financial area (Lamfalussy acts) and the GMO area discussed above. Finally, the Commission (2006a) promised to improve the flow of information to the European Parliament, especially in the financial area. This was a compromise on the Parliament’s wish to have observers present at the meetings of comitology committees. This wish, reported by both Schusterschitz and Kotz (2007, p. 85) from the Austrian Council Presidency and Szapiro (2006, p. 580) from the Commission, was flatly refused by the Council and the Commission. In terms of institutional preferences, the process leading to the revision of the 1999 comitology decision is revealing about all EU actors. Concerning the Commission, we once again learn that it works to strengthen its autonomy. Quite ingeniously, its initial proposal tried to strengthen both its own and the European Parliament’s position. As to the European Parliament, the process confirms that it considers comitology a threat to its institutional position in the EU system, and that it is willing to go to considerable lengths to defend and strengthen its position. Finally, the member states once again consider comitology a control instrument and are reluctant to relax or share control.

Lessons of the case study: Institutional preferences, interests and instruments revealed The 2006 reform of the comitology system is the result of developments in four areas. First, in the area of financial services the Lamfalussy reform and the European Parliament’s subsequent use of sunset clauses generated a strong pressure for increased legislative supervision by the European Parliament. Second, in the sensitive areas of GMOs, food safety and the environment the member states discovered that comitology does not always constrain the Commission. This generated widespread aversion to relaxing control. Third, the European Convention’s proposal for a Constitutional Treaty strengthened the position of the European Parliament in the comitology area. At the ensuing intergovernmental conference, comitology was drowned by other issues, and the member states accepted the Convention’s proposal. Although the ratification of the Constitutional Treaty failed, this set a powerful reference point for the negotiations in the fourth process, the negotiations

What is at Stake? A Case Study of the 2006 Comitology Reform

89

on amending the 1999 comitology decision. In sum, the 2006 reform can be considered the result of a game played in several arenas over a number of years. To conclude the case study, I return to the three hypotheses on comitology preferences developed in Chapter 3 and restated in the beginning of this chapter. The first hypothesis held that the Council favours strict comitology procedures, and the case study firmly supports this hypothesis. In all four processes the Council was sceptical about loosening control. In the Lamfalussy reform it worried about delegating far-reaching powers without effective control by comitology procedures. In the sensitive areas of GMO, food safety and the environment it developed a strong scepticism towards any proposal about loosening comitology control. In the process leading to the Constitutional Treaty the evidence is less clear, but the intergovernmental conference adopted a declaration on the Lamfalussy committees which indicates that strict comitology was a concern. Finally, when negotiating the revision of the 1999 comitology decision the Council refused to discuss proposals to reduce the strictness of comitology procedures. The second hypothesis held that the Commission favours permissive or no comitology procedures. Despite somewhat mixed evidence, the case study cannot refute this hypothesis. The clearest supporting evidence is found in the fourth process. In its proposal for a revised 1999 comitology decision the Commission argued for much broader use of the advisory procedure. But the case study also shows that the Commission sometimes accepts strict comitology procedures as a price for increased delegation. The Lamfalussy reform and its regulatory committees were welcomed by the Commission because it entailed more delegation. There is thus also evidence of pragmatism on the Commission’s behalf. The third hypothesis held that the European Parliament favours: (a) narrowing delegation to technical matters; (b) permissive or no comitology procedures; or, (c) access to the comitology system. This hypothesis is also difficult to refute, but the evidence is clearest on points (a) and (c). In the Lamfalussy reform the Parliament advocated less delegation than wanted by the Council and the Commission. In the processes leading to the Constitutional Treaty and the revised 1999 comitology decision the Parliament argued strongly for increased access to the comitology system. In sum, the chapter provides relatively strong evidence in favour of the three hypotheses on the EU institutions’ comitology preferences. But this evidential support should not be exaggerated because, as argued

90

The EU Comitology System in Theory and Practice

in the introduction to the chapter, the 2006 comitology reform represents a most likely case for investigating the hypotheses. The next chapter examines the hypotheses in a less favourable setting. Before turning to that analysis, however, the full benefit of the case study format should be derived. What does the study of the 2006 comitology reform reveal about interests, instruments and institutional preferences in the comitology area in a broader sense? First, in terms of institutional interests, the case study makes it possible to flesh out in some detail what this concept means in practice. Table 5.1 provides a list of contested institutional features in the 2006 reform. It shows that the actors’ positions in the comitology system depend on a number

Table 5.1 What is at stake? Manifestations of institutional interests in the 2006 reform of the comitology system The following institutional features of the comitology system were contested in the 2006 reform (examples in parentheses): 1.

2.

3.

4.

5.

The treaty foundation of the comitology system (EC Treaty article 202): Divides power over the specification of comitology procedures unevenly among the EU institutions (in the Constitutional Treaty the European Parliament wanted a new provision on comitology because the existing one only gave decision-making power to the Council). The boundary between legislative and executive decision-making: Since the position of the EU institutions varies between legislative and executive rulemaking, the Council can increase its influence by shifting decision-making from codecision to comitology (in reaction the European Parliament tries to keep its legislative influence by setting time limits on delegated powers, arguing for call-back rights or by changing the treaty). Formal competence of the comitology committees: Divides power unevenly between the Council and the Commission (in the Commission’s proposal for a revised 1999 comitology decision it proposed increased use of the advisory procedure; the affected committees would lose all formal influence on Commission proposals for implementing measures). Voting rules in the comitology committees: Divide power unevenly between the Council and the Commission (in the Commission’s proposal for a revised 1999 comitology decision it proposed the abandonment of the management procedure for matters under codecision; this would increase the autonomy of the Commission because most committees would then operate under the advisory procedure). Voting rules in the Council of Ministers for matters referred to it from comitology committees: Specify how difficult it is for the member states to block Commission proposals (when revising the 1999 comitology decision some member states, especially Denmark, wanted to re-introduce the option of rejecting Commission proposals by simple majority since this would make it easier to block Commission proposals).

What is at Stake? A Case Study of the 2006 Comitology Reform

91

6. Time limits: Short time limits to react to Commission proposals may reduce the real importance of formal influence for both the Council and the European Parliament (in the Commission’s proposal for a revised 1999 comitology decision the European Parliament was only given one month to raise objections; the Parliament wanted three months). 7. Information: The comitology rules specify how much information the European Parliament is allowed (according to the Commission’s proposal for a revised 1999 comitology decision the European Parliament was to receive agendas for committee meetings, draft implementation measures, voting results and summary records of meetings. The Parliament wanted this list extended to include revised or amended draft implementation measures). 8. Call-back provisions: A means for the legislative actors to keep influence on delegated powers (in the Lamfalussy area the European Parliament, lacking influence in the comitology committees, pressed for call-back provisions; when this was refused it inserted sunset clauses in the basic acts as an alternative). 9. Composition of comitology committees: Having participants in the comitology committees is a way to control executive rule-making (when negotiating the 2006 decision the European Parliament wanted to introduce parliamentary observers in the comitology committees; the Council and the Commission refused). 10. Guidelines for the use of comitology procedure in daily legislation: May limit or expand the use of the comitology system (the European Parliament succeeded in making the use of the new regulatory procedure with scrutiny compulsory; this is in contrast to the non-binding guidelines under the 1999 comitology decision). 11. Retrospective application of new procedures: Can be a way to increase (or limit) the importance of reforms (the joint inter-institutional statement on the new 2006 comitology decision specified 25 directives and regulations that should be changed to include the new regulatory procedure with scrutiny. The European Parliament wanted the list to be longer and, in addition, a full scrutiny of all acts adopted under codecision to see if adaptation to the new procedure was relevant. The Commission accepted the latter wish, and made a statement to this effect, but the Council refused any further commitments).

of institutional rules. The case study demonstrates that the actors are keenly aware of this and ready to fight for their position. At the general level, the endless comitology negotiations constitute a two-dimensional constitutional struggle. The first dimension concerns the relative position of the two legislative actors, the Council and the European Parliament. Should their supervisory position be equal? In hindsight it is clear that the Council was fighting a losing battle to keep its privileged position. The 2006 comitology reform was a further step in the process towards full equality that was completed with the

92

The EU Comitology System in Theory and Practice

Lisbon Treaty. The second dimension concerns the relationship between the legislative and the executive branch of the EU system. How much autonomy should the Commission be allowed? While the comitology system is uncontroversial in many policy fields – for example, agriculture – it is evident that many areas are sensitive in the eyes of the member states, and that this fact constitutes a serious barrier against a general weakening of legislative control mechanisms. In addition, the case study makes it possible to specify in concrete terms how the actors pursue their institutional interests. Table 5.2 provides a list of instruments used by the actors in the 2006 reform. It is evident that the actors have a range of instruments available, and that they use them in imaginative ways. Table 5.2 Instruments to pursue institutional interests in the 2006 reform of the comitology system The following instruments were used to pursue institutional interests in the 2006 reform (examples in parentheses): 1. Sunset clauses: May limit delegation (in the area of financial regulation the European Parliment inserted time limits in provisions delegating powers in order to pressure the Council to accept a general call-back right). 2. Money: The European Parliament’s budget authority may be used to impede the practical operations of the comitology system (when revising the 1999 comitology decision the European Parliament withheld funding for comitology committees in order to put pressure on the Council to find a compromise). 3. Time: The most patient legislator can employ delaying tactics (in 2002, the Council stalled the Commission’s proposal to amend the 1999 comitology decision for three years; in 2005–6 the European Parliament blocked a number of internal market directives in order to put pressure on the Council to compromise in the comitology area). 4. Court proceedings: Can be used to force actors to change their use of comitology (in 2006, a number of member states challenged the Commission before the European Court of Justice in order to change the use of comitology under the RoHS directive). 5. Safeguard mechanisms: Can be used for unilateral non-compliance with comitology decisions (a number of member states invoke the GMO directive’s safeguard clause in order to avoid the decisions coming out of the comitology system in this area). 6. Right of initiative: Can be used to time the initiation of reform procedures (the Commission used its right of initiative to decide when to open the formal procedure for revising the 1999 comitology decision. Care was taken to time this so that the proposal followed in the wake of the White Paper on European governance).

What is at Stake? A Case Study of the 2006 Comitology Reform

93

This is not least true for the European Parliament, which helps explain why the Parliament wielded more influence than the pre-Lisbon Treaty formally allowed. Especially the daring use of sunset clauses (and, to a lesser degree, its budgetary authority) made it possible for the Parliament to prevent decisions from slipping from the codecision to the comitology area and to win important victories in terms of more lenient time limits, amounts of information, call-back rights and, of course, the introduction of the new regulatory procedure with scrutiny, which finally put it on an equal footing with the Council in selected areas of delegated decision-making. However, the Council also has effective instruments, so the Parliament’s victory was not complete. Until the Lisbon Treaty it was favoured by the status quo because it possessed the power to decide the list of comitology procedures. Furthermore, it can employ delaying tactics to avoid or delay changes. Finally, the Council’s silent approval of unilateral non-compliance by individual member states, especially in the GMO area, puts pressure on the Commission to use delegated power carefully. The use of these instruments made it possible for the Council for many years to resist the Parliament’s call for full equal power over the Commission’s delegated powers. It has successfully limited the use of the new regulatory procedure with scrutiny and avoided its full retrospective application. It has also successfully rejected the Parliament’s wish for observers in the comitology committees. In conclusion, the chapter’s analysis supports hypotheses 4, 5 and 6. It shows that the actors behind the comitology system hold consistent preferences over this system and actively pursue them in instrumental ways. It is obvious that the actors are not unified in a search for a technically rational system. On the contrary, they have a keen eye on their institutional interests and view the shaping of the comitology system through constitutional lenses. The system can best be understood as a compromise between actors pursuing positions of institutional control. Both winners and losers have influence on the design of the comitology system’s administrative procedures. The result is a system that is not efficient from any single actor’s point of view. It is no wonder that the comitology system appears mysterious and bizarre to outside observers.

6 Comitology Preferences in Daily Legislation

Every year the Council and the European Parliament enact hundreds of regulations and directives, and about half of them contain a comitology provision. In these cases, the EU legislators delegate executive power to the Commission, but require it to follow a comitology procedure when using this power. This chapter examines the EU institutions’ comitology preferences in these situations. Do the preferences of the Commission, the Council and the European Parliament exhibit systematic patterns in daily legislation? This is an alternative investigation of the hypotheses that were also examined in the case study of the 2006 comitology reform in the previous chapter: Hypothesis 4: The Council favours strict comitology procedures. Hypothesis 5: The Commission favours permissive or no comitology procedures. Hypothesis 6: The European Parliament favours: a. narrowing delegation to technical matters b. permissive or no comitology procedures, or alternatively c. access to the comitology system But while Chapter 5 represented a most-likely case, this chapter provides a hard test of these hypotheses. First, there are guidelines for choosing comitology procedures in daily legislation. They were first introduced by the Council’s second comitology decision in 1999. Although they are non-binding, they nevertheless introduce some order and predictability. If the legislators want to introduce a comitology procedure that runs 94

Comitology Preferences in Daily Legislation

95

against the guidelines, it is legitimate for other actors to challenge this. In this sense the choice of comitology procedure in daily legislation is not completely free. Second, a high degree of open conflict over the choice of comitology procedure in daily legislation is not to be expected. The EU institutions handle hundreds of legislative proposals every year and they have considerable experience with each other’s preferences. Some routinization is likely. In cases where the actors know that their favoured comitology procedure has no chance of surviving the legislative process, they are unlikely to press their case. They are more likely to be strategic and only propose their favourite comitology procedure where it has a reasonable chance of being accepted by the other actors. This raises a tricky problem of anticipated reactions. Both the informal guidelines and the problem of anticipated reactions will be treated in greater detail later in the chapter. The point is that they make this empirical setting a least-likely case for the hypotheses. Studying comitology preferences in daily legislation is also difficult because the object of study – the comitology procedures – changes over time. The list of procedures to choose from in daily legislation was specified by the Council’s comitology decisions until the Lisbon Treaty and by the European Parliament and the Council’s comitology regulation after the Lisbon Treaty. The 1987 comitology decision specified four different procedures and several variants. The 1999 decision streamlined the comitology procedures to four standard ones. In 2006, a new fifth procedure was introduced when the Council established the regulatory procedure with scrutiny that was examined in the previous chapter. In 2011, the European Parliament and Council’s comitology regulation merged the management and regulatory procedures into variants of the new examination procedure. To keep a firm grip on the dependent variable, the analyses in this chapter focus on legislative practice under a single comitology decision, namely the 1999 decision. Under the seven-year rule of this decision the Council acting alone or the Council and the European Parliament acting together enacted a total of 686 new regulations and directives. They have all been coded to allow an analysis of the chapter’s hypotheses. The chapter is organized in the following way. Since my analysis is not the first to investigate comitology preferences in daily legislative practice, I start by reviewing earlier studies. I then explain in greater detail why I focus on legislative practice under the 1999 comitology decision. This is followed by a discussion of the two factors that make such a study a hard case for the hypotheses: the informal guidelines for the choice of comitology procedures and the problem of anticipated reactions. I then

96

The EU Comitology System in Theory and Practice

explain my data and methods before starting the empirical analyses. The chapter closes with a discussion of the combined lessons of the analyses in this and the previous chapter.

Earlier studies of the choice of comitology procedures in daily legislation Three studies have directly addressed the question of comitology preferences in daily legislation. The first was made by Dogan (1997; see also 2000), who investigated all EU legislation in the period 1987–1995, a total of 4601 acts. He found consistent evidence that the Commission and the European Parliament favour permissive or no comitology procedures, while the Council favours restrictive comitology procedures. Classifying the procedures I, IIa and IVa (under the 1987 comitology decision) as ‘executive autonomy’ and the procedures IIb, IIIa, IIIb and IVb as ‘executive dependence’, he found that the Commission proposed executive autonomy in 71 per cent of all instances involving new comitology procedures. The European Parliament favoured executive autonomy in 81 per cent of these instances – that is, it was willing to grant the Commission more autonomy than it asked for. However, the Council only allowed executive autonomy in 36 per cent of the cases and imposed executive dependence in the remaining cases. This pattern of institutional preferences supports the chapter’s hypotheses. This chapter will investigate whether it still holds. The second study is by Franchino (2007, pp. 282–5), who investigated the European Parliament’s comitology preferences. Based on a content analysis of the European Parliament’s amendments to 369 Commission proposals under the codecision procedure he found that the Parliament wishes to strengthen its own role either by greater involvement in the comitology procedures or by transferring issues out of the comitology system and back into the ordinary legislative process. At the same time the Parliament often seeks to introduce a more permissive comitology procedure than proposed by the Commission, a finding that echoes the pattern identified earlier by Dogan and supports this chapter’s hypotheses. The chapter further investigates the institutional preferences of the European Parliament both within and outside the areas covered by the codecision procedure. The third study is by Heritier and Moury (2009), who examined the temporal dynamics of comitology preferences. Based on 102 legislative acts within one issue area – environmental politics – across 14 years (1994–2008) they investigated the preferences for the regulatory

Comitology Preferences in Daily Legislation

97

comitology procedure. In a second analysis they investigated the European Parliament’s comitology preferences based on 315 parliamentary reports under the codecision procedure. They found that the Commission and the Council increasingly agree on delegation combined with the regulatory comitology procedure. They interpret this as an expression of strategic, not sincere, preferences from the Commission and argue that it shows that the Commission is willing to compromise on comitology in order to increase delegation. In essence, they argue that, at least for the Commission, the data cannot be relied upon as indicators of sincere preferences. This underscores the methodological challenge of anticipated reactions when investigating comitology preferences. They also found that the regulatory procedure is increasingly accepted by the European Parliament, but that it tries to restrict the scope of delegation, at least until the 2006 comitology decision. This finding supports the hypotheses. This chapter will probe further into the problem of anticipated reactions and investigate the institutional preferences of the European Parliament both within and outside the areas covered by the codecision procedure. In sum, the available evidence, although scattered, largely supports this book’s hypotheses. The following analysis will try to reach a more conclusive answer by using a data set that is more comprehensive in the sense that it includes all legislative actors and all EU decision-making procedures. The results of the analysis support the chapter’s hypotheses, but also show an interesting development over time that underscores the methodological problem of disentangling sincere and strategic preferences. But first I need to argue why the chapter focuses on legislative practice under a single comitology decision.

Why focus on legislative practice under the 1999 comitology decision? To have sufficient cases for a quantitative study this chapter’s analyses focus on legislative practice in the pre-Lisbon period, where legislation took place under three comitology regimes – the 1987, 1999 and 2006 comitology decisions. Although these decisions use the same labels for the various comitology procedures – advisory, management, regulatory, safeguard – the procedures vary considerably across decisions. For this reason, the following analysis focuses on legislative practice under a single decision. This allows a firm grip of the object under study: the comitology preferences of the EU institutions during the legislative process.

98

The EU Comitology System in Theory and Practice

The 1987 comitology decision was the first attempt to rationalize and introduce more consistency and predictability in the large number of implementation committees and committee procedures used at the time. But this decision still specified four different procedures and several variants. The 1999 decision simplified matters further and streamlined the comitology procedures into four standard ones (Haibach, 1999; Bergström, 2005, pp. 189–209, 249–85). In 2006, a new fifth procedure was introduced when the Council established the regulatory procedure with scrutiny that was examined in the previous chapter. As to the individual comitology procedures, only the advisory procedure has been constant over time. The management procedure underwent considerable change in 1999 when the variants were removed and replaced by a new management procedure that allowed the Commission to adopt measures even in the case of a negative opinion of the committee. The regulatory procedure underwent even greater change in 1999 as the Council’s option to reject a Commission proposal by simple majority was abolished. In 2006, the regulatory procedure underwent another profound change when the new variant – the regulatory procedure with scrutiny – was introduced. Finally, the less frequently used safeguard procedure has also changed over time. For a more precise overview of the historical development of the individual comitology procedures see Chapter 2, especially Table 2.2. Also the position of the European Parliament has changed considerably under the three comitology decisions. The 1987 decision allowed the Parliament no influence, whereas the 1999 decision provided for a limited involvement of the Parliament in the implementation of acts adopted under the codecision procedure. The Parliament was to receive all relevant information and to exercise some control of ultra vires behaviour by the Commission. But this was a weak control mechanism. If the Parliament found that the Commission’s proposal exceeded its delegated powers, it might adopt a resolution. The Commission then had to take this resolution into account, but was still allowed to continue with the procedure. In 2006, the Parliament was finally given powers equal to those of the Council, but only in relation to comitology working under the new regulatory procedure with scrutiny. In sum, apart from their labels the comitology procedures have undergone considerable change over time. In order to allow an unequivocal investigation of comitology preferences, the following analysis focuses on choices under just one comitology decision, the 1999 decision. The different comitology procedures constrain the Commission to various degrees. Under the 1999 decision, the advisory procedure is

Comitology Preferences in Daily Legislation

99

obviously the least constraining since the Commission is not obliged to follow the committee’s opinion. The safeguard procedure is obviously the most constraining since each member state’s representative may veto the Commission’s proposals. Among the two intermediary procedures, the regulatory procedure constrains the Commission more than the management procedure. According to the latter procedure, the Commission’s proposal can only be referred to the Council if a qualified majority can be mustered in the committee, while the former procedure only requires a blocking minority. Formal game theoretical analyses confirm this rank order of the comitology procedures (Steunenberg et al., 1996, 1997; Ballman et al., 2002). In terms of constraining the Commission the comitology procedures can be summarized in this way: Advisory procedure < management procedure < safeguard procedure

procedure

<

regulatory

This rank order will guide the analysis of the EU institutions’ preferences for permissive or strict comitology in this chapter.

Hard case argument I: Guidelines for the choice of comitology procedures If there were binding guidelines for the choice of comitology procedures when the Council and the European Parliament delegate executive power to the Commission, it would make no sense to use daily legislative practice to study comitology preferences since there would be no scope for disagreement. If daily legislative practice is to function as an empirical testing ground, it presupposes that the EU actors have some freedom to choose among the comitology procedures from the Council’s list of procedures. However, the 1999 comitology decision introduced guidelines for the choice of comitology procedures in individual legislative acts. The decision’s article 2 states that the management procedure ‘should’ be used within the areas of the common agricultural and fisheries policies and in relation to programmes with substantial budgetary implications. The regulatory procedure ‘should’ be used in relation to measure of general scope within the areas of health or safety of humans, animals or plants. The advisory procedure is to be used ‘in any case in which it is considered to be the most appropriate’. However, as the wording signals, these guidelines are informal. This is even stressed in the fifth consideration of

100 The EU Comitology System in Theory and Practice

the 1999 decision, which explicitly states that they are of a ‘non-binding nature’ (Council of Ministers, 1999a). The establishment of non-binding criteria was a compromise in the negotiations leading to the 1999 decision. The Commission and the European Parliament wanted binding guidelines in order to increase transparency, but the Council insisted on some freedom of manoeuvrability in daily legislation (Bergström, 2005, pp. 272–4). The compromise was aptly summarized by the Council in a press statement after its meeting on 21–22 June 1999 where it made the final decision on the new comitology rules: ‘The decision provides for criteria which will guide the legislator in the choice of committee procedure while allowing for a certain degree of flexibility’ (Council of Ministers, 1999b, p. 11; emphasis added). The non-binding nature of the guidelines has been upheld by the European Court of Justice. Nevertheless, the Court has found that their existence means that the legislators must explicitly state their reasons if they decide not to follow them (Türk, 2009, pp. 80–1). In sum, although the 1999 decision took steps to specify the choice of comitology procedures in daily legislative practice, it still left the legislators considerable freedom to decide this question on an ad hoc basis. The guidelines probably constrain the choice to some degree and thus make it less straightforward for the actors to pursue their comitology preferences. In this sense the guidelines make daily legislative practice a hard case for studying comitology preferences.

Hard case argument II: Anticipated reactions Several observers have noted that especially the Commission is strategic in its choice of comitology procedure in daily legislation. Rather than following its true preference, it proposes a procedure that is acceptable to the Council (e.g., Hix, 2005, p. 57; Heritier and Moury, 2009). This raises a problem of anticipated reactions, which makes the study of comitology preferences in daily legislation tricky. The problem is well-known from studies of the EU legislative process. The general insight is that under complete information there will be no open conflict. All objections from the European Parliament and the Council that are likely to succeed will be anticipated by the Commission. If they are acceptable to the Commission, they will be incorporated into its first proposal. If they are not acceptable, they will never surface because the Commission will use its right of initiative and not make a proposal. Objections that are not likely to succeed will never be raised by the Parliament or the Council since they realize the futility of doing

Comitology Preferences in Daily Legislation

101

so. Consequently, under complete information there will be no amendments, successful or unsuccessful, to the Commission’s proposals. All actors will accept the Commission’s initial proposal, and the legislative process will be quick and uneventful. This is the logic of backwards induction, and it has been confirmed by a number of complete information models of the EU legislative process (Steunenberg, 1994; Crombez, 1996; Moser, 1997). For instance, Steunenberg (1994, p. 663) concludes on the Commission’s right of initiative that this is a powerful weapon because it enables the Commission to prevent unfavourable changes from the status quo. The Commission ‘only initiates a legislative process if it leads to an outcome that is preferable to the status quo’. Otherwise, ‘it will keep its gates closed’, and not take any initiative. Similarly, Crombez (1996, p. 219) concludes on the cooperation procedure which gives the Parliament veto power: ‘In equilibrium the Parliament does not exercise its veto because the Commission proposes a policy the Parliament prefers to the status quo.’ Tsebelis (2002, p. 259) has aptly summed up the core prediction of complete information models: If all the actors knew each other’s preferences and payoffs . . . this way of thinking would lead to an immediate end of the legislative game: the Commission would propose a bill that would be accepted by all other actors. Indeed, the Commission would never make a proposal that would be ultimately rejected, and the other players would not raise objections if they knew they would not win in a confrontation. As a prediction of the real-world EU legislative process this scenario is, of course, far from the mark. The process abounds with amendments, even in the final stages – for example, second readings or even conciliation committees under the codecision procedure. Complete information theorists have offered a number of reasons. For instance, Moser (1997) suggests that successful Parliament amendments are possible if, first, the decisive member country in the Council of Ministers changes its position during the legislative process (e.g., because of a change of party in government); second, if the preferences of the Commission change (e.g., because the responsible Commissioner is replaced); or, third, if the perception of the reversion point changes (e.g., the possibility of unilateral action by individual member states). Amendments for these reasons cannot be ruled out, but Tsebelis (1996, 2002, pp. 259–60; Tsebelis and Garrett, 2000, p. 17; Tsebelis

102 The EU Comitology System in Theory and Practice

et al., 2001, pp. 576–7) has repeatedly advocated a more convincing reason, namely that information is not complete in the beginning of the process, but only later when the actors have exchanged considerable information. In the early stages of the process the actors may not precisely know each other’s preferences, but once they are revealed through amendments, they will be incorporated into the proposal or rejected. In addition to this systematic reason, Tsebelis has suggested other more ad hoc reasons for the existence of amendments: the Parliament may try to blackmail the Commission or the Council by threatening to delay legislation; it may hope that the Commission will live up to its role as an honest broker; or parliamentarians or individual member states may use amendments to send signals to voters or other actors in their home country. Although persuasive, Tsebelis’ arguments reduce, rather than eliminate, the problem of anticipated reactions. Even if information is incomplete, it is not absent. Many objections may still not reach the surface as amendments either because they are anticipated and incorporated into the first mover’s proposal or because actors realize the futility of raising them. The actual number of amendments is therefore likely to be lower than what would have been the case if actors used amendments to express their sincere preferences. This means that amendments should be used with caution as an indicator of the extent of true disagreement in legislative processes. However, if the objective is to study differences in preferences rather than the extent of disagreement, the problem is mainly one of missing data. Due to anticipated reactions, actors may not reveal their sincere preferences. A high degree of apparent agreement may be misleading. It may be due to sincere agreement, but it may also cover a state of affairs where actors accept the futility of airing disagreement. There is thus a problem of observational equivalence: agreement may indicate either sincere consensus or strategic behaviour. However, if there is open disagreement – that is, amendments exist – this is likely to reveal true preferences. No matter whether amendments are due to incomplete information or symbolic politics, there is no reason to expect actors to systematically propose amendments that contradict their interests. Disagreement cases are thus much more useful as data than agreement cases. The conclusion I draw from this discussion of the problem of anticipated reactions is to disregard agreement cases and focus on disagreement cases. That is, the following analysis focuses on legislative cases where the Commission, the Council and the Parliament openly disagree

Comitology Preferences in Daily Legislation

103

on the choice of comitology procedure. This is an important conclusion since, as we shall see, there are many more cases of agreement than disagreement. Furthermore, proposals and amendments are not used as indicators of absolute but rather of relative preferences. Actors are not expected to systematically express preferences that have no chance of surviving the legislative process, but to try to direct the outcome towards their true preferences.

Data and methods The data to investigate comitology preferences in daily legislation are legal acts adopted by the Council (acting alone) or the Council and the European Parliament (acting together) under the regime of the 1999 comitology decision. This decision entered into force on 18 July 1999, and a new regime started on 23 July 2006 when the 2006 comitology reform amended it. The 1999 comitology decision was thus effective for seven years. I have compiled a data set containing all new directives and regulations that were proposed by the Commission and enacted by the Council/Parliament in this period. Acts that were enacted during the period of investigation but proposed by the Commission before 1999 are excluded. Likewise, acts that were proposed by the Commission during the period, but not enacted by the Council/Parliament before the end of the period, are excluded. This is to ensure that the data set only contains acts that are products of the regime of the 1999 comitology decision. To control for the status quo, regulations and directives that are pure amendments of acts made before the 1999 comitology decision are excluded from the data set. A relatively small number of acts introduce both a new act and amendments to existing acts. These partially amending acts have been included in the data set, but coded to enable control for their status. In total, the data set contains 686 regulations and directives that have been coded to allow an analysis of the chapter’s hypotheses. All information to code the acts stems from the EU’s two official online legislative databases, Eur-Lex and PreLex. Figure 6.1 shows the distribution of the data set across time and types of acts. It is clear that the years 1999–2006 witnessed many more regulations than directives. Each year between 62 and 107 regulations were produced, but only between 6 and 27 directives (excluding the two half-years 1999 and 2006). Please see Box 6.1 for further information on the data set, data sources, identification of cases and coding procedures.

104 The EU Comitology System in Theory and Practice

120

100

80

60

40

20

0 1999

2000

2001 2002 2003 2004 Year of final adoption

2005

2006

Type of act Regulation Directive Figure 6.1 Regulations and directives proposed and enacted under the 1999 comitology decision (no. of acts) Source: Data set compiled by the author (see Box 6.1).

Box 6.1 Data set on comitology in EU regulations and directives, 1999–2006

The data set contains the following EU acts: • Directives and regulations (not decisions) • Authors: Council or Council and European Parliament (not Commission, nor European Central Bank nor other EU institutions)

Comitology Preferences in Daily Legislation

105

• New acts, that is non-amending or partially amending acts (not amending or repealing acts, not recasts, codifications or corrigenda) • Period: The whole decision-making process of the acts must fall under the 1999 comitology decision (i.e., the first Commission proposal – COM document – must be made after 17 July 1999; the final act must be enacted before 24 July 2006) In total the data set contains 686 coded EU regulations and directives. All cases have been identified by reading the page of contents of all the daily issues of the L-series of the Official Journal of the EU in the period of investigation (online edition at http://eur-lex. europa.eu/JOIndex.do?ihmlang=en). All information in the data set stems from the EU’s two official online legislative databases, Eur-Lex and PreLex. Eur-Lex (available at http://eur-lex.europa.eu/en/index.htm) contains EU law dating back to 1951. For each legal document Eur-Lex provides a bibliographic notice with information on the document’s formal title, Official Journal reference, a link to the full legal document (PDF file), the Eur-Lex directory code(s), the document’s subject matter, the author and form of the document, the document’s institutional procedure number and a short description of the document’s legal history and its legal basis. PreLex (available at http://ec.europa.eu/prelex/apcnet.cfm? CL=en) follows the major stages of the EU’s legislative process. It contains information on the stage of the process, the decisions of the involved EU institutions and links to documents. The EU acts in the data set have been coded by the author (304 cases) and three research assistants (382 cases). In order to ensure intercoder reliability 15 cases were coded and discussed by all coders before the final data set was compiled. For more information on the data set, see Blom-Hansen (2010), which can be downloaded from the author’s home page.

I map the EU institutions’ comitology preferences by their statements in the legislative process. Tsebelis’ argument that information is less complete in the early stages of the legislative process implies that preferences expressed at this stage are more likely to be sincere and less likely

106 The EU Comitology System in Theory and Practice

to be strategic anticipations of other actors’ preferences. I therefore use early statements in the legislative process to the extent possible. The Commission makes the first formal proposal, may withdraw it and present an amended proposal and, under the consultation and codecision procedures, states its opinion on the Parliament’s amendments. Among these formal statements the first proposal is likely to most precisely reveal the true preference of the Commission. Later in the process the Commission is more informed on the other actors’ preferences and may anticipate them to a higher degree. Hence, the Commission’s first proposal is used to measure its comitology preference. Earlier studies have also followed this procedure (Dogan, 1997; Heritier and Moury, 2009). The Council also makes formal statements during the legislative process, although most of its deliberations are secret. However, the formal statements vary across the different types of legislative processes. Under codecision the Council adopts a common position and a final decision. Under consultation and when acting alone its only formal statement is the final decision. To measure consistently across cases I use the final decision to tap the Council’s preference. Under the codecision procedure this is an imperfect measure since the final decision may have been influenced by the legislative powers of the European Parliament. To check the robustness of results I report codecision cases separately. Earlier studies have measured the preferences of the Council in different ways. Dogan (1997) also uses final decisions across all legislative procedures. Heritier and Moury (2009), who investigate the cooperation and codecision procedures only, use both the Council’s common position and the final decision. Finally, the European Parliament is involved in the legislative process if the cases are decided under the consultation and codecision procedures. Under the consultation procedure the Parliament has one reading and, consequently, one round of amendments. Under the codecision procedure the Parliament may make amendments during its first and second reading. Formally, first reading amendments are to the Commission’s proposal, while second reading amendments are to the Council’s common position. Formally speaking, first reading amendments reveal parliamentary conflicts vis-à-vis the Commission, while second reading amendments reveal conflicts vis-à-vis the Council. In reality, first reading amendments are a manifestation of Parliament–Council conflicts to the extent that the Parliament anticipates the preferences of the Council and incorporates them into its amendments. The degree of anticipation is difficult to measure, but is likely to be of greater relevance

Comitology Preferences in Daily Legislation

107

after the Amsterdam Treaty, which made it possible to terminate the codecision process after the Parliament’s first reading. Be that as it may, second reading amendments are likely to be less revealing about the Parliament’s true preferences because they are made at a later stage in the legislative process where the Parliament is better informed of the Council’s position. Hence second reading amendments may anticipate the reactions of the Council to a higher degree and thus mask the true preferences of Parliament. I therefore use first reading amendments to measure the Parliament’s comitology preferences. Earlier studies have also mostly used first reading amendments to measure the preferences of the Parliament. This is the procedure followed by both Dogan (1997) and Franchino (2007, pp. 282–5). Heritier and Moury (2009) use first reading amendments in their analysis of environmental politics over the years 1994–2008, but both first and second reading amendments in their supplementary analysis of parliamentary reports from the same period.

Empirical analysis The mapping of the comitology preferences of the Commission, the Council and the Parliament is reported in Tables 6.1, 6.2 and 6.3 which, respectively, compare the Commission–Council, Commission– Parliament and Parliament–Council. The most striking result in Tables 6.1, 6.2 and 6.3 is the large degree of agreement between all actors. More than 90 per cent of all cases, across all actors and all decision-making procedures, show no disagreement. Table 6.1 reports codecision cases separately, but the pattern is the same. There is still agreement in the vast majority of cases. It does not seem to matter much whether the cases are decided under codecision or another legislative procedure. As argued above, agreement cases represent a problem of observational equivalence. They may indicate both sincere and strategic preferences. They are therefore indeterminate as to the chapter’s hypotheses. Before turning to the disagreement cases it is worthwhile to linger a bit on the many cases of agreement. Their large number represents a dramatic development across time. Dogan’s (1997) analysis of all comitology cases in the period 1987–95 – that is, under the 1987 decision – revealed more disagreement. It is possible to compare my data with Dogan’s on one parameter, namely the percentage of cases in new legislation where the Council’s final decision includes

108

Table 6.1 Comitology preferences, 1999–2006: Commission versus Council Comitology procedure according to first Commission proposal

Total

No comitology

Advisory procedure

Management procedure

Regulatory procedure

Safeguard procedure

Several procedures

Panel A: Limited European Parliament involvement (Council acting alone and consultation procedure) Comitology procedure according to final act No comitology Advisory procedure Management procedure Regulatory procedure Safeguard procedure Several procedures

329 3 5 0 0 0

1 3 1 1 0 2

3 0 99 1 0 1

0 0 0 19 0 0

0 0 0 0 1 0

0 0 0 2 0 5

333 6 105 23 1 8

Total

337

8

104

19

1

7

476

Panel B: Full European Parliament involvement (codecision procedure) Comitology procedure according to final act No comitology Advisory procedure Management procedure Regulatory procedure Safeguard procedure Several procedures

43 0 0 3 0 1

0 7 1 2 0 3

0 0 11 4 0 0

3 0 0 76 0 3

0 0 0 0 0 0

0 0 0 1 0 17

46 7 12 86 0 24

Total

47

13

15

82

0

18

175

Note: Entries are absolute numbers of cases. Source: Data set compiled by the author (see Box 6.1).

Table 6.2

Comitology preferences, 1999–2006: Commission versus European Parliament Comitology procedure according to first Commission proposal

Comitology procedure according to European Parliament amendment in first reading No comitology Advisory procedure Management procedure Regulatory procedure Safeguard procedure Several procedures Total

Regulatory procedure

Safeguard procedure

Total

No comitology

Advisory procedure

Management procedure

Several procedures

149 1 1

2 17 0

0 8 48

6 1 0

0 0 0

2 0 0

159 27 49

3

0

2

90

0

1

96

0 0

1 0

0 0

0 0

0 0

0 22

1 22

154

20

58

97

0

25

354

109

Note: Entries are absolute numbers of cases. In cases where the European Parliament has no amendments to the Commission’s proposed comitology procedure, the preference of the European Parliament is coded as identical to the Commission’s. Source: Data set compiled by the author (see Box 6.1).

110

Table 6.3 Comitology preferences, 1999–2006: European Parliament versus Council Comitology procedure according to European Parliament amendment in first reading

Comitology procedure according to final act No comitology Advisory procedure Management procedure Regulatory procedure Safeguard procedure Several procedures Total

Regulatory procedure

Safeguard procedure

Total

No comitology

Advisory procedure

Management procedure

Several procedures

145 2 4

1 10 10

1 0 45

1 0 0

0 0 0

0 0 0

148 12 59

6

3

1

93

0

2

105

0 2

0 3

0 3

0 3

0 1

0 20

0 32

159

27

50

97

1

22

356

Note: Entries are absolute numbers of cases. In cases where the European Parliament has no amendments to the Commission’s proposed comitology procedure, the preference of the European Parliament is coded as identical to the Commission’s. Source: Data set compiled by the author (see Box 6.1).

Comitology Preferences in Daily Legislation

111

Table 6.4 Comparison across time of Commission–Council agreement on comitology procedures (per cent) 1987–1995 (Dogan’s study: 1987 comitology decision) No comitology (not reported) Advisory procedure 41 Management procedure – variant A 67 – variant B 53 Regulatory procedure – variant A 61 – variant B (no Commission proposals) Safeguard procedure (not reported) Several procedures (not reported)

1999–2006 (Author’s study: 1999 comitology decision) No comitology Advisory procedure Management procedure

97 48 92

Regulatory procedure

94

Safeguard procedure Several procedures

100 88

Note: Entries are percentage of cases of new legislation where Council agrees with Commission’s comitology proposal. Source: 1987–95: Dogan (1997: Figure 8). 1999–2006: Author’s data (Table 6.1: Panels A and B).

the same comitology procedure as the Commission’s original proposal (see Table 6.4). It shows that under the 1987 decision there was agreement in roughly 50 per cent of all cases, although there is some variation across procedures. Under the 1999 decision this number has jumped to about 90 per cent. At the time Dogan was surprised by the high degree of agreement, especially between the Commission’s proposals and the Council’s decisions. He interpreted this to mean that in reality there was much more agreement on the choice of comitology procedure than the public rhetoric of the EU institutions would lead observers to believe (Dogan, 1997, pp. 44–5). Today there is even more apparent agreement on the choice of comitology procedure among the EU institutions. Whether this remarkable development is caused by a convergence of sincere preferences across time or an increasing use of strategic preferences is difficult to judge due to the problem of observational equivalence. However, I find the latter explanation most plausible. The 1987 decision was the first comitology decision. Actors had no experience with the standardized comitology procedures in legislative acts and had to get acquainted with each other’s preferences. By the turn of the millennium they had learnt and the comitology preferences of all actors were by then to a higher

112 The EU Comitology System in Theory and Practice

degree common knowledge. The most plausible interpretation of the development across time is therefore that it reflects a growing routinization of the handling of comitology issues in daily legislative practice: the Commission has stopped proposing comitology procedures that it knows will not survive in the Council. I now turn to the 59 cases of open disagreement on the choice of comitology procedure. They constitute a better data set to study comitology preferences since they are more likely to reveal true preferences. No matter whether amendments are due to incomplete information or symbolic politics, there is no reason to expect actors to systematically propose amendments contrary to their interests. A first indication of sincere comitology preferences can be found by inspecting the relative number of cases above and below the diagonal of agreement in Tables 6.1, 6.2 and 6.3. In Table 6.1 relatively more cases below the diagonal would, other things being equal, indicate that the Council favours more restrictive comitology procedures than the Commission, and this is indeed the case. The same pattern can be found in Table 6.3, which indicates that, other things being equal, the Council favours more restrictive comitology procedures than the European Parliament. Table 6.2, which compares the Commission and the Parliament, reveals the opposite pattern which indicates that, other things being equal, the Parliament favours more permissive comitology than the Commission. These patterns support the chapter’s hypotheses. Although supportive, the findings should nevertheless be treated with some caution and only be taken as a first indication. The reason is that apparent disagreement over comitology may in reality be a conflict over delegation. If actors disagree on delegation of executive power to the Commission, they will dislike a delegation proposal even though it is accompanied by a proposal to install a comitology procedure. For instance, if the Commission proposes delegation and comitology, but the member states in the Council do not want to delegate any powers, they will also disagree with the Commission’s comitology proposal. But this is a by-product of the conflict on delegation. The contrary situation may also be relevant: the Commission makes a proposal on a new directive or regulation, but contains no delegation, and hence no comitology. The Council then introduces delegation and an accompanying comitology procedure. If the Commission had known that the Council would be in favour of delegation, it might have proposed the same comitology procedure. Or it might not. The point is that in such

Comitology Preferences in Daily Legislation

113

a case we cannot know for sure what the actors’ relative comitology preferences are. If we want to rule out the alternative explanation that disagreements over comitology are by-products of conflicts over delegation, we need to control for conflicts over delegation. In other words, the ideal cases to study comitology preferences have no conflict over delegation, only disagreement on comitology. These are pure comitology conflicts. Identifying pure comitology conflicts requires a closer inspection of the limited number of disagreement cases. A report on the inspection of each case is available in Blom-Hansen (2011b), which can be downloaded from the author’s home page. This inspection shows that about half the cases contain conflicts over not only comitology, but also delegation. Only 26 cases are pure comitology conflicts – that is, cases of agreement on delegation, but disagreement on comitology. These are the most revealing cases about true comitology preferences. They are reported in Table 6.5. The pure comitology conflicts exhibit a pattern that clearly supports this chapter’s hypotheses. First, comparing the Commission and the Council, there is not a single example of a case where the Council Table 6.5 Pure cases1 of inter-institutional disagreement on comitology under the 1999 comitology decision

Directive no. 122/2003 on high-activity sealed radioactive sources Regulation no. 381/2001 creating a rapid-reaction mechanism Regulation no. 389/2006 on financial support for the Turkish Cypriot community Regulation no. 1382/2003 on the environmental performance of the freight transport system Regulation no. 1287/2003 on the harmonization of gross national income Regulation no. 2006/2004 on cooperation on consumer protection laws

Commission

Council

European Parliament

No com.

Adv.

No com.

Adv.

No com.

Adv.

No com., but direct consultation2 Managem.

Adv.

Adv.

Managem.

Adv.

Adv.

Reg.

Adv.

Adv.

Reg.

No com.

114 Table 6.5 (Continued)

Regulation no. 953/2003 on avoidance of trade diversion into the EU of medicines Regulation no. 743/2002 on judicial cooperation in civil matters Directive no. 22/2004 on measuring instruments Regulation no. 2152/2003 on monitoring of forests Regulation no. 166/2006 on a European Pollutant Release and Transfer Register Regulation no. 1177/2003 on Community statistics on income and living conditions Regulation no. 1830/2003 on traceability and labelling of genetically modified organisms Directive no. 12/2006 on waste Regulation no. 510/2006 on the protection of geographical indications of origin for agricultural products and foodstuffs Regulation no. 2826/2000 on information and promotion actions for agricultural products Regulation no. 2561/2001 on the conversion of fishing vessels and fishermen dependent on the fishing agreement with Morocco Regulation no. 2328/2003 on compensation for costs of marketing fishery products from the Azores Regulation no. 555/2000 on operations in the pre-accession strategy for the Republic of Cyprus Regulation no. 806/2004 on promoting gender equality in development cooperation

Commission

Council

European Parliament

Adv.

Adv. and reg.

(not involved)

Adv.

Adv. and managem.

Adv.

Adv.

Adv. and reg.

Adv.

Managem.

Reg.

Reg.

Managem.

Reg.

(unknown)3

Managem.

Reg.

Managem.

Reg.

Reg. and adv.4

Reg.

Reg.

Reg. and managem.5 Reg.6

Reg.

Managem.

Managem.

Adv.

Managem.

Managem.

Adv.

Managem.

Managem.

Adv.

Managem.

Managem.

Adv.

Managem.

Managem.

Adv.

Managem. and reg.

Managem. and reg.

Comitology Preferences in Daily Legislation Regulation no. 1725/2001 on action against anti-personnel landmines in third countries Regulation no. 1724/2001 on action against anti-personnel landmines in developing countries Regulation no. 2130/2001 on aid to uprooted people in Asian and Latin American developing countries Directive no. 60/2005 on prevention of the use of the financial system for money laundering and terrorist financing Directive no. 23/2004 on standards of quality and safety for the donation and so on of human tissues and cells Regulation no. 2424/2001 on the second generation Schengen Information System

115

Managem.

Managem.

Adv.

Managem.

Managem.

Adv.

Managem.

Managem.

Adv.

Reg.

Reg.

No com.

Reg.

Reg.

Adv.

(not involved) 7

Managem. and Managem.8 reg.

Abbreviations: No com.: No comitology procedure; Adv.: Advisory procedure; Managem.: Management procedure; Reg.: Regulatory procedure. Notes: 1 Pure cases are cases of conflict over comitology, but agreement on delegation. See Blom-Hansen (2011b) for all cases of disagreement on comitology. 2 The Council wants the Commission to report directly to the Council, not go through a comitology committee first. 3 Document missing in PreLex. 4 The Council accepts the proposal on the regulatory procedure, but inserts the advisory procedure to control another aspect of delegated power where the Commission had proposed delegation without comitology. 5 The Council accepts the proposal on the regulatory procedure, but inserts the management procedure to control another aspect of delegated power where the Commission had proposed delegation without comitology. 6 The Council insists on the use of the regulatory procedure for all delegated powers. 7 Member state initiative. 8 The European Parliament accepts comitology, but solely wants to use the management procedure. Source: data set compiled by the author (see Box 6.1).

favours a more permissive comitology procedure than the Commission. On the contrary, in all the cases where they disagree, the Council imposes a stricter comitology procedure than the one proposed by the Commission. Second, comparing the Council and the European Parliament, there is not a single case where the Parliament prefers a stricter procedure than the Council. In all the cases where they disagree, the Council imposes a stricter procedure than wanted by the Parliament. Finally, a comparison of the Commission and the Parliament shows that in all the cases where they disagree, except one, the Parliament prefers

116 The EU Comitology System in Theory and Practice

a more permissive procedure than the Commission. These patterns support the chapter’s hypotheses on the comitology preferences of the EU institutions.

Conclusion This chapter has studied institutional comitology preferences in daily legislation under the 1999 comitology decision. In the seven-year regime of this decision a total of 686 directives and regulations were enacted by the Council and the Parliament. These cases represent a hard case for identifying true preferences. First, there are guidelines for the choice of comitology procedures in the areas where power is delegated to the Commission. Although they are non-binding, they still provide a starting point. Second, EU legislative actors are familiar with each other’s comitology preferences, which raises a tricky problem of anticipated reactions. The actors are not likely to keep proposing procedures they know will not survive the legislative process. One of the main findings of the chapter is that open conflict on comitology procedures exists in less than 10 per cent of all new regulations and directives made under the 1999 comitology decision. Compared with legislation under the 1987 comitology decision this represents a considerable increase in apparent agreement on comitology issues in daily legislative practice. Apparent agreement represents a methodological challenge. It may indicate either a true convergence or a strategic use of preferences. In other words, there is a problem of observational equivalence. For this reason, the analyses of comitology preferences disregarded agreement cases and focused on cases of open conflict. Since the overwhelming majority of cases contain no disagreement on comitology, this led to a considerable reduction in the number of cases. But to control for the alternative explanation that disagreements over comitology are byproducts of conflicts over delegation, I scrutinized the disagreement cases further in order to identify cases of pure comitology conflict. That is, cases where comitology, not delegation, was contested. This led to a further reduction of the data set. These cases of pure comitology conflict, covering all EU actors and all legislative procedures, exhibit a pattern of comitology conflicts among the EU institutions that supports the chapter’s hypotheses. They also confirm the findings of the previous studies that have focused on selected actors or legislative procedures. The central finding is that the Commission systematically proposes permissive comitology while the Council prefers strict procedures. The European Parliament is on the side

Comitology Preferences in Daily Legislation

117

of the Commission and often prefers even more permissive comitology than the Commission. However, I do not want to overdo the importance of these findings. Since the data set had to be reduced because of the large number of agreement cases the findings are not based on a strong empirical basis. It is only when the findings are combined with the results from the previous chapter that a pattern begins to materialize. The previous chapter reached a similar conclusion in a case study of the 2006 comitology decision. Since the Council’s comitology decisions represent an obvious opportunity for all actors to improve their control positions over delegated decision-making such a study constitutes a most likely case for identifying comitology preferences. In comparison this chapter constitutes a least likely case. True comitology preferences are less likely to shine through due to the existence of non-binding guidelines and rational anticipation of other actors’ preferences. Since the two chapters reach similar results, they provide relatively strong support for the hypotheses. The main explanation of the development of the comitology system over time is that it is the result of a power struggle among actors who fundamentally disagree on its design. This is the driving force of the system.

7 Comitology Across Policy Areas

The comitology system was first used in the agricultural area in the 1960s, but soon spread to other areas (see Chapter 4). Over the years the system kept growing and was gradually introduced in all areas that have come under EU regulatory authority. Today comitology committees are found in all EU policy areas, for example the committee on fuel quality, which supervises the Commission’s regulation of petrol and diesel fuels used in road vehicles;1 the coordination committee of the funds, which monitors the Commission’s distribution of means from the EU structural funds;2 the tobacco products committee, which watches the Commission’s regulation of the production and marketing of tobacco;3 and the committee on administrative cooperation, which keeps track of the Commission’s activities to combat value added tax evasion in the member states.4 According to the Commission’s latest count, there are 266 comitology committees (2010a, p. 4). The spread of comitology over policy areas has been analysed by Dogan (1997, 2000) and Falke (1996, 2000). The Commission’s annual comitology reports provide more up-to-date accounts (see e.g. Commission, 2009a, 2010a). The reports are useful for broad information purposes, but do not offer information on the relative importance of comitology in the various policy areas; that is, how often comitology is used relative to the total regulatory activity in the different areas. This information is provided in Table 7.1 for the 1999–2006 period. This table shows the use of comitology in all directives and regulations made by the Council acting alone or with the European Parliament under the 1999 comitology decision that was effective from 1999 until 2006. A total of 686 new directives and regulations were made under this decision. Table 7.1 classifies the acts according to policy area and comitology procedure, and shows the relative use of comitology across areas. 118

Table 7.1

Comitology procedures across policy areas, 1999–2006 (per cent) N = 100 per cent

Advisory procedure

Management procedure

Regulatory procedure

Safeguard procedure

Several procedures

87

3

0

11

0

0

37

37

0

56

5

0

2

43

10

0

0

85

0

5

20

18

0

71

9

0

2

45

62

0

33

2

0

4

52

97

0

0

0

0

3

34

Transport policy

18

5

0

47

0

29

38

External relations: – General – External commercial policy

63 29

2 0

32 58

0 7

0 3

3 3

65 31

General financial and institutional matters Customs union Agriculture: – Agricultural health measures – All other agriculture Fisheries: – Common fisheries policy – External fisheries policy

119

No comitology

120

Table 7.1 (Continued)

– External trade protection Industrial policy and internal market Environment, consumers and health protection Common foreign and security policy Area of freedom, security and justice Other policy areas∗ Total

No comitology

Advisory procedure

Management procedure

Regulatory procedure

Safeguard procedure

Several procedures

N = 100 per cent

100

0

0

0

0

0

110

42

0

0

42

0

16

43

18

3

12

59

0

9

34

100

0

0

0

0

0

32

64

11

2

18

0

5

44

45

9

5

38

0

3

58

58

2

18

17

0

5

686

∗ Other policy areas include the following: freedom of movement; right of establishment; competition policy; taxation; economic and monetary policy; energy; regional policy; science, information, education and culture; law relating to undertakings. Note: Entries are row percentages. Due to rounding, their sum may not add up to 100. Classification into policy areas is based on the first mentioned Eur-Lex directory code in the individual act’s bibliographic notice in Eur-Lex. Source: Data set compiled by author (see Box 6.1).

Comitology Across Policy Areas 121

Turning first to the relative use of comitology across all policy areas, the bottom row of Table 7.1 shows that 58 per cent of all directives and regulations do not use the comitology system, while 42 per cent of the acts install one or several comitology procedures specified by the 1999 comitology decision. This confirms that comitology has become a standard operating procedure. It is the EU way of controlling delegated powers. Turning next to the relative use of the different comitology procedures, the bottom row of Table 7.1 shows that, although the 1999 decision specifies four comitology procedures, EU legislators prefer only two procedures in daily legislative practice, namely the management and regulatory procedures. It is rare that the advisory and safeguard procedures are used and that several comitology procedures are inserted in a given act. As a general picture, comitology under the 1999 decision means the management or the regulatory procedure. Turning finally to the use of comitology across policy areas, Table 7.1 shows a highly varying picture. In some areas, for example external trade protection, comitology is not used at all. In other areas, for example transport policy and environmental policy, almost all directives and regulations contain a comitology procedure. As far as exact choice of comitology procedure, the picture also varies across policy areas. The management procedure is frequently used within the areas of the customs union, the Common Agricultural Policy and external commercial policy (typically dealing with rules for the import of specific products from specific countries). The regulatory procedure is frequently used in the areas of agricultural health measures (rules on animal welfare and feed), transport policy and the environment. This brings us to the question addressed in this chapter: Why this variation across policy areas? In Chapter 3, I argued that the answer is found in the rational delegation literature. In areas with institutional conflict, legislators tighten procedural controls of the executive. In areas characterized by high issue complexity, legislators need to delegate regulatory authority to the executive, but they offset this with tighter procedural controls. In this way they may capture the benefits of allowing the executive increased flexibility while minimizing the risk of non-compliance. Based on this argument, the following hypotheses were developed: Hypothesis 7: Growing conflict within the Council and the European Parliament or between the Council, the European Parliament and the Commission leads to stricter comitology procedures. Hypothesis 8: Growing issue complexity leads to stricter comitology procedures.

122 The EU Comitology System in Theory and Practice

The empirical investigation of the hypotheses is organized in the following way. My analysis is not the first to investigate the variation of comitology across policy areas, so I start by reviewing previous studies, which offer a competing perspective on the issue the chapter will evaluate. I then explain my data and methods. The analysis is based on the data set that was introduced in the previous chapter. It consists of all new directives and regulations made by the Council (acting alone) or the Council and the Parliament (acting together) under the 7 year regime of the 1999 comitology decision, a total of 686 legislative acts. Since measuring the core concepts of conflict and complexity is tricky, I discuss my empirical indicators at some length. In the empirical analysis I investigate the relationship between these indicators and the legislators’ choice of comitology procedures. The chapter closes with a discussion of the value of the delegation perspective on the choice of comitology procedures in daily legislation.

Previous studies A number of studies have investigated the spread of comitology over policy areas and offered explanations of this development. Falke (1996, 2000) traces the expansion of the comitology system across time and policy areas. He notes that the number of committees increased fivefold from 1975 to 1999, and that this increase was most significant within new EU areas such as external relations, external trade, the internal market and industry, employment and social affairs, transportation, nuclear safety, the environment, consumer policy and public health (2000, pp. 346–8). While accepting that comitology committees can be used as means to supervise the Commission, Falke suggests that the main explanation of the spread of the system is that it empowers the Commission. Since, according to Falke, the Commission lacks the necessary administrative capacity in most areas, it needs to rely on the expertise of the national administrations of the member states. In this perspective, the comitology system constitutes a mixed national– European administration that enhances the quality of the Commission’s implementation of EU policies. According to Falke (1996, p. 138), comitology in effect means that ‘the Community has thus created an extensive circle of experts upon whose advice it can call through a kind of travel diplomacy’. Falke’s explanation strikes a chord with many comitology observers. Van Schendelen (1996) finds that the main purpose of comitology is to alleviate the Commission’s administrative overload and to increase its capacity and access to expertise. Other

Comitology Across Policy Areas 123

observers find that increasing the Commission’s capacity may be a motive for installing comitology committees in new areas, but not the main one. Dogan (1997, 2000), who provided the first comprehensive empirical account of the comitology system, finds that comitology is first and foremost introduced to control the Commission, but also to offer information. Other studies explain the spread of the comitology system across policy areas solely in terms of delegation and control. Franchino (2000a) made, to my knowledge, the first multivariate statistical analysis of comitology procedures in individual legislative acts. Relying on a sample of 100 regulations and directives adopted between 1987 and 1998 (that is, under the regime of the 1987 comitology decision), he demonstrated that indicators of issue complexity and inter-institutional conflict help explain the legislators’ choice of comitology procedure. This evidence supports this chapter’s hypotheses. Pollack (2003a, pp. 140–4) made a case study of the introduction of comitology procedures in the area of financial services as part of the Lamfalussy reform in 2001. His findings show that the member states felt compelled to delegate powers to the Commission because of the technical complexity of the financial sector and the ever-changing market conditions. At the same time the member states and the European Parliament were deeply concerned about the loss of control that broad delegation might entail. They, therefore, installed a regulatory comitology committee, the Securities Committee, to control the Commission. The case study shows how legislators try to reap the benefits of executive flexibility in complex areas while minimizing the risk of non-compliance and thus also supports this chapter’s hypotheses. In sum, the literature offers competing explanations of why the use of comitology varies across policy areas. One explanation is that comitology is installed to give the Commission access to expertise in the member states’ national administrations. Another is the explanation offered by this book, namely that comitology is established to control the Commission. The following empirical analyses can gauge the relative merit of these two explanations. If the expertise explanation is correct, we would expect comitology to be associated with the complexity of the issue area. This is also to be expected from the perspective of the control explanation, but this offers the additional expectation that the stringency of comitology procedures increases with the degree of complexity. The expertise explanation would expect no such relationship since the formal competence of the committee is irrelevant for informational purposes. In addition, the control explanation expects an

124 The EU Comitology System in Theory and Practice

association between conflict and comitology, which would be surprising from the perspective of the expertise explanation since this is irrelevant for informational purposes. The chapter’s analyses cannot make a final judgement on the relative merits of the two explanations since they both expect a positive relationship between comitology and issue complexity. But the chapter can evaluate whether the control perspective offers a fuller explanation, a perspective that explains more of the empirical phenomenon.

Data and methods The data for this chapter’s analyses are directives and regulations made by the Council and the Parliament under the regime of the 1999 comitology decision. This decision was effective for 7 years (from July 1999 until July 2006) and the EU legislators made a total of 686 new directives and regulations during these years. The data set was introduced in the previous chapter and is explained in more detail in Box 6.1 to which the reader is referred for further information on data sources, identification of cases and coding procedures. Although the 1987, 1999 and 2006 comitology decisions use the same labels for the various comitology procedures – advisory, management, regulatory, safeguard – the procedures in fact vary considerably across decisions. This point was discussed at length in the previous chapter. The conclusion was that it is necessary to hold the comitology regime constant in order to keep a firm grip on the dependent variable – the choice of comitology procedure. The previous chapter also discussed the fact that the 1999 comitology decision introduced non-binding guidelines for the choice of comitology procedure in daily legislation. Although these guidelines constrain the legislators’ freedom to some extent, there is still considerable freedom to decide the exact comitology procedure on a case-by-case basis. The reader is referred to the previous chapter for an extensive discussion of these points. Turning to the explanatory factors, although the chapter’s two hypotheses are by now standard within the delegation literature, there is no consensus in the literature on how to measure institutional conflict and issue complexity. Therefore I discuss my measures in some detail below.

Measuring institutional conflict Empirical delegation studies often rely on partisanship as a proxy for preferences (Epstein and O’Halloran, 1996, 1999, pp. 77–9, 129–33;

Comitology Across Policy Areas 125

Huber and Shipan, 2002, pp. 147–8, 183–7; Volden, 2002). However, partisanship as a measure is not readily applicable in the EU context. First, although the partisan background of the individual commissioners is far from irrelevant, the Commission is not a partisan actor. Second, although both the Council of Ministers and the European Parliament are partisan actors, party politics is different from the member state level where politicians sharing ideology are tied together by a common party organization. More direct measures of the EU institutions’ policy preferences are also difficult to establish in any systematic fashion. In recent years considerable progress has been made in understanding the policy preferences of the Council by analysing voting records, although no consensus has emerged on the exact dimensions in Council voting patterns (Hayes-Renshaw et al., 2006; Hagemann, 2007; Mattila, 2008). Likewise, our understanding of the policy preferences of the European Parliament has increased considerably due to a growing literature on parliamentary roll-call votes, which shows that the dominant dimension in the Parliament’s decision space is the traditional left–right cleavage (Hix, 2002; Hix et al., 2003, 2006). However, there is no consensus in the literature on the general policy preferences of the third EU institution, the Commission. Formal scholars portray it as a pro-integrationist competence-maximizer (Garrett and Tsebelis, 1996; Tsebelis and Garrett, 2000; Tsebelis, 2002, pp. 248–82), a conclusion that is also reached by several case studies of selected policy areas (Schmidt, 1998; Guiraudon, 2003; Pollack, 2003a, pp. 263–377). Other observers doubt that substantial differences can persist between the preferences of the Commission and the member states because of the procedures for selecting and appointing commissioners (Hug, 2003; Döring, 2007; Wonka, 2007). A third set of observers see the Commission as relatively impartial (Nugent, 2001, pp. 109–17) or as balancing among competing roles (Egeberg, 2006) and find no systematic evidence that Commission top officials are pro-integrationist (Hooghe, 2001). In short, the preferences of the Commission defy easy categorization. Since it is difficult to identify the policy preferences of the individual EU institutions, it is obviously also difficult to find a single measure that can be applied to all three institutions and used to establish the degree of preference divergence among them. However, recently two ambitious attempts at measuring the relative policy positions of EU legislative actors in a large-N setting have been launched. The first is the Decision-Making in the European Union (DEU) project reported in Thomson et al. (2006). The project selected 66 legislative acts proposed by the Commission between 1999 and 2002, and identified

126 The EU Comitology System in Theory and Practice

the policy positions of the Commission, the individual member states and the European Parliament by interviewing a large number of policy area experts. The selection criterion – all acts were controversial and had made it into the news – was necessary for the acts to provide a testing ground for the project’s theoretical models (Thomson and Stokman, 2006, p. 28). The second attempt was made by Franchino (2007, pp. 122–43, 257–61), who used national party election manifestos to identify the policy positions of the member states, the Commission and the European Parliament over the entire post-war period. These attempts at measuring the policy positions of EU legislative actors are both innovative and impressive but, unfortunately, they cannot be used for my purpose. I seek to analyse legislation that is as representative as possible, including a large number of legislative acts that never made it into the news and did not generate much open controversy. This makes it impractical to use Thomson et al.’s (2006) approach because interviewees are unlikely to have any clear remembrance of low-controversy acts. Franchino’s (2007) approach is of most value over a considerable time span since it cannot measure conflict at the level of the individual act within given election periods. This is less than ideal since there may be considerable issue-specific conflict at this level. The scholars in the DEU project deliberately refrained from using underlying dimensions to measure policy positions because they found a lack of consistency in the alignment of actors on the different issues in their data set (Thomson and Stokman, 2006, pp. 40–1). Actors who supported similar positions on one issue supported different positions on others. Therefore, the DEU project favoured the case-specific approach. Since my data consist of many cases across relatively few years, a casespecific approach has the most potential. But since expert judgements are not available to identify the actors’ policy positions, I need other indicators. I make use of the fact that absolute policy positions are not necessary for my purpose. It is sufficient to find indicators of preference divergence – that is, the distance between the actors’ policy positions. For this purpose I have collected data on three indicators. The first indicator is the length of the decision-making process. My contention is that, ceteris paribus, the longer it takes for the legislators to decide an act, the more they disagree. Among the 686 cases in the data set, the time between the first Commission proposal and the final enactment varies from less than 1 year to 7 years. However, this indicator should be used with caution because, not surprisingly, the time varies considerably and systematically across legislative procedures.

Comitology Across Policy Areas 127

Eighty four per cent of the acts where the Council acts alone (that is, without involvement of the Parliament) are decided in less than one year. The corresponding figures for acts adopted under the more cumbersome consultation and codecision procedures are, respectively, 34 and 5 per cent. So the more the Parliament is involved, the longer it takes to decide legislative acts. It means that if the length of the decision-making process is to be used as a proxy for institutional conflict, it is necessary to control for the type of legislative procedure. The second indicator is the number of times an act has been on the Council agenda as a ‘B-point’. In the Council system, legislative proposals from the Commission are first treated in working groups consisting of national officials. At this level the majority of issues are agreed upon, but some issues are sent on to Coreper for further negotiation. At this level, most issues are then solved and sent on to the ministerial level as ‘A-points’ – issues that only require formal ministerial approval. In a small number of cases, Coreper fails to agree, and these cases are sent on to the Council as ‘B points’ that still need to be resolved (Hix, 2005, pp. 80–3). Although ‘false B-points’ sometimes exist – issues with no real disagreement, but which for political reasons are put on the Council agenda for discussion (Lewis, 2002, p. 288) – my contention is that, on average, ‘B-points’ indicate conflict. The majority of the cases in the data set have only been on the Council agenda as ‘A-points’, but 32 per cent have been on the agenda as a ‘B-point’ one or more times. Seven per cent of the cases have been on the Council agenda as ‘B-points’ more than one time. This means that the ministers have discussed the case and sent it back to Coreper for further clarification and then discussed it again at the ministerial level. One extreme case has been on the Council agenda as a B-point nine times. The third indicator measures whether the Commission makes use of its right to change its legislative proposal before the Council acts. When this occurs, it may signal conflict. As Nugent (2001, p. 255) notes, it may indicate that the Commission accommodates a negative opinion in the Council. But it may also signify the opposite situation – that the Commission tries to force the Council to deal seriously with new aspects of its proposal. For instance, if the Parliament under the consultation procedure proposes amendments which the Commission likes, but which it fears that the Council is going to ignore, it may change its proposal to include the amendments. The Council can then only reject them with unanimity. In sum, an amended proposal may indicate a defensive or an aggressive Commission, but in either case it is a signal of institutional conflict. It is relatively rare that the Commission uses the option

128 The EU Comitology System in Theory and Practice

to change its proposal. In the 686 cases in the data set it only happens in 17 per cent of the cases. The three measures of institutional conflict – length of the decisionmaking process, ‘B-points’ on the Council agenda and changed Commission proposals – are correlated (0.67
Measuring issue complexity Developing empirical indicators of the complexity of the regulatory field facing legislators, and thus their need for technical expertise when regulating, has been a challenge in the delegation literature. Some studies avoid the problem by deliberately focusing on one policy area and thus holding complexity constant (Epstein and O’Halloran, 1996; Huber and Shipan, 2002, p. 142). Other studies have found ingenious indicators. In a delegation study of state welfare policies in the United States, Volden (2002) measures complexity by the composition of the client group, arguing that welfare policies are more complex to handle for legislators when demographics are changing. In a study of US federal agencies, Lewis (2008, pp. 115–18) measures the complexity of the agencies’ regulatory fields by the composition of their workforce, arguing that a large share of technical or professional employees indicate complexity. In their study of delegation in major US laws Epstein and O’Halloran (1999, pp. 206–8) develop a complexity measure based on committee hearings in the Congress. Their argument is that hearings are used as a legislative tool to gather technical and political information. Hearings can therefore be used as a proxy for complex issue areas. Franchino (2004, 2007, pp. 143–6) develops complexity indicators for the EU system based on a similar logic. Since the Commission is the initiator of EU legislation, it also needs to gather technical information to formulate proposals in complex issue areas. In the most complex areas, it gathers the relevant information on the issue and drafts a report. These reports are both a sign of information gathering and an instrument to induce the relevant political actors to signal support or opposition to the measures. Franchino’s contention is that reports are produced in areas with a greater need to gather technical and political information before formulating legislative proposals, and therefore can be used as indicators of issue complexity. Franchino also makes use of the fact that the Commission in some cases relies on expert committees for advice when drafting legislative proposals. These committees collect information, carry out studies, draw up reports and support the

Comitology Across Policy Areas 129

Commission in the drafting of concrete proposals. Committees can consist of experts from the national technical ministries or independent experts. Franchino’s contention is that the existence of these committees is a sign of the need to gather technical and political information. They deal with complex and specialized issues and therefore serve as indicators of issue complexity. I follow the logic employed by Epstein and O’Halloran and Franchino and use information gathering in the policy formulation stage to develop two proxies for issue complexity. First, I use reports and other documents issued by the Commission as part of the preparation of legislative proposals as a complexity indicator. Since such preparatory documents may be referred to in various ways, I coded instances where a reference is made to a ‘programme’, ‘report’, ‘plan’ or ‘communication’. A dichotomous variable ‘preparatory documents’ will take the value 1 if recitals in the relevant directive or regulation refer to one of these documents, and the value 0 if not. Examples of such references include regulation no. 871/2004 on new functions for the Schengen information system, which refers to the Community action plan of 21 September 2001 against terrorism; regulation no. 870/2004 establishing a Community programme on the conservation of genetic resources in agriculture, which refers to the Community action plan for biodiversity conservation and the protection of genetic resources in agriculture; and regulation no. 2560/2001 on cross-border payments in euros, which refers to the Commission’s communication on means to assist economic actors in switching to the euro and reports by the European Central Bank on improving cross-border payment services. Second, I utilize the fact that the Commission in some cases relies on expert committees for advice when drafting legislative proposals. If a directive or regulation is based on preparatory work by such a committee, I use this as an indicator of issue complexity. A dichotomous variable ‘preparatory committee’ will take the value 1 if recitals in the relevant directive or regulation refer to an advisory expert committee, and the value 0 if not. Examples of such references include directive no. 43/2006 on statutory audits of annual accounts, which is based on work by the special committee on auditing; directive no. 23/2004 on quality standards for the donation of human tissues and cells, which is based on work by the scientific committee for medicinal products and medical devices; and regulation no. 701/2006 on rules for the harmonized index of consumer prices, which is based on work by the statistical programme committee. As these examples indicate, the use of an advisory committee may serve as an overall indicator of complexity.

130 The EU Comitology System in Theory and Practice

Empirically, this has been confirmed in a number of analyses of EU legislation by Franchino (2004, 2007, pp. 160–95). However, in the data set used in this book one policy area needs careful attention for this indicator to function properly. One hundred and six cases, almost one-sixth of all cases, are regulations imposing countervailing or antidumping duties on third countries trying to export goods to the EU area (in Table 7.1 these acts are classified under ‘external trade protection’). These acts are retaliatory measures imposed because of unfair trade practices by the selling country. It is an area where the Commission has been delegated considerable power. It can impose provisional duties, but they must be confirmed later by the Council to be definitive. When introducing anti-dumping or countervailing duties the Commission must consult with an advisory committee of member state representatives. This procedure is forced upon the Commission by the member states as a formal requirement in the basic regulations.6 It is a control mechanism rather than a technical support mechanism, and it is comparable to a comitology procedure, although the Council is automatically involved when duties are made definitive (Holmes and Kempton, 1996; Woolcock, 2000, pp. 388–91). Consequently in this case, it is misleading to use the existence of an advisory committee as an indicator of complexity. This problem is probably relevant for some other committees too, but this can be treated as ‘noise’ in the data. But given the large number of cases of anti-dumping and countervailing duties, including them would bias the complexity indicator. To deal with this problem. I use a dummy variable to control for these cases. In addition to the two complexity indicators, I use an indicator based on the final enacted directive or regulation. Some contain one or more annexes, which specify in greater detail the issues dealt with in the main text of the act. My contention is that their existence indicates that regulation needs to deal with issues of considerable detail, and that legislators face a complex field. A dichotomous variable ‘annex’ will take the value 1 if the relevant act contains one or more annexes, and the value 0 if not. For example, directive no. 99/2002 on animal health rules on products of animal origin for human consumption contains an annex on relevant animal diseases; directive no. 122/2003 on the control of high-activity sealed radioactive sources contains an annex on activity levels for radionuclides; and regulation no. 91/2003 on rail transport statistics contains an annex on the detailed reporting of annual statistics on goods transport. In sum, the following empirical analysis includes three indicators of issue complexity: preparatory documents, preparatory committees

Comitology Across Policy Areas 131

and annexes to legislative acts. The indicators are uncorrelated (0.20
Measuring the dependent variable: The legislative choice of comitology procedure The comitology system has existed since the early 1960s, but the exact comitology procedures have varied considerably over time. Until 1987, the system had no general rules on procedures that could be applied in individual cases. This led to a bewildering number of procedures in practice. Bergström (2005, p. 189) estimates that no less than 30 procedural variants were used when the Council made its first comitology decision in 1987, which introduced a simplified list of procedures that the legislators could use in daily legislation. But it still specified four different procedures and several variants. The second comitology decision in 1999 simplified matters further and streamlined the comitology procedures into four standard ones (Haibach, 1999; Bergström, 2005, pp. 189–209, 249–85). In 2006, a new fifth procedure was introduced when the Council established the regulatory procedure with scrutiny that was examined in Chapter 5. In 2011, the management and regulatory procedures were turned into variants of the new examination procedure (Brandsma and Blom-Hansen, 2011). Until the Lisbon Treaty, the same labels were used for the various comitology procedures – advisory, management, regulatory and safeguard – but as argued at length in the previous chapter, the procedures in fact varied considerably across comitology decisions. For this reason the following analysis focuses on legislative practice under a single decision, namely the 1999 comitology decision that was effective from 1999 until 2006. This allows a firm grip of the object under study – the comitology preference of the EU institutions during the legislative process. The 1999 decision specifies four comitology procedures: • The advisory procedure: The Commission must take ‘the utmost’ account of the committee’s opinion, but may enact the measure regardless of that opinion. • The management procedure: The Commission can enact the measure unless a qualified majority of the committee opposes, in which case the measure is referred to the Council.

132 The EU Comitology System in Theory and Practice

• The regulatory procedure: The Commission can enact the measure only if a qualified majority of the committee supports it. Otherwise the measure is referred to the Council. • The safeguard procedure: Any committee member may refer the Commission’s measure to the Council. As argued in the previous chapter and as game theoretical analyses confirm (Steunenberg et al., 1996, 1997; Ballman et al., 2002), the constraining effect on the Commission of the comitology procedures can be ranked in this way: Advisory procedure < management procedure < regulatory procedure < safeguard procedure This rank order will guide this chapter’s analysis of the EU legislators’ choice of stringency of control of the Commission by comitology procedures. However, in practice, there is not sufficient variation in the choice of comitology procedure under the 1999 decision to allow an analysis of degrees of control stringency that includes all four procedures. As Table 7.1 shows, daily legislative practice makes use of only two procedures – the management and the regulatory procedures. The advisory procedure is only used in 2 per cent of all cases, and the safeguard procedure is hardly used at all. An analysis of control stringency therefore needs to focus on the management and the regulatory procedures.

Empirical analysis Below I conduct two tests of the two hypotheses. The first investigates when the EU legislators decide to install a comitology committee and when they do not, and compares the choice of some kind of control by comitology committees with the choice of not installing a comitology committee. If correct, the hypothesis should be able to explain this choice. Both institutional conflict and issue complexity should lead the legislators to install comitology committees. However, the alternative explanation focusing on the Commission’s need for expertise that was introduced at the beginning of the chapter would also expect complexity to lead to the instalment of a comitology committee. Therefore, the first test cannot firmly establish the reason for the effect of issue complexity. The second test focuses on the stringency

Comitology Across Policy Areas 133

of comitology control. Since this is irrelevant for the Commission’s informational needs, the alternative expertise explanation provides no expectation. In contrast, the book’s control explanation would expect institutional conflict and issue complexity to have a positive impact on the stringency of comitology control. Specifically, this analysis focuses on the choice of the regulatory procedure over the management procedure. The 2 tests are made as a series of logistic regression analyses of the 686 new directives and regulations made by the Council (acting alone) or the Council and the Parliament (acting together) under the 7 year regime of the 1999 comitology decision. The dependent variable in the first test takes the value 1 if the act contains a clause installing a comitology committee, irrespective of committee type, and the value 0 if not. In the second test the dependent variable takes the value 1 if the act contains a clause installing a regulatory comitology committee, and the value 0 if the act installs a management comitology committee. The independent variables in the two tests are the indicators of institutional conflict and issue complexity that were introduced above plus a number of control variables that will be explained as they are introduced. Table 7.2 provides summary statistics on all variables. The logistic regression coefficients in the two tests indicate whether the various explanatory factors increase or decrease the likelihood of EU legislators choosing comitology/stringent comitology, respectively. Since logistic regression is non-linear, the coefficients do not indicate the substantive magnitude of the effects, only their direction and statistical significance. For instance, a positive and statistically significant coefficient means that if a given legislative act has more of the characteristic measured by the relevant explanatory factor, it increases the likelihood that EU legislators will choose comitology/stringent comitology. To provide an idea of the substantive size of the coefficients, I report the probability of choosing comitology/stringent comitology for selected types of institutional conflict and issue complexity. The first test, the legislative choice of some kind of comitology control, is reported in Table 7.3. This test comprises the full data set, all 686 new regulations and directives made by the Council (acting alone) or the Council and the Parliament (acting together) under the 7 year reign of the 1999 comitology decision.7 The test is composed of five models of increasing complexity. The first two models are essentially bivariate

134 Table 7.2 Summary statistics on variables used in logistic regression models Definition

Range

Mean

I. Dependent variables: Comitology over no comitology

Dichotomous variable (1 = act installs a comitology committee; 0 = otherwise)

0–1

0.42

Regulatory procedure over management procedure

Dichotomous variable (1 = act installs a regulatory comitology committee; 0 = act installs a management comitology committee)

0–1

0.49

Index of three conflict indicators (length of the decision-making process; ‘B-points’ on the Council agenda; changed Commission proposals)

0–3

1.01 (st.dev.: 1.05)

– Preparatory document

Dichotomous variable (1 = recitals refer to preparatory programme, report, plan or communication; 0 = otherwise)

0–1

0.18

– Preparatory committee

Dichotomous variable (1 = recitals refer to an advisory expert committee; 0 = otherwise)

0–1

0.25

– Annex

Dichotomous variable (1 = act contains one or more annexes; 0 = otherwise)

0–1

0.56

– Council acting alone

Dichotomous variable (1 = Council acts without Parliament; 0 = other legislative procedures)

0–1

0.44

– Codecision procedure

Dichotomous variable (1 = codecision procedure; 0 = other legislative procedures)

0–1

0.26

II. Independent variables: Institutional conflict

Complexity:

Legislative procedure:

Comitology Across Policy Areas 135 – Consultation procedure

Dichotomous variable (1 = consultation procedure; 0 = other legislative procedures)

0–1

0.29

Countervailing or anti-dumping duty

Dichotomous variable (1 = act imposes a countervailing or anti-dumping duty on a third country; 0 = otherwise)

0–1

0.16

Type of act (directive)

Dichotomous variable (1 = directive; 0 = regulation)

0–1

0.18

Partially amending act

Dichotomous variable (1 = act partially amends an existing act; 0 = new act)

0–1

0.13

Dichotomous variable (1 = act in area not mentioned in article 2, point a or b, in the 1999 comitology decision; 0 = otherwise)

0–1

0.77

– Management procedure

Dichotomous variable (1 = act in area mentioned in article 2, point a, in the 1999 comitology decision; 0 = otherwise)

0–1

0.15

– Regulatory procedure

Dichotomous variable (1 = act in area mentioned in article 2, point b, in the 1999 comitology decision; 0 = otherwise)

0–1

0.08

Informal comitology guideline: – No guideline

analyses of the two major explanatory factors, institutional conflict and issue complexity, respectively. The first model shows that institutional conflict, as expected, has a positive impact on the likelihood of inserting a comitology clause into EU legislative acts. This model also includes a control for legislative procedures. As argued above this is necessary for the length of the legislative process (one of the items in the institutional conflict index), not simply to reflect the various legislative procedures.8 Model 2 shows that the three measures of issue complexity – preparatory documents, preparatory committees and annexes – as expected also have a positive impact on the legislative choice of comitology, although the strength of the relationship varies across the three measures. The model also includes a control for countervailing

Table 7.3 Logistic regression analysis of the choice of comitology over no comitology Model 2:

Model 3:

Model 4:

Conflict

Complexity

Conflict and complexity

Conflict plus controls

0.26 (0.13)∗∗ 0.13 (0.25) 0.22 (0.32) 1.30 (0.20)∗∗∗

0.36 (0.13)∗∗∗ – – –

(ref.cat.) 1.29 (0.33)∗∗∗ 0.39 (0.25)

(ref.cat.) 1.74 (0.34)∗∗∗ 0.34 (0.25)

0.37 (0.12)∗∗∗ – – –

– 0.59 (0.23)∗∗ 0.57 (0.29)∗ 1.24 (0.18)∗∗∗

Model 5: Complexity plus controls – 0.45 (0.24)∗ 0.40 (0.31) 1.10 (0.19)∗∗∗

Legislative procedure: – Council acting alone – Codecision procedure – Consultation procedure Countervailing or anti-dumping duty Type of act (directive) Partially amending act

(ref.cat.) 1.64 (0.31)∗∗∗ 0.60 (0.23)∗∗ –

– – – –21.40 (3,712)

–20.38 (3,712)

–

– – – –20.92 (3,725)

– –

– –

– –

–0.56 (0.31)∗ 0.93 (0.28)∗∗∗

–0.34 (0.25) 0.83 (0.28)∗∗∗

– – –

– – –

– – –

(ref.cat.) 1.23 (0.25)∗∗∗ 1.70 (0.47)∗∗∗

(ref.cat.) 0.66 (0.24)∗∗∗ 1.63 (0.42)∗∗∗

–1.62∗∗∗ 0.34 711.5 667

–1.24∗∗∗ 0.40 692.6 686

Comitology guidelines: – No guideline – Management procedure – Regulatory procedure Constant Nagelkerke R2 –2 log likelihood N ∗

–1.35∗∗∗ 0.25 768.8 667

–0.95∗∗∗ 0.34 733.0 686

–1.67∗∗∗ 0.42 654.9 667

p < 0.1; ∗∗ p < 0.05; ∗∗∗ p < 0.01. Dependent variable: Dichotomous measure of the choice of comitology over no comitology (1 = act installs a comitology committee; 0 = otherwise). Entries are logistic regression coefficients (standard errors in parenthesis).

136

Institutional conflict Preparatory documents Preparatory committees Annexes

Model 1:

Comitology Across Policy Areas 137

or anti-dumping duties, which, as argued above, is necessary for the preparatory committee variable to function properly. Model 3 presents a simultaneous test of the conflict and complexity explanations of the legislative choice of comitology. As is evident both explanations are weakened in this test. This is due to a certain collinearity between the measure of institutional conflict and two of the complexity measures, preparatory documents and preparatory committees (gamma = 0.57 and −0.40, respectively). As this makes it difficult to disentangle the independent effects of these variables, I proceed with separate tests of the two main explanatory variables. Model 4 tests the impact of institutional conflict when controlled for a number of additional factors. As is evident, the conflict variable survives this test intact compared with model 1. The model includes three control variables. First, a control is made for the type of legislative act which shows that comitology clauses are somewhat less likely to occur in directives compared with regulations. Next, a control is made for partially amending acts. As explained in the previous chapter, the data set only includes new acts, not amending acts, in order to control for the status quo. But in practice this distinction is blurred since a number of acts introduce both new acts and amendments to existing acts. These acts – 91 out of the total of 686 acts – are included in the data set, but coded so that a control for their status can be made. It turns out that comitology clauses are more likely in these acts. Finally, a control is made for the 1999 comitology decision’s guidelines for the legislative choice of comitology procedure. As explained in detail in the previous chapter, these guidelines are non-binding and leave the legislators considerable discretion in practice. Nevertheless, it cannot be ruled out that they constrain the legislative choice of comitology procedure to some degree. Therefore, a control is relevant and is done by coding whether the policy area, to which the act primarily belongs, is among the areas mentioned by the 1999 comitology decision as suitable for either the management or regulatory procedure.9 The guidelines turn out to be strong predictors of comitology clauses in legislative acts. Model 5 presents the corresponding test of the complexity explanation when controlled for additional factors. The explanation does not survive this test to the same extent as the conflict explanation did in model 4. One of the complexity measures, preparatory committees, fails to reach statistical significance, and another, preparatory documents, is only significant at the 10 per cent level. But the third measure,

138 The EU Comitology System in Theory and Practice

annexes, is still a strong predictor. In sum, the complexity explanation is weakened, although not to be ruled out. The control variables reach approximately the same results as in model 4. In sum, Table 7.3 cannot refute that institutional conflict and issue complexity have a positive impact on the EU legislators’ choice of inserting comitology clauses in legislative acts. This finding is robust across alternative specifications, especially for the conflict variable.10 To illustrate the substantive impact of the conflict and complexity variables, Table 7.4 reports the probability of making this choice for selected types of institutional conflict and issue complexity. Comparing low and high conflict, while holding control variables constant, shows that the probability that a comitology procedure is installed increases by almost 50 per cent (from 0.53 to 0.76). Moving from low to high complexity acts, while holding control variables constant, more than doubles the likelihood of comitology (from 0.22 to 0.58). These changes are substantial. The second test focuses on the stringency of comitology control. As argued above, this is operationalized as the choice of the regulatory procedure over the management procedure. This test consequently only comprises legislative acts including a clause installing a regulatory or a management committee – 236 of the 686 cases in the full data set. The analysis is reported in Table 7.5. As in the first test, the analysis is composed of models of increasing complexity. Again, the first two models are essentially bivariate analyses Table 7.4 Probability of choosing comitology over no comitology for selected types of conflict and complexity Type of conflict and complexity

Low conflict and fixed levels of control variables High conflict and fixed levels of control variables Low complexity and fixed levels of control variables High complexity and fixed levels of control variables

Probability of choosing comitology over no comitology 0.53 0.76 0.22 0.58

Note: Probabilities computed on the basis of models 4 and 5 in Table 7.3. Low conflict = 0; High conflict = 3; Low complexity designates preparatory documents = 0, preparatory committees = 0 and annexes = 0; High complexity designates preparatory documents = 1, preparatory committees = 0 and annexes = 1. Fixed levels of control variables: Codecision procedure = 1; Consultation procedure = 0; Type of act = 0; Partially amending act = 0; Comitology guideline for management procedure = 0; Comitology guideline for regulatory procedure = 0.

Table 7.5

Logistic regression analysis of the choice of stringency of comitology control

Institutional conflict Preparatory documents Preparatory committees Annexes

Model 1:

Model 2:

Model 3:

Model 4:

Conflict

Complexity

Conflict and complexity

0.66 (0.26)∗∗ – – –

– 0.47 (0.34) 2.12 (0.48)∗∗∗ 0.89 (0.34)∗∗∗

0.78 (0.28)∗∗∗ −0.29 (0.45) 1.78 (0.62)∗∗∗ 1.46 (0.48)∗∗∗

0.96 (0.36)∗∗∗ 0.09 (0.55) 2.79 (0.90)∗∗∗ 2.18 (0.68)∗∗∗

(ref.cat.) 3.55 (0.70)∗∗∗ 1.18 (0.63)∗

– – –

(ref.cat.) 3.51 (0.73)∗∗∗ 1.26 (0.65)∗

(ref.cat.) 2.96 (0.85)∗∗∗ 3.50 (0.88)∗∗∗

– –

– –

– –

2.40 (0.93)∗∗ –0.05 (0.59)

– – –

– – –

– – –

(ref.cat.) –3.53 (0.78)∗∗∗ 0.65 (0.79)

Conflict and complexity plus controls

Legislative procedure: – Council acting alone – Codecision procedure – Consultation procedure Type of act (directive) Partially amending act Comitology guidelines: – No guideline – Management procedure – Regulatory procedure Constant Nagelkerke R2 –2 log likelihood N

–2.96∗∗∗ 0.61 180.2 232

–1.12∗∗∗ 0.19 290.0 236

–4.41∗∗∗ 0.66 163.2 232

–5.44∗∗∗ 0.79 115.1 232

∗

139

p < 0.1; ∗∗ p < 0.05; ∗∗∗ p < 0.01. Dependent variable: Dichotomous measure of the stringency of comitology control (1 = act installs a regulatory comitology committee; 0 = act installs a management comitology committee). Entries are logistic regression coefficients (standard errors in parenthesis).

140 The EU Comitology System in Theory and Practice

of the two major explanatory factors, institutional conflict and issue complexity, respectively. The first model shows that institutional conflict, as expected, has a positive impact on the stringency of comitology control. As in the first test, this model includes a control for legislative procedures. Results are comparable. Model 2 shows that two of the three measures of issue complexity – preparatory committees and annexes – have a positive impact on the choice of stringency of comitology control. The third complexity measure is positive, but fails to reach statistical significance.11 Model 3 presents a simultaneous test of the conflict and complexity explanations of the legislative choice of comitology. Substantially identical results are obtained. This is noteworthy given the collinearity problems in the corresponding model in Table 7.3. But in the subset of the data set analysed in Table 7.5, collinearity between the conflict and complexity measures is less of a problem.12 I, therefore, proceed with a simultaneous test of the two main explanatory variables. Model 4 finally tests the impact of institutional conflict and issue complexity when controlled for a number of additional factors. As is evident, both explanations survive this test intact compared with the previous models. Indeed, they both end up stronger than in the previous models. As in the first test, this model includes three control variables. First, directives are more likely than regulations to install regulatory committees relative to management committees. Second, whether the act partially amends an existing act has no discernable influence on the choice of the regulatory over the management procedure. Finally, the 1999 comitology decision’s guidelines for the legislative choice of comitology procedure are again strong predictors. In sum, Table 7.5 cannot refute that institutional conflict and issue complexity have a positive impact on the EU legislators’ choice of comitology stringency. Again, this finding is robust across alternative specifications.13 To illustrate the substantive impact of these variables, Table 7.6 reports the probability of choosing stringent comitology for selected types of institutional conflict and issue complexity. Comparing low and high conflict, while holding complexity low and control variables constant, makes the probability of stringent comitology jump from 0.08 to 0.60. Moving from low to high complexity, while holding conflict low and control variables constant, dramatically increases the probability of stringent comitology from 0.08 to 0.92. Again, these changes are substantial.

Comitology Across Policy Areas 141 Table 7.6 Probability of choosing stringent comitology for selected types of conflict and complexity Type of conflict and complexity

Low conflict, low complexity and fixed levels of control variables High conflict, low complexity and fixed levels of control variables Low conflict, high complexity and fixed levels of control variables

Probability of choosing stringent comitology 0.08 0.60 0.92

Note: Probabilities computed on the basis of model 4 in Table 7.5. Low conflict = 0; High conflict = 3; Low complexity designates preparatory documents = 0, preparatory committees = 0 and annexes = 0; High complexity designates preparatory documents = 0, preparatory committees = 1 and annexes = 1. Fixed levels of control variables: Codecision procedure = 1; Consultation procedure = 0; Type of act = 0; Partially amending act = 0; Comitology guideline for management procedure = 0; Comitology guideline for regulatory procedure = 0.

Conclusion This chapter has investigated the variation in the use of comitology procedures across policy areas. It started by demonstrating this variation. In some policy areas comitology is not used at all; in other areas almost all directives and regulations include a comitology clause. If comitology is used, some areas prefer the management procedure, while other areas rely on the regulatory procedure. Some areas use a mix of comitology procedures. Why this variation? According to this book’s perspective, the answer is found in the rational delegation literature which points to two types of explanations: in areas with institutional conflict the legislature will tighten procedural controls of delegated powers; in areas characterized by high issue complexity, the legislature needs to delegate powers to the executive, but offsets this with tighter procedural controls. The purpose of the chapter was to investigate the merit of these explanations for the legislative choice of comitology procedures in the EU system. At the same time the chapter has gauged the merit of an alternative perspective on comitology, namely the argument that this system is installed to give the Commission access to expertise necessary for formulating technically complex rules. To investigate these explanations, the chapter used a data set of all new legislation in 1999–2006, the 7 year regime of the 1999 comitology decision during which the Council and the Parliament enacted 686 new

142 The EU Comitology System in Theory and Practice

directives and regulations. These acts were all coded to allow a test of the explanations. The results show that both institutional conflict and issue complexity have a positive impact on the introduction of some kinds of comitology control. They also have a positive impact on the stringency of comitology control. This evidence supports the explanations based on the rational delegation literature. Although the argument on the Commission’s need for expertise would also expect a positive relationship between issue complexity and comitology control, it is silent on the possible impact of institutional conflict and on any explanations of the stringency of comitology control. Actually any relationships here would appear surprising because they are unrelated to the Commission’s informational needs. For this reason the explanations based on the rational delegation literature constitute a fuller understanding of the variation of comitology across policy areas. They help make sense of not only when the EU legislators install comitology committees, but also which supervisory powers they allow them.

8 The Comitology System in Daily Practice

Introduction When subjected to control by comitology procedures, the Commission must ask the relevant committee of member state representatives for a formal opinion before implementation measures can be adopted. The committee then delivers its opinion by voting on the Commission’s proposal. In 2009, the comitology committees delivered a total of 2091 formal opinions. Only ten opinions – less than one per cent – were unfavourable (Commission, 2010a). 2009 was not an unusual year. It is very rare that comitology committees vote against the Commission. This fact is often taken as an indication that the daily workings of the comitology system are peaceful and consensual, and that the cumbersome comitology procedures do not really constrain the Commission. However, this interpretation is vulnerable to the objection that the voting results are also consistent with the exact opposite situation. If the comitology committees function as perfect guardians of national interests, and if the Commission perfectly anticipates what is acceptable to the committees, the committees never need to exercise a negative vote. In other words, there is a problem of observational equivalence. This is a general problem for studies of legislative oversight of the executive. A lack of active legislative monitoring may indicate either a rubber-stamping legislature or a perfect alignment of legislative–executive preferences. The problem of observational equivalence has been recognized at least since Weingast and Moran’s (1983) demonstration that passive Congressional oversight of US regulatory agencies may indicate either legislative control or bureaucratic independence. 143

144 The EU Comitology System in Theory and Practice

Like the US literature on Congressional control of regulatory agencies, the literature on comitology control of the Commission is divided on how effective the control system is for the daily operation of the Commission. As noted by Pollack (2003b), two images of comitology are advanced in the literature. According to the first image, comitology committees are arenas where policy experts deliberate in order to help the Commission find the professionally best solutions to policy problems. In contrast, the second image portrays comitology committees as bargaining arenas where the member states keep a close watch over the Commission’s regulatory activities. As argued by Weingast and Moran (1983), other data than monitoring activities are necessary to settle the effectiveness of legislative control with the executive. In the area of comitology, studies of the daily operation of the system have relied on two types of evidence in addition to voting results: case studies and surveys. However, as I will argue below, this evidence is inconclusive, but it suggests that both images may be true. In other words, the exact nature of the daily operations of the numerous comitology committees is still an open question. This is puzzling. It is an accepted view in the literature that the comitology system is designed to control the Commission, and the preceding chapters support this view. The member states spend considerable time and energy on the exact specifications of the various comitology procedures, and on when and where to install them. Can it be true that the member states forget about the system once the committees are established and let their representatives operate as autonomous experts? To understand this puzzle, I argued in Chapter 3 that we need to take four analytical steps. To briefly summarize this argument, we need to start by shifting our focus from Brussels to member state capitals. To understand how the national representatives operate in Brussels, we need to understand their instructions. To what extent are the representatives controlled by their home government? This question, which can only be answered by studying the internal workings of the national government administrations, is key to understanding the representatives’ behaviour in Brussels. Second, having understood the representatives’ instructions, we can shift our focus back to Brussels and study the meetings of the comitology committees. Representatives armed with clear instructions are not likely to engage in deliberation but to defend their negotiation mandate by bargaining. In contrast, representatives who are allowed discretion can afford to deliberate. In other words, it is likely that there is a

The Comitology System in Daily Practice

145

link between the representatives’ discretion and their behaviour in the committees. Third, the behaviour of the individual national representatives determines the interaction style in the committees. If committees are dominated by deliberating or bargaining representatives, the interaction style will be, respectively, deliberative or bargaining. Committees with representatives exhibiting different types of behaviour are likely to have less clear-cut interaction styles. Fourth, understanding the interaction style in the committees is central for understanding the effect of the comitology system. Does it constrain the Commission? It is likely that not only the comitology procedures, but also the interaction styles in the committees determine the constraining effect of the system. Based on this argument, the following hypotheses were developed: Hypothesis 9: Growing issue complexity leads to increased discretion for the national representatives in the comitology committees. Hypothesis 10: Growing conflict between the national government and the national representatives leads to reduced discretion for the representatives in the comitology committees. Hypothesis 11: Reducing/increasing the national representatives’ discretion leads to more/less bargaining behaviour and less/more deliberative behaviour by the representatives in the comitology committee. Hypothesis 12: Committees dominated by deliberating or bargaining representatives are characterized by, respectively, a deliberative or bargaining interaction style. Hypothesis 13: The more comitology committees are characterized by a deliberative interaction style, the more active the Commission is. The purpose of this chapter is to investigate these hypotheses. I start by explaining in more detail the two prevailing images of the daily workings of the comitology system and the existing empirical evidence. I then discuss the methods and data I use to investigate the hypotheses. The data stem from a population survey of all comitology representatives from one member state, Denmark, and before embarking on the empirical analyses I introduce the Danish setting and the Danish comitology representatives. The empirical analyses are then conducted in four steps corresponding to the steps described above and summarized in Figure 3.2 in Chapter 3. The chapter concludes with a discussion of the lessons of the analyses for the relative merit of the two images of comitology and suggests avenues for further research.

146 The EU Comitology System in Theory and Practice

Two images of the daily workings of comitology The literature offers two images of the daily workings of comitology (Pollack, 2003b). The first image – associated with authors such as Joerges and Neyer (1997a, b), Dehousse (2003) and Wessels (1998) – builds upon sociological institutionalism and constructivism. It suggests that comitology committees provide a forum in which experts meet and discuss to find the best or most efficient solutions to common policy problems. The advocates of this image argue that comitology is government by persuasion, argumentation and discursive processes rather than by command, control and strategic action. Joerges and Neyer (1997a, b) introduced the term ‘deliberative supranationalism’ to describe the interaction style in the comitology committees. They base this on three arguments. First, they find formal structures to be of limited importance: ‘The boundaries of the committee system cannot be equated with its formal structures’ (1997a, p. 279). The lines are blurred between the comitology committees and other groups, such as the Commission’s informal advisory groups, which prepare new rules, and the working groups under Coreper, which negotiate the adoption of rules in the Council of Ministers. Issues as well as persons overlap across these various committee arenas. Second, the decision style in the comitology system is not one of bargaining, but of problem-solving (Joerges and Neyer, 1997a, pp. 288–9). Power rests in the ability to present and substantiate arguments. To be convincing, arguments must be backed by scientific evidence (see also Wessels, 1998, p. 225; Dehousse, 2003, p. 803). Finally, advocates of the deliberative image argue that comitology has a strong socializing effect on its participants. They note that committee negotiations sometimes last for years with the same national representatives participating. During their course of working together, the representatives develop common understandings of problems and solutions. The prevalence of this deliberative image of comitology has been stressed many times by Joerges and Neyer since they presented it in the mid-1990s (Joerges, 2002, 2006; Neyer, 2003, 2006). Today, they even refer to it as ‘the comitology mode of decision-making’ (Joerges, 2006, p. 779). The second image of comitology is very different from the deliberative one. It builds upon rational choice theory and is associated with authors such as Steunenberg et al. (1996, 1997), Pollack (2003a, pp. 114–46), Ballman et al. (2002) and Franchino (2000b). It portrays comitology as a

The Comitology System in Daily Practice

147

mechanism designed by the member states to control the Commission’s executive duties. In this view, the comitology committees function as miniature versions of the Council of Ministers. They use the same qualified majority voting weights, and, as in the policy formulation phase, the right of initiative belongs to the Commission. Consequently, the decision style is intergovernmental bargaining, well-known from the Council (cf. Moravcsik, 1998). Member state representatives are careful watchdogs of their national interest. In contrast to the deliberative image, formal rules are of paramount importance since they constrain the Commission to varying degrees; for instance, the advisory procedure has no particular importance or scientific interest because it ‘does not restrict the Commission in the slightest way’ (Steunenberg et al., 1996, p. 341). According to this image, comitology meetings are implementation games or principal–agent relationships. Comitology is necessary because legislative acts passed by the Council are incomplete contracts whose daily operation needs to be monitored to avoid problems of agency loss and defection by the original contracting partners. Comitology is a continuation of Council negotiations by stand-ins, and committee procedures are instruments for pursuing national interests (Steunenberg et al., 1996, 1997; Franchino, 2000b; Ballman et al., 2002; Pollack, 2003a, pp. 114–46, 2003b). The two images present widely different understandings of the daily workings of the comitology system. Empirical investigations of their validity have primarily rested on two types of evidence. The first type of evidence is case studies of selected individual committees. Several authors have documented the empirical relevance of the two images in studies of individual committees. Joerges and Neyer (1997a; see also Neyer, 1998) – the primary advocates of the deliberative image – built their case on a careful study of the Standing Committee for Foodstuffs. Daemen and van Schendelen (1998) investigated the Advisory Committee on Safety, Hygiene and Health Protection. Like Joerges and Neyer, they found a dominant culture of consensus based on an expertise orientation. But at the same time, they discovered that national representatives seem to be rather cynically pursuing national interests in the setting of common standards for workers’ protection. In the Ecolabel committee, which awards ecolabels to environmentally friendly products, Philip (1998) found little trace of deliberative processes. The label is commercially sensitive, and the national representatives have a keen eye on their national competitive advantages. In a study of the Environmental Management and Audit Scheme (EMAS)

148 The EU Comitology System in Theory and Practice

Committee, Töller (1998) found strategic use of voting rules to coexist with a consensus-oriented decision process. In a review of the Genetically Modified Organisms (GMO) Committee’s treatment of a multinational company’s application to market genetically modified maize in the EU territory, Bradley (1998) reports a cynical and daring use of formal voting rules. Gatt (2009), who also investigates comitology within the GMO area, reaches a similar conclusion: ‘Member state representatives are always severely divided on the matter, defeating any claim to consensual decision-making’ (p. 181; see also Pollack and Shaffer, 2008). In the area of financial regulation, de Visscher et al. (2008) find that the Lamfalussy reform, which introduced comitology committees in this field, succeeded in increasing the role of expertise and reducing the role of politics. This supports the deliberative image. Finally, in a study of three environmental comitology committees – the drinking water committee, the nitrate committee and the packaging waste committee – Gehring (1999) found that member state representatives refuse to ignore their national interests, but are also compelled to provide reasonable arguments rather than to rely on their power resources. He concludes (p. 215) that ‘comitology committees are hybrids that must accommodate two different, and possibly conflicting, modes of interaction’. In sum, case studies show traits of both deliberative and rationalist images of comitology. The second type of evidence is survey data on individual representatives, which is limited to two studies. Egeberg et al. (2003) surveyed 218 participants in comitology committees and other EU committees from 14 EU member states. They found support for a distinct EU negotiation culture characterized by arguing (rather than bargaining), expertise (rather than power) and a high level of loyalty and trust. This supports the deliberative image. However, they also found differences across types of committees. Compared with Commission expert groups, both Council working groups and comitology committees appear to be intergovernmental arenas, where participants act as representatives of their home government and feel closely aligned to it. This supports the rationalist image. Similar results have been reached by Sannerstedt (2005) in a survey of Swedish EU committee members. In sum, the existing survey evidence is also inconclusive. All in all, the available evidence cannot refute either image of comitology. There is support, albeit scanty, for both of them. The remainder of this chapter offers a new approach to understanding the two images and their relative merits by shifting the focus from Brussels to national capitals and by bringing in new and more systematic data.

The Comitology System in Daily Practice

149

Methods and data Data to investigate hypotheses 9–13 were collected by a survey of Danish representatives in nearly all comitology committees that were operative in the autumn of 2006. The survey was conducted in the following steps. First, comitology committees were identified. No precise up-to-date register of comitology committees existed when the survey was planned,1 but help was available in the Commission’s annual comitology report, which lists all committees at the time of writing. An initial list of committees was compiled on the basis of the Commission’s (2006b) comitology report for 2005. Committees that had not met in 2005 were excluded as ‘inactive’ committees. Committees established after the report was printed were identified by consulting the Danish representatives and the EU coordinating units in the Danish ministries. The second step, identifying the Danish representatives, soon proved to be an arduous task. Member states are free to choose any representative they want to attend the comitology meetings. At least in Denmark, this means that a variable number of civil servants, sometimes spanning more than one ministry, are attached to the various committees. The choice of representative for a given meeting depends on the agenda. This problem was discussed with the relevant ministry’s or agency’s EU coordinating unit in order to identify the most frequent, important or experienced representative on the individual committees. In this way 191 Danish representatives, each covering one committee, were identified, contacted by phone or e-mail and asked to fill out a questionnaire, which was subsequently sent to them by ordinary mail. One hundred and sixty-one respondents filled out the questionnaire, which gives an overall response rate of eighty-four per cent. Please see Box 8.1 for more information on the survey.

Box 8.1

The Danish comitology survey

Operational definition of a comitology committee Defining a comitology committee is not straightforward. First, some committees are divided into subcommittees. In the survey, subcommittees count as regular committees if they have a permanent character and if they make decisions according to the formal comitology rules. One example is the Standing Committee on the

150 The EU Comitology System in Theory and Practice

Food Chain and Animal Health which is divided into a number of permanent sections. Second, some committees meet very rarely. The survey only comprises committees which met at least once in 2005 according to the 2005 comitology report by the Commission (2006b). Third, new committees are constantly added to the system. The survey was initially based on the Commission’s 2005 report. Committees established after this report was printed were identified as the participating actors and were consulted in the data-collection process. Given these selection criteria, the survey comprised 191 comitology committees. Identifying the national representative The member states are free to appoint their national representative to the individual comitology committees. At least in Denmark, this means that a variable number of civil servants, sometimes spanning more than one ministry, are attached to the various committees, but there is no central registration of these persons. The first problem was to identify the responsible ministry or agency. This problem was discussed with the central EU coordinator in each of the Danish ministries, who provided an overview of the comitology committees for which their ministry is responsible. They also specified the responsible unit or agency within the ministry. These units were then contacted by telephone in order to identify the most frequent, important or experienced Danish representative in the individual committees. This representative was then contacted by telephone and asked to fill out the questionnaire. Preparing the questionnaire and collecting answers The questions were tried out in exploratory interviews with experienced comitology representatives from the Danish Environmental Protection Agency and the Directorate for Food, Fisheries and Agri Business in October 2006. When the questionnaire was ready, all respondents were contacted by telephone in advance. The purpose of the survey was explained to them, and they were all guaranteed anonymity. The questionnaire was then sent by ordinary post in November 2006. If it was not returned within 10–14 days, the respondents were contacted again at least twice by e-mail

The Comitology System in Daily Practice

151

or telephone. By February 2007, 161 out of a total of 191 questionnaires had been completed and returned. This gives an overall response rate of 84 per cent. Evaluation of non-response The overall response rate of 84 per cent means that 16 per cent of the questionnaires were not completed. To assess the impact of non-response, three characteristics of the population – responders and non-responders – were evaluated: the institutional affiliation of the committees in the Commission, the institutional affiliation of the committees in the Danish central administration and the type of committee (advisory, management or regulatory). Measured against these criteria there is almost no difference between the population and the responders. The survey is thus not characterized by any serious bias stemming from non-response. Further information See Blom-Hansen (2007), which can be downloaded from the author’s home page, for more information on the above points, documentation of analysis of non-response, an exact list of committees, formulation of survey questions and frequency distributions of answers.

Before embarking upon the analysis of this data set, its strengths and weaknesses should be made clear. The major strengths are threefold. First, it is a population analysis. It is not restricted to a sample, but contains information on all comitology committees and thus includes maximum variation on all relevant parameters. Second, the data have been collected in a systematic way. Considerable effort has been taken to systematically find and contact representatives on all committees. This stands in contrast to the previous two surveys of comitology representatives. Egeberg et al. (2003) obtained their data by distributing questionnaires to seminar participants at the European Institute of Public Affairs (EIPA) in Maastricht. As the authors readily acknowledge, this resulted in a non-representative sample of committee members. Sannerstedt’s (2005) data collection is based on a list of Swedish officials created by the Swedish Chamber of Commerce.

152 The EU Comitology System in Theory and Practice

Apart from a note that this list comprises a broad section of Swedish officials with EU experience, there is no explanation of the representativeness of the sample of respondents.2 Third, the questionnaire contains information about the discretion allowed to the representatives by their home ministry. This makes it possible to investigate the connection between national capitals and the committee meeting rooms in Brussels. While the data set thus makes it possible to bring the study of the daily operations of the comitology committees one step forward, the limitations of the data should also be acknowledged. First, the data are restricted to one member country, Denmark. This raises a question about potential country bias. However, there is no reason to expect Danish representatives to exhibit any systematic bias when asked about the nature of the meetings in their committees. Comparisons across committees of interaction style, issue complexity and the like are therefore unlikely to suffer from selection bias. However, answers to questions about the discretion allowed by the home ministry obviously only cover the Danish situation. Second, the data consist of the respondents’ impressions of the nature of committee meetings, not ‘hard information’. Subjective impressions may to some extent be influenced by individual characteristics of the respondents. The following analyses therefore control for the gender, age and education of the respondents. In addition, the analyses include the length of the respondents’ participation in the committees to control for potential socialization effects. Third, having only one respondent per committee raises concerns about reliability, but since there is no reason to expect systematic bias, this simply introduces ‘noise’ in the data and makes the hypothesis testing conservative. Fourth, the response rate to the survey is not 100 per cent, but 84 per cent, but there are no indications of bias stemming from non-response, see further Box 8.1. Fifth, since both dependent and independent variables stem from the data set, they are measured at the same time, which raises concern about the direction of causality. I return to this problem in the conclusion. Finally, a complete investigation of all four steps in the argument – that is the full causal chain outlined in Figure 3.2 – is a tall order, and one which the data cannot fully deliver. I discuss this problem as the analysis proceeds and return to it in the conclusion. Given these limitations of the data, the following analysis may bring the debate over the daily workings of the comitology system one step further, but it does not provide the final answer. In the chapter’s analysis, I discuss future avenues for research.

The Comitology System in Daily Practice

153

The Danish comitology representatives In Brussels each comitology committee has a clear institutional affiliation in the Commission. Although several directorate-generals (DGs) may be involved in the work of a given committee, one DG is responsible and provides the committee’s chairman and main secretarial assistance. In the member states’ national administrations institutional affiliation is less clear-cut. The member states are free to choose their representatives and may decide to attach several representatives to a given committee. In the Danish central administration in Copenhagen this means that a variable number of civil servants, sometimes spanning more than one ministry, are attached to the various committees. Consequently, identifying the Danish representative is not straightforward. As explained above, the respondent to the questionnaire is the most frequent, important or experienced Danish representative from the various committees. With this reservation, the institutional affiliation of the comitology committees in the Danish central administration and the match with their affiliation in the Commission is provided in Table 8.1. This table shows that there is a close match between the institutional affiliation of the comitology committees in Brussels and Copenhagen. For instance, almost all committees belonging to the DG Agriculture in Brussels are affiliated with the Danish Ministry of Food, Agriculture and Fisheries in Copenhagen. Conversely, almost all committees belonging to the Danish Ministry of Food, Agriculture and Fisheries are affiliated with the DG Agriculture in Brussels. The only exception is the Danish Ministry of Economic and Business Affairs, but this is an unusually large and heterogeneous ministry and it is not surprising that its functions cover a number of DGs in Brussels. Within the Danish ministries, the comitology committees may be affiliated with the ministerial department, but most are affiliated with agencies or other institutions under the ministries. Of the 33 committees belonging to the Danish Ministry of Food, Agriculture and Fisheries, seven are affiliated with the Ministry’s Plant Directorate, 22 with the Ministry’s Food Industry Agency, two with the Ministry’s Directorate of the Fisheries and two with ministerial research institutions (the Agricultural Research Center and the Institute of Food and Resource Economics). In total, only 28 out of the 191 identified committees are anchored in ministerial departments. The remaining 163 are organized under ministerial agencies and institutions. Turning to the individual representatives, the typical Danish participant in a comitology committee is found at the bottom of the

154

Table 8.1 The Brussels–Copenhagen comitology connection Commission

Danish Ministry

Total

DG Economic

Food,

Transportation

and

Agriculture

and Energy

Business

and

Affairs

Fisheries

Environment

Science

Taxation Family and

Justice

Consumers

Foreign

Other

Affairs

Ministries∗

Enterprise

2

1

2

4

2

−

−

−

−

2

Agriculture

−

22

−

1

−

−

2

−

−

−

25

Transport and

2

−

17

−

−

−

−

1

−

1

21

Environment

−

−

−

20

−

−

−

−

−

−

20

Research

−

1

−

−

14

−

−

−

−

−

15

Information

−

−

−

−

5

−

−

−

−

1

6

6

−

1

−

−

−

−

−

−

1

8

1

−

−

−

−

20

−

−

−

1

22

13

Energy

Society Internal Market Taxation and Customs

−

−

−

−

2

−

−

−

1

4

7

1

6

−

1

−

−

9

−

−

3

20

−

−

−

−

−

−

−

5

−

2

7

Europe Aid

−

−

−

−

−

−

−

−

7

−

7

Statistics

6

−

−

−

−

−

−

−

−

−

6

Other DGs∗∗

2

3

−

−

−

2

−

−

3

4

14

20

33

20

26

23

22

11

6

11

19

191

Education and Culture Health and Consumers Justice, Liberty and Security

Total

Notes: The table includes the 191 comitology committees where it was possible to identify the Danish representative. Entries are absolute numbers of committees. ∗ ‘Other Ministries’ include ministries with fewer than five committees: Employment, Education, Culture, Finance, Defence, Social Affairs, Refugees and Immigration, Interior and Health. ∗∗ ‘Other DGs’ include DGs with fewer than five committees: Employment and Social Affairs, Fisheries, Regional Policy, Trade, Enlargement, Humanitarian Aid, Budget and Anti-Fraud Office.

155

156 The EU Comitology System in Theory and Practice

ministerial hierarchy. Approximately 80 per cent of the participants are heads of units or consultants. These are positions under a head of office, which is normally the first position where occupants have responsibility over staff. At the same time, most are experienced committee participants. On average the respondents have participated in their committee for more than five years. They are also experienced civil servants since their average age is 50 and they are highly educated. More than 90 per cent have a tertiary education, typically from a university. Approximately one-third have a specialist education such as engineering or natural science. The remaining two-thirds are generalists with degrees in law, political science or economics. Finally, it is a male-dominated world; women constitute only about one-third of the respondents.3

Step 1: The discretion of the national comitology representative The first step in the argument on the daily workings of the comitology committees is to investigate the instructions of the national representative. According to hypotheses 9 and 10, issue complexity and conflict between the national government and the comitology representatives have an impact on the discretion of the individual representatives. The empirical indicators used to investigate these hypotheses, as well as the remaining hypotheses, stem from the survey of the Danish comitology representatives. Most indicators are indices constructed from two or more questions in the survey. All indices are coded to vary between 1 and 5, and high scores on all indices indicate high levels of the phenomenon in question. All indices have been subjected to reliability testing, and the result (Chronbach’s alpha) is reported in Table 8.2, which also contains summary statistics on all indices and other variables used in this chapter’s analyses. Measuring the discretion of the national representative. The respondents have been asked to evaluate their freedom to determine the Danish position when votes are taken in their comitology committee. They have indicated the extent to which they agree with the following two statements: ‘In reality I have considerable freedom to decide the national position in my committee’ and ‘I take the position in my committee that I think is best according to my own professional judgement. The answers to the two questions have been combined into an index which yields a continuous measure of perceived discretion at the level of the individual representative.

157 Table 8.2 Descriptive statistics on all variables used in regression analyses in Chapter 8 Range

Mean value

St. dev.

N

Discretion Index of two survey questions (alpha = 0.59):a In reality I have considerable freedom to decide the national position in my committee; I take the position in my committee that I think is best according to my own professional judgement

1–5

3.9

0.9

161

Issue complexity Index of three survey questions (alpha = 0.71):a You need considerable technical insight to work with the cases in my committee; It is difficult for outsiders to understand the cases we work with in my committee; The cases in my committee are often technically complex

1–5

3.9

0.8

161

Political interest Index of five survey questions (alpha = 0.87):b How much are the following actors interested in the work in your comitology committee? Your minister, the government, the Parliament, the press and interest organizations

1–5

2.2

0.8

160

Age (years) Specialist education Female Committee experience (years)

27–69 0–1 0–1 0.4–28.0

49.5 0.3 0.3 5.4

10.9 − − 5.1

160 161 161 154

Deliberative disposition Index of two survey questions (alpha = 0.64):a It is important for me that the committee reaches the technically best solution; It is important for me that I can defend the technical contents of the committee decisions

1–5

4.3

0.6

161

Bargaining disposition Index of two survey questions (alpha = 0.46): a It is important for me to defend Denmark’s interest in the committee; It is important for me to observe my negotiation mandate

1–5

4.5

0.6

161

Deliberative interaction style Index of four survey questions (alpha = 0.52):a It happens that participants are persuaded by good arguments to change their position; Arguments on common solutions are especially important; The participants normally present detailed arguments for their positions; All participants can freely express their opinion

1–5

4.1

0.5

160

158 The EU Comitology System in Theory and Practice Table 8.2 (Continued) Range

Mean value

St. dev.

N

Bargaining interaction style Index of four survey questions (alpha = 0.63):a Compromises are normally political horse-trades; It is really the large countries that decide; The participants often resort to bluffing; National interests dominate our work

1–5

2.6

0.7

160

Mediator Index of three survey questions (alpha = 0.77):c How often does the EU Commission play the following roles in your comitology committee? The one who makes sure that everybody feels comfortable; The one who makes sure that the result is satisfactory to everybody; The one who mediates

1–5

3.7

0.7

159

Policy advocate Index of four survey questions (alpha = 0.62): c How often does the EU Commission play the following roles in your comitology committee? The one who prepares things; The one who investigates and explains things; The one who promotes his own proposals; The one who drives negotiations forward

1–5

4.3

0.4

159

Comitology procedures Categorical variable: Advisory, management regulatory, several procedures

–

−

−

159

Fixed DG effects Categorical variable: DG Enterprise and Industry, DG Agriculture, DG Transport and Energy, DG Environment, DG Research, DG Taxation and Customs Union, DG Health and Consumer Protection and other DGs (residual category of remaining DGs with fewer than ten observations)

–

−

−

191

a Based on five-point response scale (‘agree’, ‘partly agree’, ‘neither agree nor disagree’, ‘partly disagree’ and ‘disagree’). b Based on five-point response scale (‘very much’, ‘much’, ‘somewhat’, ‘a little’ and ‘not at all’). c Based on five-point response scale (‘always’, ‘often’, ‘now and then’, ‘seldom’ and ‘never’).

Measuring issue complexity. As discussed at length in Chapter 7, it is a challenge to find empirical indicators of issue complexity. In the present context I rely on information provided by the respondents to the survey. They have indicated the extent to which they agree with the following

The Comitology System in Daily Practice

159

statements: ‘You need considerable technical insight to work with the cases in my committee’; ‘It is difficult for outsiders to understand the cases we work with in my committee’ and ‘The cases in my committee are often technically complex. I have combined their answers to the three statements into an index of issue complexity. Since complexity is such a challenging concept to operationalize, the validity of empirical indicators is an issue of concern. For the present indicator, two validity checks are possible. First, like other national administrations, the Danish central administration is divided into a number of ministries which again are divided into a ministerial department and a number of agencies. The organizational rationale is that the ministerial department deals with general matters and offers policy advice to the minister, while the agencies deal with the details of day-to-day administration and provide the informational basis for new legislation. Consequently, we expect comitology committees that are anchored in agencies to deal with more technically complex issues than committees anchored in the ministerial departments. And indeed, the average value on the complexity measure is 4.0 for comitology committees anchored in agencies, and 3.3 for committees anchored in ministerial departments. Since the complexity measure ranges from 1 to 5, this is a notable difference. Second, some committees deal with issues that, a priori, are plausibly technically more complex than others. A rough correspondence between the complexity measure and prima facie understandings of the committees’ tasks is thus to be expected. Table 8.3 provides an overview of the ten committees that score the lowest and the highest values on the complexity measure. The table shows that the low complexity committees deal with general policy initiatives, research and welfare issues. In contrast, high complexity committees deal with technical environmental issues, technical aspects of trade questions, statistical and customs issues. This rank order makes sense and has a certain degree of face validity. In sum, the complexity measure appears to be a useful indicator of the complexity of the various comitology committees’ tasks. The second main independent variable is conflict between the national government and the comitology representatives. However, since the Danish government has followed the ally principle to the greatest extent possible, conflict is not a variable but a constant in the data set. Although free to choose its representatives, the Danish government has chosen civil servants from the ministerial apparatus in all cases. There is not a single example of a representative from local government, an interest organization, an independent research

160 The EU Comitology System in Theory and Practice Table 8.3 Comitology committees scoring low and high values on the issue complexity measure The ten committees with the lowest complexity score: Committee for Community actions in support of consumer policy (2004–2007) Research programme committee (subcommittee on research infrastructure) Committee of the Community action programme to encourage cooperation between member states to combat social exclusion Advisory committee on the approximation of the laws of the member states relating to machinery Committee for the fourth multiannual programme for small and medium-sized enterprises in the EU Committee for the application of the regulation authorizing voluntary participation by undertakings in the Community eco-management and audit scheme (EMAS) The anti-fraud committee on mutual assistance in customs and agricultural matters Management committee for agricultural promotions Research programme committee (subcommittee on science and society) Research programme committee (subcommittee on human resources and mobility) The ten committees with the highest complexity score: Committee for the implementation of the directive concerning the placing of biocidal products on the market Committee for the implementation of the directive on the control of major accidental hazards involving dangerous substances European Agricultural Guidance and Guarantee Fund Committee Committee on the harmonization of gross national income at market prices Committee on article 6 visa Committee on the second general system for the recognition of professional education and training Customs Code Committee (subcommittee on agriculture and chemicals) Combined heat and power committee Committee for the harmonization of national measures on the indication by labelling and standard product information of the consumption of energy and other resources by household appliances Committee on safe seas and prevention of pollution from ships

institution or a private enterprise. Although it appears plausible that people from outside the government apparatus in some cases may be more insightful, the government prefers to have its own people represent it. Consequently, conflict cannot be investigated actively, but is held constant across all cases. Finally, since the analysis focuses on the representatives’ subjective impression of their discretion, it includes controls for individual characteristics (gender, age, education and committee experience). In addition,

The Comitology System in Daily Practice

161

it is possible to control for the wider interest taken by influential actors in Denmark in the comitology committees. The respondents have indicated the extent to which their minister, the government, the Danish Parliament, the press and interest organizations are interested in the work in their committee. The answers have been combined into an index which controls for the political salience of the committees, or the attention paid to them. The analysis is reported in Table 8.4, which shows that issue complexity, as expected, is positively related to the discretion allowed to the individual comitology representative. The effect is not only strongly statistically significant, but also of substantial magnitude. Complexity and discretion are both measured on a five-point scale. A one-point increase in the complexity index leads to an increase of one-third of a point in the discretion index. This finding, which is consistent with both the deliberative and the rationalist image of comitology, indicates that some representatives are less controlled by their home government than others and therefore may behave in a less constrained and more autonomous way at committee meetings. But this is no coincidence. Discretion is sometimes granted because of the technical complexity of the issues dealt with by the committees, not because the government has forgotten about the control mechanism it has installed.

Table 8.4 Regression analysis of the discretion of the Danish comitology representatives Full model Issue complexity Political interest Individual characteristics of the respondent:

0.32 (0. 10)∗∗∗ −0.19 (0. 09)∗∗

– Female – Age – Specialist – Committee experience

0.20 (0. 17) 0.02 (0. 01)∗ −0.04 (0. 16) 0.02 (0. 02)

Constant Adj. R2 N ∗

2.12∗∗∗ 0.10 152

p < 0.1; ∗∗ p < 0.05; ∗∗∗ p < 0.01. Dependent variable: Index of discretion. Entries are OLS-regression coefficients (standard errors in parenthesis).

Reduced model 0.30 (0. 09)∗∗∗ −0.20 (0. 09)∗∗

– 0.02 (0. 01)∗∗ – – 2.35∗∗∗ 0.10 159

162 The EU Comitology System in Theory and Practice

The control variables in Table 8.4 show that political interest, not unexpectedly, leads to less discretion for the representatives. It is not surprising that the government wants to control representatives dealing with salient issues. Furthermore, the respondent’s age is positively related to discretion, perhaps because older representatives perceive their instructions as more lenient, or because age to some extent functions as a proxy for general job experience. If so, the positive relationship indicates that more experienced civil servants are allowed more discretion than their younger and less experienced colleagues.

Step 2: The behaviour of the national comitology representative The second step in the argument on the daily workings of the comitology committees is to link the instructions to the national representatives with their behaviour in the committees. According to hypothesis 11, the representatives’ discretion has an impact on the extent to which they exhibit bargaining or deliberative behaviour in the committee. Unfortunately, the questionnaire does not contain information on the respondents’ actual behaviour, but it allows an investigation of their behavioural disposition. This concept is discussed in various terms in the literature; from a sociological institutional perspective it is often called a role orientation. Egeberg (1999, pp. 458–9) explains: A single individual may have several roles and identities. . . . By ‘role’ we usually mean a set of expectations (norms or rules) that more or less specify the desired behaviour of the role incumbent. . . . To say that individuals follow roles and identities is not to say that their behaviour is always easily predicted. . . . The crucial question is: Which role or identity becomes evoked in a particular decision situation? . . . Organizations and institutional arrangements are normative systems that are supposed to structure the occasions for evoking one identity or another. They are also incentive systems. Officials learn to evoke a particular role orientation by experiencing the rewards and punishments of having done so in the past . . . (emphasis in original). From a rational institutional perspective, behavioural dispositions are often called ‘preferences’. In her analysis of top Commission officials, Hooghe (2001) focuses on their general views and beliefs about the

The Comitology System in Daily Practice

163

EU, which she calls ‘preferences’, or ‘basic orientations’. She explains (Hooghe, 2001, p. 11): Preferences are general guidelines – heuristic aids to action – not a set of algorithms. They are context-sensitive propensities to action. . . . They profoundly influence, but do not determine, action. Since behavioural dispositions are not perfectly correlated with behaviour, an investigation of the Danish comitology representatives’ behavioural dispositions does not constitute a strong test of hypothesis 11. The hypothesis focuses on actual behaviour, but if discretion influences actual behaviour, this influence is likely to go through the representatives’ behavioural disposition. Investigating the impact of discretion on behavioural dispositions is thus an analysis of one of the causal mechanisms linking discretion to behaviour. To measure their behavioural dispositions, the Danish respondents were asked what is important to them in their committees. The questions were formulated to make possible an evaluation of the respondents’ disposition for deliberative or bargaining behaviour. Table 8.5 shows the empirical indicators of the behavioural dispositions and the results of a factor analysis. The indicators clearly load on two different dimensions that can be interpreted as a deliberative and a bargaining disposition. If the representatives’ discretion affects their behavioural dispositions in the hypothesized way, increasing their discretion should make them more disposed for deliberation and less for bargaining. To investigate this, I combine the first two statements in Table 8.5 into an index of deliberative disposition and the last two statements into an index of bargaining disposition (see Table 8.2 for summary statistics). I regress the behavioural dispositions on the discretion variable and control variables, and I control for issue complexity although it, according to the four-step argument, is an indirect cause. However, the intervening variable – discretion – may not capture the full effect of complexity. I also control for political interest, which arguably may have an independent effect on behavioural dispositions. Finally, I again control for individual characteristics of the respondents. The analysis is reported in Table 8.6. Table 8.6 confirms that increasing the comitology representatives’ discretion makes them more disposed for deliberative behaviour. The effect is of moderate substantial magnitude. Discretion and deliberative disposition are both measured on a five-point scale. A one-point increase in

164 Table 8.5 Factor analysis of the behavioural dispositions of the Danish comitology representatives Loading on factor 1

Loading on factor 2

(deliberative disposition)

(bargaining disposition)

0.854

0.087

0.859

0.049

0.068

0.806

0.059

0.802

1.62

1.16

It is important for me: . . . that I can defend the technical contents of the committee decisions . . . that the committee reaches the technically best solution . . . to observe my negotiation mandate . . . to defend Denmark’s interest in the committee Eigen value

Note: All items are Likert scale statements to which the respondents could answer ‘agree’, ‘partly agree’, ‘neither agree nor disagree’, ‘partly disagree’ or ‘disagree’. Extraction method: principal component analysis. Rotation method: Varimax with Kaiser normalization.

Table 8.6 Regression analysis of the behavioural dispositions of the Danish comitology representatives Deliberative disposition

Discretion Issue complexity Political interest

Bargaining disposition

Full model

Reduced model

0.23 (0. 05)∗∗∗

0.23 (0. 05)∗∗∗ 0.24 (0. 06)∗∗∗ –

−0.08 (0. 06) 0.09 (0. 07) 0.16 (0. 06)∗∗

– – 0.18 (0. 06)∗∗∗

– – – –

0.13 (0. 12) 0.00 (0. 00) 0.15 (0. 11) 0.01 (0. 01)

– – – –

0.24 (0. 07)∗∗∗ 0.01 (0. 06)

Full model

Reduced model

Individual characteristics of the respondent: – Female – Age – Specialist – Committee experience Constant Adj. R2 N ∗

0.02 (0. 11) 0.00 (0. 00) −0.03 (0. 10) 0.01 (0. 01) 2.24∗∗∗ 0.23 152

2.46∗∗∗ 0.23 161

3.89∗∗∗ 0.05 152

p < 0.1; ∗∗ p < 0.05; ∗∗∗ p < 0.01. Dependent variables: Indices of behavioural disposition. Entries are OLS-regression coefficients (standard errors in parenthesis).

4.15∗∗∗ 0.05 160

The Comitology System in Daily Practice

165

the discretion index leads to an increase of one-fifth of a point in the index of deliberative disposition. While discretion has an impact on the representatives’ disposition for deliberative behaviour, it seems unrelated to their bargaining disposition. The control variables show mixed results. Issue complexity is related to deliberative dispositions, but not to bargaining dispositions. Political interest is related to bargaining dispositions, but not to deliberative dispositions. The individual characteristics of the respondents do not seem to play any role. In sum, although the analysis of step two has not allowed an analysis of the representatives’ actual behaviour, it has demonstrated a link between their instructions and their behavioural dispositions. Since the impact of discretion on actual behaviour is likely to go through behavioural dispositions, one of the causal mechanisms linking discretion to behaviour has been illuminated.

Step 3: The interaction style in the comitology committees The third step in the argument on the daily workings of the comitology committees is to link the behaviour by the national representatives with the interaction style in the committees. According to hypothesis 12, the representatives’ behaviour has an impact on the extent to which the committees have a bargaining and deliberative interaction style. This hypothesis is challenging to investigate because it requires data on the behaviour of all, not only the Danish representatives. Since these data do not exist, I proceed in a less ambitious way. Although I cannot investigate the determinants of the committees’ interaction styles, the data allow a descriptive analysis of the styles. In particular, I can analyse the extent to which a deliberative and bargaining style exists. Hypothesis 12 presupposes that a deliberative and bargaining interaction style exists. However, this cannot be taken for granted. The two interaction styles correspond to the deliberative and rationalist image of comitology, but since their empirical backing mostly consists of case studies of selected individual committees, the exact empirical prevalence of the two interaction styles is still an open question. A descriptive analysis also makes it possible to resolve an issue that is often left unclear in the literature: is the interaction style among participants in the comitology system a one-dimensional concept, the extreme values of which are deliberation and bargaining? Or are they different dimensions of interaction styles that should be treated as separate phenomena?

166 The EU Comitology System in Theory and Practice

The first challenge for a systematic study of deliberative and bargaining interaction styles is to define the two concepts in more detail and to specify empirical indicators. I now turn to this challenge. Measuring deliberative interaction. In a review of applied deliberative democratic theory, Chambers (2003) defines ‘deliberation’ as debate and discussion aimed at producing reasonable, well-informed opinions in which participants are willing to revise preferences in light of discussion, new information and claims made by fellow participants. Excluded from this definition is the result of the deliberative process. Many believe that deliberation leads to consensus, and often consensusoriented politics is often equated with deliberation (certainly, this is the case in debates on comitology). But, as stressed by several empirical students of deliberation (Lascher, 1996; Chambers, 2003; Steiner et al., 2004, pp. 25–6, 41–2), whether deliberation promotes toleration and understanding or disagreements and conflict should be treated as an empirical question. Consequently, the following analysis focuses on deliberative processes, not their results. At the heart of deliberation is talking. Deliberative politics is discursive or communicative politics. But there is more to deliberation than talking. The quality of the talking matters. High-quality deliberation consists of four central components (cf. Steiner et al., 2004, pp. 52–61). First, speakers must be able to participate freely in the debate. Second, speakers should justify their positions with explicit arguments. Third, arguments referring to the common good should carry greater weight. Finally, deliberation is constructive. It should be able to change minds and transform opinions. In the survey the Danish comitology representatives were asked about these four factors. Measuring bargaining interaction. In a review of realist, liberal intergovernmental and rational choice institutionalist studies of EU politics, Pollack (2001) argues that these approaches converge in a single rationalist model based on common assumptions, especially the notion that EU negotiations consist of a set of states in rational pursuit of relatively stable and well-ordered interests determined by national constituents. More specifically, four central components can be identified. First, states are self-interested. They act instrumentally on the basis of national interests. To Moravcsik (1998), the first step in international cooperation is the formation of national preferences which are determined by powerful domestic constituents. In Putnam’s (1988) terms, international negotiators’ ‘win-set’ depends on the preferences, coalitions and institutions of ‘level II’ (i.e., national) constituents. Second, as especially emphasized by realists, negotiations are power plays. The broad parameters of

The Comitology System in Daily Practice

167

inter-state cooperation are set by the most powerful actors in the international system. It is telling that Moravcsik’s (1998) influential account of the major turning points in the EU’s history does not include the three small Benelux countries or (after 1973) Ireland and Denmark, but focuses solely on Germany, France and Britain. Third, negotiation results are often obtained by compromises covering several issues (‘issue linkage’, ‘side payments’, ‘package deals’ etc.). This is an efficient negotiation form since it reaps the gains of cooperation by exploiting asymmetrical preferences among the negotiators. Finally, insincere behaviour should be expected due to information asymmetries among negotiators, who are often incompletely informed about each other’s domestic politics. In this situation, ‘negotiators have an incentive to understate their own win-sets’ (Putnam, 1988, p. 452). In the survey the Danish comitology representatives were asked about these four factors. Table 8.7 shows the empirical indicators of the deliberative and bargaining interaction styles in the comitology committees and the results of a factor analysis. The indicators clearly load on two dimensions, which strongly suggests that deliberation and bargaining are two separate dimensions of interaction style. Deliberation is not the opposite of bargaining, but a separate phenomenon that may, or may not, coexist with bargaining. To further explore the two interaction styles, I combine the first four statements in Table 8.7 into an index of bargaining interaction, and the last four statements into an index of deliberative interaction (see Table 8.2 for summary statistics). This leads to two findings. First, the indices vary almost across their full potential range, which means that there is a lot of variation in the interaction styles. Second, the indices are not correlated (r = −0.05; sign.:0.49), which means that a bargaining interaction style prevails in some committees, and a deliberative style in others. In some committees bargaining coexists with deliberation; in other committees neither interaction style prevails. This fact may help explain the mixed findings in case studies of interaction styles in individual committees. In sum, although the data do not allow a full analysis of step three in the causal chain constituting the argument on the daily workings of comitology, it has been possible to demonstrate that a deliberative as well as a bargaining interaction style exists in the comitology committees. This indicates that both the deliberate and the rationalist image of comitology may have merit. This is a premise of the argument, and a stylized fact in the literature, but a fact which so far has lacked a systematic empirical basis.

168 The EU Comitology System in Theory and Practice Table 8.7 Factor analysis of the Danish comitology representatives’ evaluation of the nature of meetings in their committee

In the meetings of my comitology committee: . . . compromises are normally political horse-trades . . . it is really the large countries that decide . . . the participants often resort to bluffing . . . national interests dominate our work . . . it happens that participants are persuaded by good arguments to change their position . . . arguments on common solutions are especially important . . . the participants normally present detailed arguments for their positions . . . all participants can freely express their opinion Eigen value

Loading on factor 1

Loading on factor 2

(bargaining interaction style)

(deliberative interaction style)

0.756

−0.115

0.729

0.064

0.701

−0.025

0.547

−0.113

0.103

0.683

0.055

0.673

−0.149

0.645

−0.147

0.547

2.02

1.60

Note: All items are Likert scale statements to which the respondents could answer ‘agree’, ‘partly agree’, ‘neither agree nor disagree’, ‘partly disagree’ or ‘disagree’. Extraction method: principal component analysis. Rotation method: Varimax with Kaiser normalization.

Step 4: Constraining the Commission? The final step in the analysis of the daily workings of the comitology committees is to investigate the connection between the interaction styles in the committees and the Commission’s behaviour. According to hypothesis 13, a deliberative interaction style makes it possible for the Commission to play a more active role in the committees. The effect of a bargaining interaction style is theoretically unclear, and I have not formulated a hypothesis on this question. What behaviour is the Commission likely to exhibit in the comitology committees? Case studies of selected individual committees suggest that

The Comitology System in Daily Practice

169

the Commission sometimes pursues its own agenda and sometimes takes on a more neutral mediating role. Joerges and Neyer’s (1997a) study of the Standing Committee for Foodstuffs showed that most representatives in this committee found that the Commission pursues its own interests; others found that the Commission acts as a neutral broker (p. 280). In their study of the Advisory Committee on Safety, Hygiene and Health Protection, Daemen and van Schendelen (1998) found evidence of an active Commission that sometimes completely ignores the position of the committee when formulating draft rules. Philip (1998) investigated the Ecolabel committee and found evidence of a cautious Commission. The ecolabel is commercially sensitive, and the national representatives succeed in dominating how the labelling system is structured. In a study of the EMAS Committee, Töller (1998) found ambiguous evidence on the Commission’s behaviour. The national representatives thought that the committee influences the Commission, but also that the committee does not really constrain the Commission’s discretion (pp. 197–8). What these case studies suggest is that the Commission’s behaviour in the comitology committees reflects its general behaviour in the EU system, which is well-known from a number of insightful studies. On the one hand, this literature portrays the Commission as the ‘engine of integration’, that is, a policy entrepreneur that actively pursues a supranational agenda and tries to direct the member states into ‘ever closer union’. On the other hand, this literature also portrays the Commission as an ‘honest broker’, that is, a loyal agent of the member states. It arbitrates between their conflicting preferences and tries to keep the community from falling apart. These roles have been identified by observers since the beginning of the Community. David Coombes’ (1970) classic study unveiled a ‘political or promotive’ role for the Commission, but also a more ‘bureaucratic or implementative’ one. In a similar vein, Page (1997, pp. 141–58) points out that the Commission sometimes acts as a ‘bureaucratic entrepreneur’, sometimes as a mere administrator of routine policies. Likewise, Nugent (2001, pp. 202–35) argues that the Commission provides leadership in the EU, but also that this role is often constrained by operational contexts such as member states’ preferences and the decision-making framework, which may force the Commission into a more advisory and administrative role. Against this background my strategy is to map the extent to which the Commission plays two roles in the comitology committees: a policyadvocating role and a mediating role. Having mapped these roles, I then

170 The EU Comitology System in Theory and Practice

investigate whether the interaction styles in the committees have an influence on how actively the Commission plays them. To measure the Commission’s behaviour, the Danish representatives were asked to characterize how the Commission behaves in their committee and indicate, based on a number of behavioural roles presented to them, how often the Commission plays these roles in their committee. Table 8.8 shows the empirical indicators of the Commission’s roles in the comitology committees and the results of a factor analysis. The indicators clearly load on two dimensions: a mediating and a policyadvocating role. I then use the indicators to construct indices of the roles. The first three indicators are combined into an index of a mediator, the last four into an index of a policy advocate. As can be seen from the summary statistics in Table 8.2, there is variation around the mean of these indices, suggesting that the Commission sometimes plays these roles in an active way, sometimes in a more passive way. Table 8.8 Factor analysis of the Danish representatives’ evaluation of the Commission’s roles in the comitology committees

How often does the EU Commission play the following roles in your comitology committee? . . . The one who makes sure that everybody feels comfortable . . . The one who makes sure that the result is satisfactory to everybody . . . The one who mediates . . . The one who prepares things . . . The one who investigates and explains things . . . The one who promotes his own proposals . . . The one who drives negotiations forward Eigen value

Loading on factor 1

Loading on factor 2

(mediator)

(policy advocate)

0.837

0.020

0.825

0.043

0.750 0.167

0.306 0.780

0.280

0.712

−0.295

0.708

0.394

0.509

2.76

1.51

Note: All items are Likert scale statements to which the respondents could answer ‘always’, ‘often’, ‘now and then’, seldom’ or ‘never’. Extraction method: principal component analysis. Rotation method: Varimax with Kaiser normalization.

The Comitology System in Daily Practice

171

The question is whether Commission activism is related to the interaction styles in the committees. This is investigated by regressing the measures of the Commission’s roles on the measures of the interaction styles and control variables. In addition to the usual controls for individual characteristics, I control for two things. First, for the comitology procedures, which makes it possible to investigate whether the effect of interaction styles is spurious because they correlate with these procedures, but also whether the procedures affect the Commission’s activism. Second, I use fixed effects for the Commission’s DGs, which vary in terms of discretionary powers, prestige and reputation and organizational culture (Page, 1997; Cini, 1997; Hooghe, 2001). These factors may cause the Commission’s representatives in the comitology committees to behave in different ways depending on their institutional affiliation.4 However, disentangling the independent effects of the comitology procedures and the DGs is hampered by a collinearity problem since comitology procedures are not randomly spread across the DGs. For example, DG Agriculture almost only uses the management procedure, while DG Environment almost only uses the regulatory procedure. But since the main function of the DGs and comitology procedures in the present analysis is to serve as control variables, I refrain from further investigation of their relative importance. The present analysis of their effects should therefore be considered preliminary and in need of further refinement. The analysis of the impact of the committees’ interaction styles on the Commission’s behaviour is reported in Table 8.9, which shows the results both with and without the fixed DG effects. The analysis of the Commission’s mediating behaviour shows that it is strongly influenced by a deliberative interaction style. The more deliberative the interaction style, the more active the mediating behaviour of the Commission. The impact is not only strongly statistically significant, but also substantially important. Both variables are measured on a five-point scale, and a onepoint increase in the measure of deliberative interaction style leads to an increase in mediating behaviour of half a point. Including the fixed DG effects in the analysis does not affect this result. No control variables are related to the mediating role in a statistically significant way. The analysis of the Commission as a policy advocate in the last two columns of Table 8.9 shows that this role is influenced by a bargaining interaction style. Compared with the effect of the deliberative interaction style on the mediating role, however, the substantial magnitude of this effect is much smaller. Furthermore, this result is not robust to alternative specifications of the analysis, see further endnote 4, so it should be interpreted with caution. The policy-advocating role also seems to be

172 The EU Comitology System in Theory and Practice Table 8.9 Regression analysis of the Commission’s behaviour in the comitology committees Mediator Model 1 Deliberative interaction style Bargaining interaction style

0.53 (0. 10)

Policy advocate

Model 2 ∗∗∗

−0.11 (0. 08)

Model 1 ∗∗∗

0.54 (0. 11)

−0.09 (0. 08)

−0.02 (0. 07) 0.13 (0. 05)∗∗

Model 2 0.04 (0. 07) 0.11 (0. 05)∗∗

Comitology procedures: – Management procedure – Advisory procedure – Regulatory procedure – Several procedures

(ref.cat.)

(ref.cat.)

(ref.cat.) −0.18 (0. 21)

(ref.cat.)

−0.05 (0. 30)

0.01 (0. 32)

−0.02 (0. 13)

−0.10 (0. 15)

−0.18 (0. 09)∗∗

−0.05 (0. 09)

−0.10 (0. 14)

−0.20 (0. 16)

−0.10 (0. 09)

−0.13 (0. 10)

0.12 (0. 12) 0.01 (0. 01) 0.00 (0. 11) 0.00 (0. 01)

0.12 (0. 12) 0.01 (0. 01) 0.06 (0. 13) 0.00 (0. 01)

−0.07 (0. 08) 0.01 (0. 00) −0.04 (0. 08) −0.00 (0. 01)

−0.07 (0. 08) 0.01 (0. 00)∗∗ −0.06 (0. 08) −0.00 (0. 01)

0.02 (0. 20)

Individual characteristics of the respondent: – Female – Age – Specialist – Committee experience Fixed DG effects?a Constant Adj. R2 N

NO 1.44∗∗ 0.14 149

YES 1.41∗∗ 0.13 149

NO 3.94∗∗∗ 0.05 149

YES 3.32∗∗∗ 0.20 149

∗

p < 0.1; ∗∗ p < 0.05; ∗∗∗ p < 0.01. Dependent variables: Indices of the Commission’s behaviour. Entries are OLS-regression coefficients (standard errors in parentheses). Note: a Fixed DG effects include: DG Enterprise and Industry, DG Agriculture, DG Transport and Energy, DG Environment, DG Research, DG Taxation and Customs Union, DG Health and Consumer Protection, and a residual category of remaining DGs which all have fewer than ten observations.

influenced by the comitology procedures. The strict regulatory procedure seems to have a dampening effect on the Commission’s behaviour, but not surprisingly this effect disappears when the fixed DG effects are included. Their inclusion means that the analysis is restricted to investigating variation within DGs. Since comitology procedures primarily vary across rather than within DGs, it is uncertain whether the dampening effect is really caused by the regulatory procedure or some other DG

The Comitology System in Daily Practice

173

characteristic such as organizational culture. Further analysis is required to settle this question. In sum, the analysis of step four in the argument on the daily workings of the comitology committees shows that a deliberative interaction style in the committees has an impact on Commission activism. The analysis focused on two types of Commission behaviour suggested by case studies of selected individual comitology committees, namely mediating and policy-advocating behaviour. This is behaviour that rhymes with the general literature on the Commission which for many years has discussed the Commission’s role as either an ‘honest broker’ among the member states or as the entrepreneurial ‘engine of integration’. As predicted by hypothesis 13, a deliberative interaction style creates room for Commission activism. However, this does not take the form of policy-advocating behaviour, but of active mediation.

Conclusion This chapter has studied how the comitology system functions in daily practice. Using a data set from a survey of all Danish comitology representatives, it has investigated the four-step argument developed in Chapter 3. The first step was to inspect the representatives’ instructions, the second to look at the representatives’ behaviour in the comitology committees, the third to analyse the interaction styles in the committees and the fourth to investigate the Commission’s behaviour in the committees. Before drawing final conclusions, the limitations of the research design should be reiterated. In step one it was not possible to analyse the impact of conflict between the national government and the comitology representatives. In the Danish case, conflict is a constant since the government uses civil servants as representatives in all committees. Conflict was thus controlled for, but not actively investigated. In step two it was not possible to analyse the representatives’ actual behaviour, but only their behavioural dispositions. This, however, is a causal mechanism that links discretion to actual behaviour. In step three it was not possible to investigate the determinants of the interaction style in the committees as this would require data on the behaviour of all, not only the Danish, representatives. However, the data made it possible to clarify whether both a bargaining and a deliberative interaction style really exist in the comitology meeting rooms in Brussels. This is a stylized fact in the literature and a precondition for step three. Finally, in step four it was possible to link interaction styles in the committees to the

174 The EU Comitology System in Theory and Practice

Commission’s behaviour. However, behaviour is an output, rather than outcome, type of variable. In the final analysis, the nature of implementation measures rather than Commission behaviour is the interesting effect of the comitology system. Given these limitations, conclusions should be drawn with caution, but the following seem justified. First, the basic logic of the fourstep argument seems sound. That is, since the committee participants are representatives of national governments, an understanding of the daily operation of the comitology system must start in the member state capitals, not in Brussels. The analysis of step one demonstrated that the representatives operate with varying discretion, which at least partly is a result of a well-known factor from the delegation literature, namely issue complexity. Step two showed that discretion is related to behavioural dispositions. Representatives allowed discretion are more inclined towards deliberative behaviour. The proposition that the representatives’ instructions have an impact on what is going on in the meeting rooms of the comitology committees is thus strengthened. Step three showed that deliberative and bargaining interaction styles are distinct interaction styles that may, or may not, coexist in the comitology committees. Finally, the analysis of step four found that the interaction styles have an effect, at least on the activism exhibited by the Commission in the committees. Returning to the five hypotheses on the daily operation of the comitology system, this chapter has provided supporting, although also mixed, evidence for hypotheses 9, 11, 12 and 13. Due to lack of data, it was not possible to investigate hypothesis 10. The evidence thus increases our understanding of the two images of the daily workings of the comitology system. It not only demonstrates the empirical validity of the two images, but also helps us understand their causes and effects. Having said that, there is an obvious need for further research on the daily operation of the comitology system, and two avenues seem especially promising. The first is to collect survey data of the sort presented in this chapter from more countries. Several factors make this a worthwhile project. First, it would increase the reliability of empirical indicators. The data set in this chapter has relied on one observation per committee. More would obviously be desirable. Second, it would reduce concerns of country bias in the data. As mentioned at the beginning of the chapter, I believe that this concern is easily exaggerated. Gijs Jan Brandsma has collected comparable data from Dutch comitology representatives, and our joint analysis of Dutch and Danish data shows

The Comitology System in Daily Practice

175

that many of the relationships discussed in this chapter hold with cross-national data (Blom-Hansen and Brandsma, 2009; Brandsma and Blom-Hansen, 2010). Data from more countries would obviously eliminate the country bias concern. Third, more survey data would make it possible to uncover the instructions and behaviour of all representatives and to directly link this to what is going on in the committee meeting rooms in Brussels. The second avenue is a mixed-method strategy, that is, to combine large-N quantitative analysis with small-N comparative case studies (Lieberman, 2005). Quantitative analysis of the kind used in this chapter is useful for uncovering correlations, which is a precondition for, but not a demonstration of causality. Furthermore, the survey data measure independent and dependent variables at the same point in time, which makes the direction of causality questionable. Strategically selected case studies would make it possible to inspect causal mechanisms and thus provide further evidence on causal relations. One option would be to study a limited number of committees whose score, according to the quantitative analysis, on the Commission’s behaviour is well-predicted by their score on the measures of interaction style. Such a study could clarify whether this association is truly causal and, if so, in which direction causality runs. Once causality has been established, an analysis of off-the-regression-line cases could be useful for uncovering additional potential determinants of Commission behaviour in the comitology committees. The generalizability of these potential extra determinants could then be investigated in future quantitative analyses.

9 Conclusion

In this concluding chapter I have four purposes. The first is to make a succinct presentation of the main empirical findings of the book and explicate its contribution to the empirical study of the comitology system. The second purpose is to draw out the theoretical lessons of the book. I have applied a delegation perspective on the comitology system, but as I made clear in the empirical chapters there are alternative theoretical perspectives on all the questions analysed. I now settle the score and evaluate the relative value of the different theories as well as the book’s contribution to the delegation perspective. The third purpose is to discuss how the empirical study of the comitology system can be extended. While the book advances our understanding of the comitology system, many aspects of the system remain under-researched and the book provides a basis for discussing a relevant research agenda. Finally, I conclude the book by discussing its normative implications. The comitology system is often criticized for intransparency and complexity and for increasing the EU’s democratic deficit. I find this criticism misplaced. In the final section I discuss the relationship between the comitology system and the democratic deficit. I discuss this problem in general, but with a particular eye to the changes brought to the comitology system by the Lisbon Treaty.

Main empirical findings I set out to find answers to three fundamental unresolved questions on the comitology system. The first concerned the origins of the comitology system. Parliamentary control of delegated rule-making by the executive is a well-known problem in many political systems, but solving it by a comitology system is not, as the system is unique to the 176

Conclusion

177

EU. Why was the comitology system created? The answer given in the literature is a functional one. The system was established as a response to the Council’s need to delegate implementing powers to the Commission without losing control. However, this functional explanation is not entirely convincing since it overlooks the importance of choice. Why did the member states choose the comitology system when other solutions could easily be imagined, for example solutions used in their own national systems to solve similar problems? Building on the delegation literature, I argued that the member states in the early 1960s were caught in a coordination problem that could be solved by a carefully constructed focal point. In Chapter 4, I demonstrated that the member states delegated informal agenda-setting power to the Commission, and that the Commission, not the member states, proposed committees as a solution. To the member states this represented an obvious solution to their control problem. It was a focal point, and a powerful one. It simplified matters and reduced the complex control problem to one of more or less comitology. Once introduced, it functioned as a precedent not only in agriculture, where comitology was first installed, but also in other policy areas. When the same problem of supranational administration and national control arose in relation to development aid to Africa and customs, the member states found solutions similar to the agricultural committees and never seriously considered alternative solutions, for example well-known solutions to similar problems in their own national systems. In contrast to the existing literature my explanation focuses on the Commission, not the member states, as the decisive actor. The second question concerns the driving forces of the comitology system. Almost all observers are struck by the inter-institutional rivalry about the design of the system. But what exactly is at stake? I wanted to understand the development of the comitology system over time and investigated, over three chapters, the interests of the negotiating actors, their strategies and their instruments to pursue interests. I started by analysing the institutional comitology preferences of the Council, the Commission and the European Parliament – the three primary actors in the comitology system. In Chapter 5, I conducted a case study of the 2006 comitology reform and demonstrated that the Council favours strict comitology procedures while the Commission prefers permissive procedures. The European Parliament agrees with the Commission, but also wants to narrow delegation to technical matters or to gain access to the comitology system. The case study also made it possible to flesh out in considerable detail what these institutional interests mean in practice

178 The EU Comitology System in Theory and Practice

and how the actors can pursue them. The case study of the 2006 reform represents a most likely case for revealed institutional preferences since it represents an obvious opportunity for the actors to improve their control over delegated decision-making. In Chapter 6, I conducted a least likely study of revealed institutional comitology preferences by investigating their importance in daily legislation. This is a hard case because the actors are likely to be strategic in their choice of comitology procedure in daily legislation. This raises a problem of anticipated reactions, which makes open disagreement less likely. The chapter studied all new directives and regulations made under the 1999 comitology decision – a total of 686 acts. It found that open disagreement on comitology procedures exists in less than 10 per cent of all these acts. But due to anticipated reactions it is difficult to evaluate whether this high degree of apparent agreement indicates a true convergence of preferences or a strategic use of preferences. There is, in other words, a problem of observational equivalence. Consequently, conclusions had to be based on disagreement cases, and they exhibited a pattern of institutional preferences similar to the case study of the 2006 reform. The driving forces of the comitology system were further explored in Chapter 7, where I analysed the variation in the use of comitology procedures across policy areas. Building on the delegation literature I argued that institutional conflict and issue complexity are key factors shaping the EU legislators’ choice of comitology procedures. Analysing the data set of the 686 new directives and regulations made under the 1999 comitology decision, I demonstrated that both factors have a positive impact not only on the introduction of comitology procedures, but also on their stringency. In sum, Chapters 5, 6 and 7 point to two fundamental driving forces of the comitology system. First, the system is born out of conflict as the result of a power struggle among actors who fundamentally disagree on its design. Second, the system is born out of necessity. Some regulatory areas are so complex that the Council and the Parliament need to delegate powers to the Commission, but they can impose comitology procedures to minimize agency loss. The third and final question concerns the daily operation of the comitology system. Most observers agree that the system is put in place to control the Commission. However, several case studies of the daily operation of selected individual comitology committees show that the committees do not always function as guardians of national interests, but sometimes as forums where policy experts meet and deliberate to find the best or most efficient solutions to common problems. Informal norms, deliberation and good arguments seem to matter more than

Conclusion

179

national interests and formal voting rules. These findings are puzzling because they indicate that even though the member states spend considerable time and energy installing control mechanisms, they do not use them once they are in place. Based on the delegation literature, I argue that to understand the daily operations of the comitology system, we need to start by inspecting the instructions given to national representatives. Representatives armed with clear instructions from their home government are likely to defend their negotiation mandate by bargaining, while representatives who are allowed some latitude by their home government can better afford to engage in deliberation. So, depending on the representatives’ instructions, the comitology committees are likely to be characterized by a bargaining and deliberative interaction style to different degrees. This again influences the Commission’s manoeuvrability since deliberative committees provide better opportunities for Commission activism. In Chapter 8, I used data from a survey of representatives on all comitology committees from one member state – Denmark – to trace the representatives’ instructions and to investigate their impact on behaviour, interaction styles and Commission activism in the individual comitology committees. This analysis confirmed that an understanding of the daily operation of the comitology system must start in the member state capitals, not in Brussels. The representatives operate with varying discretion, which at least partly is a result of a well-known factor from the delegation literature, namely issue complexity. Further, the representatives’ discretion seems to matter for their behavioural dispositions. Representatives allowed discretion are more inclined towards deliberative behaviour. Committees dominated by representatives operating with discretion are therefore likely to be characterized by a deliberative interaction style, whereas committees with tightly instructed representatives are likely to exhibit a bargaining interaction style. However, while the link between instructions and behavioural dispositions could be established empirically, analysing the link between behavioural dispositions and interaction styles requires data from all member states, not only Denmark. But the Danish data could demonstrate that deliberative and bargaining interaction styles are distinct and existing interaction styles that may, or may not, coexist in the comitology committees. Finally, it was possible to link interaction styles in the committees to Commission activism. In sum, my findings make sense out of the mixed findings of case studies since they show and explain why some committees are deliberative forums, others bargaining arenas.

180 The EU Comitology System in Theory and Practice

The theoretical lessons of the book The book has applied a delegation perspective on the comitology system and used this literature to derive hypotheses on the three questions: Why does the comitology system exist? What are the driving forces of the system’s development over time? How does the system work in daily practice? The overall argument is that these questions represent delegation challenges to the member states. They seek to derive the benefits of cooperation and delegation without losing control. This involves finding a balance between delegating powers to the Commission and controlling it. The comitology system is the instrument. The delegation perspective is not the only theoretical lens applied to the comitology system. Contenders to the delegation perspective in relation to all three questions have been presented and discussed in the relevant chapters. Now is the time to make a concluding evaluation of their relative merits. In relation to the origins of the comitology system – the first question addressed – the functionalist perspective presented in Chapter 1 and discussed further in Chapter 4 argues that the comitology system was created to enable the member states to delegate power to the Commission without losing national control. My analysis in Chapter 4 confirmed this explanation. The system was indeed created by the Council to monitor the Commission. But the functionalist perspective only provides a theoretical tool for understanding half the story. It is correct on the motives behind the comitology system, but it cannot explain why comitology was chosen among the many potential solutions to the Council’s delegation problem. In contrast, my focal point explanation, derived from the delegation perspective, can explain the exact choice of comitology as the institutional solution to the member states’ problem. In relation to the driving forces of the comitology system – the second question – the expertise explanation presented and discussed in Chapter 7 argues that comitology committees are established to give the Commission access to expertise in the member states’ national administrations. This is necessary because, so the argument goes, the Commission does not possess sufficient administrative capacity and needs to rely on the member states’ central administrations. According to this perspective, the driving force of the comitology system is empowerment of the Commission. My analysis in Chapter 7 showed that there is a link between issue complexity and the instalment of a comitology committee. This is consistent with the expertise explanation, but the analysis also showed that there is a link between institutional conflict

Conclusion

181

and the instalment of a comitology committee. It further showed that both issue complexity and institutional conflict have an impact not only on the instalment of a comitology committee, but also on the stringency of comitology control. While the expertise explanation offers no understanding of these findings, they are consistent with my control explanation based on the delegation perspective. The expertise explanation may therefore offer a partial understanding of the driving forces of the comitology system, but my control explanation seems to offer a more complete understanding. In relation to the daily operations of the comitology system – the third and final question – the deliberative image presented in Chapter 1 and discussed further in Chapter 8 argues that comitology is government by persuasion, argumentation and discursive processes rather than by command, bargaining and protection of national interests. My analysis in Chapter 8 confirmed that some committees indeed function in this way, but the deliberative perspective is far from offering a full story. First, it offers no understanding of why some committees are deliberative. This is puzzling because the member states have spent considerable time and energy installing the committees as watchdogs. Second, it offers no understanding of why only some committees function in this way. My explanation, based on the delegation perspective, offers an understanding of these questions. It argues that the interaction styles and the constraining effect of the comitology system can be traced back to instructions given to national representatives. There is a long causal chain involved in this argument, and my data did not allow a final empirical investigation. But the evidence seems sufficient to conclude that the explanation must start in the member state capitals, not in Brussels. The key is the representatives’ instructions. While the deliberative perspective thus may offer an understanding of some aspects of the daily operation of the comitology system, my delegation perspective offers a more complete explanation. In sum, functionalist reasoning on the origins of the comitology system, expertise explanations of the system’s development over time and deliberative perspectives on the daily operation of the comitology system all offer insights into the comitology system and enrich our understanding of it. However, they represent partial explanations of selected aspects. The delegation perspective offers a coherent and more complete explanation. While the delegation literature thus has a lot to offer to our understanding of the comitology system, it is also relevant to turn the question upside down. What can the book’s empirical study of the

182 The EU Comitology System in Theory and Practice

comitology system offer the delegation literature? At least four contributions deserve to be stressed. First, as to classification of oversight mechanisms it has, since McCubbins and Schwartz (1984), been standard to use the distinction between police patrol and fire alarm oversight. The former represents active and centralized oversight, the latter reactive and decentralized oversight. As argued by Ogul and Rockman (1990), this typology conflates the dimensions of activism and centralization of oversight mechanisms. They therefore advocate a four-cell classification scheme based on the dimensions of centralized/decentralized and active/reactive oversight. The comitology system is fragmented into several hundred committees, but still represents proactive and systematic control. The book’s demonstration that comitology is first and foremost about controlling the Commission therefore provides a clear example of a decentralized police patrol system. This finding confirms the relevance of Ogul and Rockman’s (1990) modification of McCubbins and Schwartz’s (1984) original classification scheme. Second, the book adds insight into legislators’ exact choice of oversight mechanisms; a question that has received surprisingly little attention in the delegation literature. Empirical delegation studies tend to take existing oversight mechanisms for granted and explore how they function or how legislators employ them in given delegation situations. For instance, Epstein and O’Halloran (1999, pp. 99–102) identify 14 procedural constraints in their study of delegation provisions in major US laws, but do not investigate why the menu of procedural constraints looks exactly like this. Franchino (2007, pp. 91–6) follows the same approach in his study of delegation in the EU system. Likewise, Huber and Shipan (2002, pp. 56–65) identify 12 procedural controls in their study of delegation in the Medicaid programme in US states, but do not investigate their origins. The standard approach is to take the menu of oversight mechanisms for given. My study of the origins of the comitology system in Chapter 4 stresses the value of actively investigating the origins of oversight mechanisms. In my case, it led to a revision of the standard view of the system’s origins. It was not born by the member states, but by the Commission. This finding underscores how delegators and delegatees may interact in the design of oversight mechanisms. Third, the book’s study of institutional comitology preferences in Chapters 5 and 6 underscores the importance of control considerations when legislators delegate power to the executive – a recurrent theme in the delegation literature. Legislators seek to obtain the benefits of delegation, but worry about losing control (Kiewiet and McCubbins,

Conclusion

183

1991, pp. 22–39; Bawn, 1995). The analysis of comitology preferences showed that both the Council and the European Parliament first and foremost worry about control positions in relation to delegated powers. This finding is a confirmation of Moe’s (1990a, 1990b) argument that legislators try to freeze politics in time by designing administrative structures and McCubbins et al.’s (1987, 1989) argument that legislators try to ‘stack the deck’ in favour of their constituencies by carefully designing implementation procedures for delegated powers. Finally, the book confirms that conflict and issue complexity are crucial factors in legislative decisions on delegation. These determinants of delegation decisions have been in focus in the literature at least since McCubbins’ (1985) theoretical model. Subsequent empirical studies have confirmed their importance in both a US context (Epstein and O’Halloran, 1999; Huber and Shipan, 2002, pp. 139–71) and an EU context (Pollack, 2003a; Franchino, 2007). Chapter 7’s analysis of the variation in the use of comitology control across policy areas lends further support. Institutional conflict and issue complexity seem very robust explanations of legislative delegation decisions.

A research agenda This book has addressed three prominent unresolved questions on the comitology system. Why does the comitology system exist? What are the driving forces of the system’s development over time? How does the system work in daily practice? By providing answers, the book has enhanced our understanding of the comitology system, but the system is still to a large extent a white spot on the EU map. Many aspects remain under-researched, and there is a glaring imbalance between the impressive amount of research into ordinary EU legislative decisionmaking and the limited research into executive EU rule-making by the comitology method. To specify the most relevant areas of future research I return to the four-level institutional structure of the comitology system introduced in Chapter 2. The first level – the treaty level – specifies how rules on the operation of the system are made. This level has not received a lot of attention in the comitology literature, but the limited research that exists (Demmke et al., 1996; Haibach, 2000; Bergström, 2005, pp. 178–84, 335–51) suggests that the delegation provision inserted in the EC Treaty’s article 202 by the Single European Act and the hierarchy of legal acts introduced by the Lisbon Treaty (but originally proposed by the European Convention for the Constitutional Treaty) played a limited role in the

184 The EU Comitology System in Theory and Practice

intergovernmental negotiations leading to these treaty revisions. However, this is not necessarily an indication that comitology and delegated decision-making are uncontroversial at this level. On both occasions, the Commission and the European Parliament pressed for more executive autonomy, but the member states were reluctant to grant this. They preferred broad and flexible rules at the treaty level. This approach in effect pushes the institutional battle over comitology down to the next institutional level. In this sense the treaty has not had a dramatic impact on the development of the comitology system, and this is why this book has not investigated the treaty level. However, I stress that this interpretation rests on limited empirical evidence, so more research is warranted. The second level in the comitology system consists of the framework rules: that is, rules on committee procedures, time limits, the involvement of the Council and the European Parliament. These rules provide a menu for the legislators to choose from in daily legislation. Framework rules have been made since the Single European Act. Until the Lisbon Treaty they were made as Council decisions adopted under the consultation procedure. Now they are made as regulations adopted under the codecision procedure. My analysis of the 2006 comitology decision in Chapter 5 and other research of this and prior decisions show that the framework rules are objects of major institutional battles. This is where the Commission, the Council and the European Parliament settle the rules of the comitology game. Investigations of prior framework rules – the 1987, 1999 and 2006 comitology decisions – provide a good basis for understanding what goes on at this level. However, more research and continued scholarly attention to this level are in high demand because the Lisbon Treaty makes the European Parliament a full co-decider and thus changes the actors’ bargaining powers. The third level in the comitology system is daily legislation where the Council and the European Parliament decide whether to install a comitology committee when they delegate power to the Commission in individual legislative acts. This book’s analysis of daily legislative practice under the 1999 comitology decision shows that comitology seldom causes disagreement at this level. This is hardly because of an alignment of preferences, but more likely because the actors know each other’s preferences and anticipate them in their proposals. Since the Lisbon Treaty changes the actors’ bargaining powers by introducing codecision in more areas (e.g., agriculture) and by introducing more control mechanisms of delegated powers (rights of revocation and opposition under article 290), it is likely to change the status quo. The

Conclusion

185

European Parliament is likely to set its mark on comitology in legislative acts, moving from the consultation to the codecision procedure, and to closely guard the distinction between delegated acts and implementation acts. It is likely to be a subtle change of the status quo, however, since both the Commission and the Council know the Parliament’s preferences and are likely to anticipate them. Creative research is needed to trace this change at level three in the comitology system. The fourth and final level is the daily operation of the individual comitology committees. This book has made some progress in understanding what goes on at this level by analysing survey data from comitology representatives from one member state. But, as I argued at length in Chapter 8, more research is in high demand. We need evidence from more member states, and we need data triangulation and ‘mixed-method’ (Lieberman, 2005) combinations of case studies and quantitative analyses. In sum, much remains to be done. Comitology procedures have developed into the standard control mechanism of delegated powers in the EU system. It would seem to be worthy of more scholarly attention than it has received so far.

Comitology, the democratic deficit and the Lisbon Treaty The discussion of the democratic credentials of the EU has a long story. Worried observers especially point at the following problems (Scharpf, 1999, pp. 6–13; Moravcsik, 2002; Follesdal and Hix, 2006; Majone, 2009, pp. 151–79). First, given the historical, linguistic, cultural, ethnic and institutional diversity of the member states, there is not enough of a collective identity to constitute a European demos. Second, the institutions in the EU system are not representative, transparent or genuine. The Commission, which enjoys a powerful role as agenda-setter and initiator of the legislative process, is not elected. The national ministers in the Council deliberate in secret. The European Parliament, the only directly elected institution, is the result of second-order elections in which an ever-diminishing number of voters participate and select among national parties on the basis of national issues. Third, the decision-making procedures are unnecessarily complex, allocate considerable power to the unelected Commission through its right of initiative (and thereby an effective pre-veto over EU legislation) and make it difficult to hold actors accountable. However, observers disagree about whether these features add up to a democratic deficit. Proponents and opponents are split according to

186 The EU Comitology System in Theory and Practice

the weight they place on arguments of input and output legitimacy. The former refers to the democratic character of the political decisionmaking procedure (government by the people), while the latter refers to the degree to which the substance of political decisions promotes collective interests (government for the people). Observers who value input legitimacy often compare the EU to a national political system and judge the EU according to the same democratic ideals. Using this yardstick one quickly reaches the conclusion that the EU suffers from a democratic deficit. Proponents of input legitimacy argue that mechanisms must be established that more directly link outcomes with citizen preferences. They deplore the lack of an effective opposition in the EU political system. Because if voters do not have a specific sense of an alternative leadership or policy agenda, it is difficult for them to determine whether leaders could have done better or identify who is responsible for policies. As argued by Follesdal and Hix (2006, p. 548): Consider those who favour an alternative set of policy outcomes to the current policies of the Commission, the Council and the parliament. As the EU is currently designed, there is no room to present a rival set of leadership candidates (a government ‘in waiting’) and a rival policy agenda. . . . Indeed, it is precisely because there is no visible quasi-official ‘opposition’, that citizens cannot distinguish between opposition to the current EU policy regime and opposition to the EU system as a whole. Further, advocates of input legitimacy also call for more effective competitive elections in the EU. This is important both for giving politicians incentives to be responsive to the preferences of the citizens and for fostering political debates about the best means and objectives of policies. Debates promote the formation of public opinion and therefore have an important function in relation to preference formation among voters. Finally, input procedures matter not only because they link voter preferences and political decisions, but also because they promote accountability. Voters should be able to sanction politicians who are incompetent or whose policies prove flawed. Advocates of input legitimacy point to several practical solutions to the EU’s democratic legitimacy problem. First and foremost, they argue for more powers to the European Parliament, but they also propose measures such as more transparency in the Council, direct election of the Commission President, and extension of the right of legislative initiative

Conclusion

187

to the Council and European Parliament (Majone, 1998, p. 7, 2009, p. 33; Follesdal and Hix, 2006, pp. 553–6). In contrast to observers who value input legitimacy, advocates of output legitimacy do not start by comparing the EU to a national political system. Their point of departure is the multi-level nature of the EU system. They argue that the member states have delegated a specific set of regulatory functions to the EU: the creation of a single market for products and services, rules for the movement of factors of production, harmonization of product standards, health and safety rules, trade-related environmental and consumer policy and monetary policy. The purpose of the EU is to correct market failures and produce Pareto-efficient outcomes, rather than to engage in redistribution. These specific functions have been delegated to the EU to deliberately isolate them from domestic majoritarian government and to enhance the credibility of long-term policy commitment. For the same reason these functions are often delegated to non-majoritarian agencies at the national level. In this view, the Commission is a kind of super-agency, and politicization would result in redistributive rather than Paretoefficient outcomes and so in fact undermine the legitimacy of the EU. Advocates of output legitimacy add that the EU’s functions are complex and require expert advice. Deliberation among experts should thus play a large role in policy formulation. Furthermore, the EU’s regulatory functions have, in the minds of voters, low salience. This is also the case for similar functions at the national level. High salience areas such as welfare, education and health policy have simply not been delegated to the EU. Hence, the lack of voter interest in the EU is no surprise and constitutes no particular problem (Majone, 1998, 2009, pp. 151–204; Moravcsik, 2002). Advocates of output legitimacy also argue that since the EU is a multilevel entity, democratic impulses do not only go from the national level to the EU, but also in the other direction. Delegation to the EU may enhance democracy at the national level in at least three ways. First, it may empower the public by more effectively combating special interests. In areas such as trade policy where concentrated and well-organized losers from liberalization often dominate diffuse and less organized winners from liberal trade, such as consumers, it may be advantageous to leave control over the trade agenda to more insulated leaders with a broader and longer-term mandate. Second, the EU may enhance democracy at the national level by protecting individual and minority rights. For instance, according to the Copenhagen criteria, a state must respect human rights and protection of minorities to be eligible to join the

188 The EU Comitology System in Theory and Practice

EU. Finally, pooling resources in the EU may enlighten national policies by expanding the range of information available to national politicians and publics. The expertise available in the EU does not only feed into EU policies, but also back into the decision-making processes in the individual member states (Keohane et al., 2009). To increase the democratic legitimacy of the EU, proponents of output legitimacy point to means that are very different from those proposed by the advocates of input legitimacy. The major danger, in the eyes of output-oriented scholars, is the continuous expansion of EU powers. The solution lies in limitation and more precise definition of the powers transferred to the EU level. In particular, the Commission should be freed from its ‘political’ functions such as initiating legislation and preparing budgets and multiannual work programmes and focus on its role as a supranational agency responsible for specific regulatory functions (Majone, 2009, pp. 9, 165–73; Moravcsik, 2002). Reconciling the views of input- and output-oriented observers may seem insurmountable, or at least to require a great deal of pragmatism. This is exactly what Bellamy (2010) offers. Although an adherent of input standards, he nevertheless acknowledges that since the EU is primarily an organization for cooperation between democratic states, rather than a mechanism for transcending them, the output legitimacy may be sufficient. Non-majoritarian mechanisms are legitimate, but only as long as the EU is primarily an arrangement for interstate regulatory cooperation, and as long as the non-majoritarian mechanisms are controlled by the majoritarian systems of the member states (Bellamy, 2010, pp. 15–16). In essence, Bellamy argues that the EU for the present – that is, as long as it has not developed into a fully-fledged political system – can be justified by arguments of output legitimacy. This argument underscores how the debate on the EU’s democratic legitimacy is to a large extent really a debate about the true nature of the EU system. Is it a political system in its own right, or is it a cooperative arrangement for sovereign nations? Bellamy’s middle ground seems especially relevant for the comitology area. From an input perspective, the comitology system is easily criticized. Its bewildering complexity makes accountability vanish and transparency non-existent (cf. also the introduction to Chapter 2). Comitology has even been characterized as a democratic ‘netherworld’ (Flynn, 2000, p. 82). However, it is also possible to defend the comitology system from an input perspective. This can be done in three ways. First, formal votes are taken in the comitology committees. This is a well-known democratic procedure which secures a certain

Conclusion

189

degree of accountability. Second, the comitology system is now much more open than in the past. The Commission now publishes an annual detailed report on the comitology system (e.g., Commission, 2009a). In addition, there is now an online comitology register that provides detailed information on agendas for meetings in comitology committees, attendance lists, summary records, vote results and draft measures (http://ec.europa.eu/transparency/regcomitology/index.cfm). Although these transparency measures may not function optimally (Brandsma et al., 2008), they still represent a step forward. Third, the role of the European Parliament, the preferred institution of the input scholars, has been considerably enhanced in the comitology system over the past 10–15 years. The Parliament has fought a battle since the early 1960s with the Council over influence on the comitology system (Bradley, 1992, 1997, 2008; Bergström, 2005). However, this did not pay off until the late 1990s. The 1999 comitology decision gave the Parliament some insight and supervisory powers. For the first time it was now to receive information and to exercise some control of ultra vires behaviour by the Commission. Seven years later the 2006 comitology decision gave the Parliament veto rights over comitology decisions. Under the new regulatory procedure with scrutiny not only the Council, but also the European Parliament might block the Commission’s proposals. Three years later the Lisbon Treaty, entering into force in December 2009, introduced the distinction between delegated and implementation acts. For delegated acts the Parliament was granted the same parliamentary control mechanisms as the Council. For implementation acts the Lisbon Treaty foresaw a new comitology system and changed the decision rule for making rules on this system from the consultation to the codecision procedure and thus placed the Parliament on an equal footing with the Council also in this field. In sum, the Parliament’s influence on the comitology system has been steadily increasing over time – a development that advocates of input legitimacy are likely to welcome. In sum, judged from an input perspective, the legitimacy of the comitology system is questionable, but increasing over time. However, the value of input arguments in the comitology area can be questioned. Especially, the increasing influence of the European Parliament appears ill-founded. Two lines of criticism can be raised: one pragmatic, one principled based on output standards. The pragmatic criticism is to question whether parliamentary control is possible in practice. One thing is to pursue institutional parliamentary interests in the comitology area by a small number of dedicated

190 The EU Comitology System in Theory and Practice

parliamentarians, primarily from the constitutional or legal affairs committee; pursuing policy interests in relation to the individual measures adopted by the comitology procedure is quite another. Every year the Commission adopts 1500–2500 measures under the comitology procedures. The logistic challenge for the Parliament in handling this number and diversity of measures within the specified time limits is enormous. Although to different degrees, the task involves all the Parliament’s committees. Securing the right and timely flow of documents from the Parliament’s Receptions and Referrals Office to the individual committee secretariats and on to the individual members of Parliament is demanding. Although the Parliament has increased the number of staff dealing with comitology issues, handling the document flow is still a challenge. Not surprisingly, some committees function better than others. Furthermore, even if the documents reach the committee members in time, securing interest and competent reactions is also challenging. The comitology measures are technically complex, and often the parliamentarians need to rely on the expertise of the committees’ limited secretarial resources (Kaeding and Hardacre, 2010; Brandsma, 2010b). In short, given the information load, the Parliament’s ambition of exercising ‘police patrol control’ (McCubbins and Schwartz, 1984) in the comitology area is extremely demanding and unlikely to be systematically effective. This means that, to be effective, parliamentary control must probably rely on ‘fire alarms’. This would require MEPs to assume that a comitology measure is fine until the relevant committee is notified by a private or public interest group to the contrary (cf. also Hix, 2000). But in the absence of a formal ‘Notice and Comment’ system, for instance known from the US Administrative Procedure Act, where interested external parties are notified of proposed new comitology measures and given a formal opportunity to voice concern (cf. Bignami, 1999), interactions between members of the European Parliament and interest groups about comitology measures must be informal, and thus nontransparent. However, secretive parliamentary interactions with interest groups would seem to undermine the whole logic of involving the Parliament in comitology cases and strengthening input mechanisms in the first place. In sum, the European Parliament’s increased influence in the comitology area is questionable on purely practical grounds. Second, judged by output standards, it is questionable whether increased Parliament influence is desirable in a normative sense. It is here useful to return to Bellamy’s (2010) pragmatic middle ground between input and output arguments: the non-majoritarian aspects of the EU are legitimate, first, as long as the EU is primarily an arrangement

Conclusion

191

for interstate regulatory cooperation and, second, as long as the nonmajoritarian mechanisms are controlled by the majoritarian systems of the member states. If these conditions are fulfilled, the arguments of the advocates of output legitimacy carry considerable weight. In the regulatory arena, citizens look for Pareto-efficient improvements that correct for market failure. Such measures are technical, uncontentious and have low electoral salience. They require expertise, equity and efficiency – qualities best provided by experts representing national interests yet isolated from short-term electoral considerations or powerful pressure groups. Even Follesdal and Hix (2006, p. 542) accept that purely regulatory policies may legitimately be isolated from the ordinary majoritarian democratic process. But, as Follesdal and Hix correctly point out, the EU’s many activities are not limited to purely regulatory functions without redistributive effects or to apolitical technical issues. However, while this observation may be true for the EU in general, it is less true for the comitology area. The comitology system deals with regulation, not with initiating legislation or proposing and negotiating budgets. Making implementation measures is regulation par excellence and it is, not surprisingly, one of the Commission’s least politicized functions. If there is one area of the EU system where purely regulatory functions dominate, it is the comitology system. Hence, the arguments of the output-oriented scholars would seem to carry considerable weight in the comitology area. But Bellamy’s (2010) second condition for accepting output arguments – that the member states retain control – should be kept in mind. In this perspective, the comitology procedures have two desirable traits. First, they secure that the member states control the Commission’s implementation measures. The member states can even be held to account since formal votes are taken in comitology committees. Second, they add considerable technical information to the decision-making process since the member states are represented by experts from their national administrations. From the general perspective advocated in this book, adding technical insight to the process is a by-product of the comitology system. But from a democratic–normative perspective it is not an unimportant by-product. The conclusion from the above reasoning is that the rise of the European Parliament in the comitology area represents a loss of democratic legitimacy. Especially the Lisbon Treaty is a step in the wrong direction. The European Parliament is now more involved than ever in delegated decision-making. It controls delegated acts and has full influence over the rules on the comitology system, and it has used this

192 The EU Comitology System in Theory and Practice

influence to secure a right of scrutiny and access to information on implementation acts. At the same time the member states’ control over the Commission’s implementation measures is weakened. In the case of delegated acts, they must share control with the Parliament. In the case of implementation acts, more measures are to fall under the advisory procedure where the member states have no formal control. Accountability is consequently less clear in the new system. All this is likely to lead to more politicization of comitology cases, which again will lead to more redistributive rather than Pareto-efficient outcomes. As a result, legitimacy decreases.

Notes

1

Introduction

1. The expression ‘EU’/‘European Union’ is used throughout the book to streamline the presentation, although at times ‘European Communities’ (EC) or ‘European Economic Community’ (EEC) would strictly speaking be more appropriate. 2. The expression ‘the Commission’ is used throughout the book to streamline the presentation, although at times ‘the Commission of the European Economic Community’, ‘the Commission of the European Communities’ or ‘the European Commission’ would strictly speaking be more appropriate. 3. The story of the single sky committee and its meeting on 8 July 2010 is based on material from dossier CMTD(2010)0774 available at the Commission’s comitology register (http://ec.europa.eu/transparency/regcomitology/ index.cfm). The title of the Commission’s amended regulation is Commission regulation (EU) no. 1191/2010 amending regulation (EC) no. 1794/2006 laying down a common charging scheme for air navigation services. It is published in OJ L 333, 17 December 2010, pp. 6–20. 4. The expression ‘the Council’ is used throughout the book to streamline the presentation, although at times ‘the Council of Ministers’ or ‘the Council of the European Union’ would strictly speaking be more appropriate. 5. Simple search on all secondary acts in force divided according to author of act (http://eur-lex.europa.eu/en/index.htm). 6. The expression ‘the codecision procedure’ is used throughout the book to streamline the presentation, although at times ‘the ordinary legislative procedure’ would strictly speaking be more appropriate.

4 The Origins of Comitology 1. See Lindberg (1963, pp. 219–83), Groeben (1987, pp. 70–8, 101–8) and Moravcsik (1998, pp. 208–17). See also Commission President Walter Hallstein’s (1962, pp. 54–6) vivid account of the marathon negotiations in the Council of Ministers in December 1961 and January 1962. 2. The following sections draw on Blom-Hansen (2008). 3. Before the merger treaty of 1965 was implemented, the European Economic Community (EEC), the European Coal and Steel Community (ECSC) and the European Atomic Energy Community (Euratom) published their own separate Bulletins. In 1968, a common monthly Bulletin of the European Communities took over. The different bulletins are abbreviated in the text – the EEC-Bulletin and the EC-Bulletin. 193

194 Notes

5 What is at Stake? A Case Study of the 2006 Comitology Reform 1. The case study also provides a basis for understanding the post-Lisbon negotiations on the comitology system (see Blom-Hansen, 2011a). 2. See, for example, article 17(4) in the market abuse directive (OJ L96/16, 12 April 2003); article 27(4) in the transparency directive (OJ L390/38, 31 December 2004); article 24(4) in the prospectus directive (OJ L345/64, 31 December 2003); article 64(3) in the directive on markets in financial instruments (OJ L145/1, 30 April 2004); and article 150(4) in the directive on the business of credit institutions (OJ L177/1, 30 June 2006).

7

Comitology Across Policy Areas

1. Directive no. 30/2009 on the specification of petrol, diesel and gas-oil and the monitoring of greenhouse gas emissions, article 11. 2. Regulation no. 1083/2006 laying down the general provisions on the European Regional Development Fund, the European Social Fund and the Cohesion Fund, article 103. 3. Directive no. 37/2001 on the approximation of the laws, regulations and administrative provisions of the member states concerning the manufacture, presentation and sale of tobacco products, article 10. 4. Regulation no. 1798/2003 on administrative cooperation in the field of valued added tax, article 44. 5. Since the three items have unequal ranges, I use dichotomized versions to construct the index. For the length of the decision-making process ‘0’ equals less than 1 year (324 cases), while ‘1’ equals more than 1 year (345 cases). For ‘B-points’ on the Council agenda ‘0’ equals no B-points (468 cases), while ‘1’ equals 1 or more ‘B-points’ (218 cases). For changed Commission proposal ‘0’ equals no changed proposal (554 cases), while ‘1’ equals a changed proposal (115 cases). A reliability test confirms that the three items measure the same underlying variable (Chronbach’s alpha = 0.66), although this result should be interpreted with caution due to the dichotomous nature of the items. 6. Article 15 in Council regulation (EC) no. 384/96 of 22 December 1995 on protection against dumped imports from countries not members of the European Community; and article 25 in Council regulation (EC) no. 2026/97 of 6 October 1997 on protection against subsidized imports from countries not members of the European Community. 7. The number of cases varies across models in Table 7.3 because the dataset includes 13 member state initiatives for which it is not possible to compute the institutional conflict variable (since the Commission does not make the legislative proposal, two of the items in the conflict index cannot be computed). In addition, a few cases have missing values due to data missing in the PreLex or Eur-Lex databases. 8. The statistically significant sign of the legislative procedures that involve the European Parliament may appear puzzling. How can the codecision and consultation procedures, controlled for conflict, have a positive impact on the likelihood of inserting a comitology clause into a legislative act? It would

Notes

195

be misleading to interpret this as an indication that the European Parliament pressures for comitology committees since we know from the previous chapter that the Parliament’s comitology preferences are more permissive than the Council’s. The positive coefficient is more likely due to the fact that the legislative procedures to some extent also function as proxies for policy areas combined with the fact that comitology committees are more prevalent in areas with parliamentary involvement (cf. Table 7.1). 9. The non-binding guidelines for the choice of comitology procedure are stated in the 1999 comitology decision’s article 2: The choice of procedural methods for the adoption of implementing measures shall be guided by the following criteria: (a) management measures, such as those relating to the application of the common agricultural and common fisheries policies, or to the implementation of programmes with substantial budgetary implications, should be adopted by use of the management procedure; (b) measures of general scope designed to apply essential provisions of basic instruments, including measures concerning the protection of the health or safety of humans, animals or plants, should be adopted by use of the regulatory procedure; where a basic instrument stipulates that certain non-essential provisions of the instrument may be adapted or updated by way of implementing procedures, such measures should be adopted by use of the regulatory procedure; (c) without prejudice to points (a) and (b), the advisory procedure shall be used in any case in which it is considered to be the most appropriate. Coding whether acts fall under these guidelines has been done conservatively. Only areas that are directly mentioned by the guidelines are coded. Consequently, areas covered by the guideline for the management procedure are the common agricultural policies and common fisheries policies, which are both directly mentioned in point (a). In addition, regional policy and ‘science, information, education and culture’ are coded as falling under this guideline because they arguably have ‘substantial budgetary implications’. This amounts to 102 out of 686 cases. Areas covered by the guideline for the regulatory procedure are ‘environment, health and consumer protection’ and ‘agricultural health measures’ which are both directly mentioned in point (b). This amounts to 54 out of 686 cases. The guideline for the advisory procedure – point (c) – is so loose that a coding is not possible. Consequently all remaining cases are coded as not being subject to any guideline. This amounts to 530 out of 686 cases. All codings of acts into policy areas are based on the act’s first mentioned directory code in the EU’s Eur-Lex database. 10. As an alternative to the fixed effect approach of controlling for comitology guidelines in models 4 and 5 in Table 7.3, I have run a test of these models without the fixed effects, but with robust standard errors clustered at the level of policy areas (defined as in Table 7.1). I reach substantially

196 Notes similar results. The conflict variable and the annex variable are still statistically significant, but the preparatory document variable just fails to reach statistical significance in this analysis. Results are available upon request. Since the fixed effect approach represents a more direct test of the comitology guidelines, I report this analysis in Table 7.3. 11. In the test in Table 7.5 the preparatory committee variable is not accompanied by a control for cases imposing countervailing or anti-dumping duties because these cases do not install any comitology committee and, hence, are not included in the dataset analysed in this table. 12. The correlation between the measure of institutional conflict and the two complexity measures, preparatory documents and preparatory committees, is lower in the subset of data analysed in Table 7.4 (gamma = 0.35 and 0.43, respectively). 13. Again, as an alternative to the fixed effect approach of controlling for comitology guidelines in model 4, in Table 7.5, I have run a test of this model without the fixed effects, but with robust standard errors clustered at the level of policy areas (defined as in Table 7.1). Again, I reach substantially similar results. The conflict variable, the preparatory committee variable and the annex variable are still statistically significant in this analysis. Results are available upon request. Since the fixed effect approach represents a more direct test of the comitology guidelines, I report this analysis in Table 7.5.

8

The Comitology System in Daily Practice

1. Now such a list is available at the Commission’s online comitology register at http://ec.europa.eu/transparency/regcomitology/index.cfm. 2. Sannerstedt’s data-collection process is described in Elgström et al. (2001, pp. 112–13). 3. Descriptive statistics on all these parameters can be found in Table 8.2. 4. As an alternative to the fixed effects approach, I have run the regressions in Table 8.9 without the fixed effects, but with robust standard errors clustered at the level of DGs. The analysis of the Commission as a mediator yields substantially identical results. The deliberative interaction style is still strongly related to the mediating role. However, the analysis of the Commission as a policy advocate reaches different results. The bargaining interaction style no longer obtains statistical significance.

Literature

Agence Europe. Bulletin quotidien Europe (various years and days). Alfé, M., G. J. Brandsma and T. Christiansen (2009) ‘The Functioning of Comitology Committtees in Practice’, in T. Christiansen, J. M. Oettel and B. Vaccari (ed.), 21st Century Comitology. Implementing Committees in the Enlarged European Union (Maastricht: European Institute of Public Administration). Andenas, M. and A. Türk (ed.) (2000) Delegated Legislation and the Role of Committees in the EC (The Hague: Kluwer Law International). Avgerinos, Y. V. (2002) ‘Essential and Non-essential Measures: Delegation of Powers in EU Securities Regulation’, European Law Journal, 8, 269–89. Bache, I. (1999) ‘The Extended Gatekeeper: Central Government and the Implementation of EU Regional Policy in the UK’, Journal of European Public Policy, 6, 28–45. Bache, I. (2004) ‘Themes and Issues in Multi-level Governance’, in I. Bache and M. Flinders (ed.), Multi-Level Governance (Oxford: Oxford University Press). Ballman, A., D. Epstein and S. O’Halloran (2002) ‘Delegation, Comitology, and the Separation of Powers in the European Union’, International Organization, 56, 551–74. Bates, R. H., A. Greif, M. Levi, J.-L. Rosenthal and B. Weingast (1998) ‘Introduction’, in R. H. Bates, A. Greif, M. Levi, J.-L. Rosenthal and B. Weingast (ed.), Analytic Narratives (Princeton, NJ: Princeton University Press). Bawn, K. (1995) ‘Political Control versus Expertise: Congressional Choices about Administrative Procedures’, American Political Science Review, 89, 62–73. Bellamy, R. (2010) ‘Democracy without Democracy? Can the EU’s Democratic “Outputs” be separated from the Democratic “Inputs” Provided by Competitive Parties and Majority Rule?’, Journal of European Public Policy, 17, 2–19. Bendor, J., A. Glazer and T. Hammond (2001) ‘Theories of Delegation’, Annual Review of Political Science, 4, 235–69. Bergman, T., W. C. Müller, K. Strøm and M. Blomgren (2003) ‘Democratic Delegation and Accountability: Cross-national Patterns’, in K. Strøm, W. C. Müller and T. Bergman (ed.), Delegation and Accountability in Parliamentary Democracies (Oxford: Oxford University Press). Bergström, C. F. (2005) Comitology – Delegation of Powers in the European Union and the Committee System (Oxford: Oxford University Press). Bergström, C. F., H. Farrell and A. Heritier (2007) ‘Legislate or Delegate? Bargaining over Implementation and Legislative Authority in the EU’, West European Politics, 30, 338–66. Bertram, C. (1967–68) ‘Decision-Making in the E.E.C: The Management Committee Procedure’, Common Market Law Review, 5, 246–64. Beyers, J. (2005) ‘Multiple Embeddedness and Socialization in Europe: The Case of Council Officials’, International Organization, 59, 899–936. 197

198 Literature Bignami, F. E. (1999) ‘The Democratic Deficit in European Community Rulemaking: A Call for Notice and Comment in Comitology’, Harvard International Law Journal, 40, 451–515. Blom-Hansen, J. (2005) ‘Principals, Agents, and the Implementation of EU Cohesion Policy’, Journal of European Public Policy, 12, 624–49. Blom-Hansen, J. (2007) ‘The Danish Comitology Survey. Technical Report’, Unpublished report. Department of Political Science, Aarhus University. Downloadable from the author’s home page. Blom-Hansen, J. (2008) ‘The Origins of the EU Comitology System: A Case of Informal Agenda-Setting by the Commission’, Journal of European Public Policy, 15, 208–26. Blom-Hansen, J. (2010) ‘Dataset on Comitology Provisions in EU Regulations and Directives – Technical Description, Variables, And Frequency Distributions’, Unpublished report. Department of Political Science, Aarhus University. Downloadable from the author’s home page. Blom-Hansen, J. (2011a) ‘Interests, Instruments, and Institutional Preferences in the EU Comitology System: The 2006 Comitology Reform’, European Law Journal, 17, 343–364. Blom-Hansen, J. (2011b) ‘Technical Note: Non-Identical Comitology Preferences of EU Institutions in Daily Legislation, 1999–2006’, Unpublished note. Department of Political Science, Aarhus University. Downloadable from the author’s home page. Blom-Hansen, J. and G. J. Brandsma (2009) ‘The EU Comitology System: Intergovernmental Bargaining and Deliberative Supranationalism?’, Journal of Common Market Studies, 47, 719–40. Boisson De Chazournes, L. and M. M. Mbengue (2004) ‘GMOs and Trade: Issues at Stake in the EC Biotech Dispute’, Review of European Community & International Environmental Law, 13, 289–305. Bradley, K. St. C. (1992) ‘Comitology and the Law: Through a Glass, Darkly’, Common Market Law Review, 29, 693–721. Bradley, K. St. C. (1997) ‘The European Parliament and Comitology: On the Road to Nowhere?’, European Law Journal, 3, 230–54. Bradley, K. St. C. (1998) ‘Alien Corn, or the Transgenic Procedural Maze’, in M. P. C. M. van Schendelen (ed.), EU Committees as Influential Policymakers (Aldershot: Ashgate). Bradley, K. St. C. (2008) ‘Halfway House: The 2006 Comitology Reforms and the European Parliament’, West European Politics, 31, 837–54. Brandsma, G. J. (2010a) Backstage Europe. Comitology, Accountability and Democracy in the European Union (Utrecht: University of Utrecht). Brandsma, G. J. (2010b) ‘The Effect of Information on Legislative Oversight: The European Parliament’s Response to Increasing Transparency of the Executive’, Paper prepared for the USG Anniversary Conference, 19–20 November 2010. Brandsma, G. J., and J. Blom-Hansen (2010) ‘The EU Comitology System: What Role for the Commission?’, Public Administration, 88, 496–513. Brandsma, G. J., and J. Blom-Hansen (2011). ‘The post-Lisbon Battle over Comitology: Another Round of the Politics of Structural Choice?’, Working paper, Utrecht School of Governance, University of Utrecht & Department of Political Science, Aarhus University (available from the authors upon request).

Literature

199

Brandsma, G. J., D. Curtin and A. Meijer (2008) ‘How Transparent are EU “Comitology” Committees in Practice’, European Law Journal, 14, 819–38. Broscheid, A. and D. Coen (2007) ‘Lobbying Activity and Fora Creation in the EU: Empirically Exploring the Nature of the Policy Good’, Journal of European Public Policy, 14, 346–65. Buonanno, L. (2006) ‘The Creation of the European Food Safety Authority’, in C. Ansell and D. Vogel (eds.), What’s the Beef? The Contested Governance of European Food Safety (Cambridge, MA: The MIT Press). Chambers, S. (2003) ‘Deliberative Democratic Theory’, Annual Review of Political Science, 6, 307–26. Christiansen, T. and E. Kirchner (ed.) (2000) Europe in Change. Committee Governance in the European Union (Manchester: Manchester University Press). Christiansen, T. and B. Vaccari (2006) ‘The 2006 Reform of Comitology: Problem Solved or Dispute Postponed?’, Eipascope, Bulletin 2006/3, 9–17. Christiansen, T. and T. Larsson (ed.) (2007) The Role of Committees in the Policy-Process of the European Union. Legislation, Implementation and Deliberation (Cheltenham: Edward Elgar). Cini, M. (1997) ‘Administrative Culture in the European Commission: The Cases of Competition and Environment’, in N. Nugent (ed.), At the Heart of the Union (London: MacMillan). Commission (1968) ‘Commission Proposal on Valuation for Customs Purposes’, EC-Bulletin, 1/1968, 27–28. Commission (1999a) Implementing the Framework for Financial Markets: Action Plan, COM(1999) 232 Final. Commission (1999b) Commission Statement (ad article 5). Published in ‘Declarations on Council Decision 1999/468/EC of 28 June 1999 laying down the procedures for the exercise of implementing powers conferred on the Commission’, OJ C 203, 17 July 1999, 1. Commission (2001a) European Governance – A White Paper, COM(2001) 428 Final. Commission (2001b) ‘Commissioner Bolkestein Welcomes Final Report from Wise Men on Securites Markets Regulation’, Press Release, 15 February 2001 (available at: http://ec.europa.eu/internal_market/securities/lamfalussy/index_ en.htm, date accessed 30 May 2011). Commission (2001c) ‘Standard Rules of Procedure – Council Decision 1999/468/EC’, OJ C 38, 6 February 2001, 3–5. Commission (2002a) Implementation of Financial Services Legislation in the Context of the Lamfalussy Report, Intervention by President Romano Prodi to the European Parliament’s plenary session Strasbourg, 5 February 2002 (available at: http://ec.europa.eu/internal_market/securities/lamfalussy/index_ en.htm, date accessed 30 May 2011). Commission (2002b) Proposal for a Council Decision Amending Decision 1999/468/EC laying down the procedures for the exercise of implementing powers conferred on the Commission, COM(2002) 719 Final. Commission (2004) Amended proposal for a Council Decision amending Decision 1999/468/EC laying down the procedures for the exercise of implementing powers conferred on the Commission, COM(2004) 324 Final. Commission (2006a) ‘Statements by the Commission to be entered in the Council minutes of 17 July 2006’, OJ C 171, 22 July 2006, 21.

200 Literature Commission (2006b) Commission Staff Working Document. Annex to the report from the Commission on the working of committees during 2005, SEC(2006) 1065. Brussels. Commission (2009a) Report from the Commission on the Working of Committees during 2008, COM(2009) 335 Final. Commission (2009b) Communication from the Commission to the European Parliament and the Council. Implementation of Article 290 of the Treaty on the Functioning of the European Union, COM(2009) 673 Final. Commission (2010a) Report from the Commission on the Working of Committees during 2009, COM(2010) 354 Final. Commission (2010b) Proposal for a Regulation of the European Parliament and of the Council laying down the rules and general principles concerning mechanisms for control by member states of the Commission’s exercise of implementing powers, COM(2010) 83 Final. Committee of Wise Men (2001) Final report of the Committee of Wise Men on the Regulation of European Securities Markets (available at: http://ec.europa.eu/ internal_market/securities/docs/lamfalussy/wisemen/final-report-wise-men_ en.pdf, date accessed 30 May 2011). Coombes, D. (1970) Politics and Bureaucracy in the European Community (London: George Allen and Unwin). Corbett, R. (2003) Statement by rapporteur Corbett in the debates of the European Parliament on 2 September 2003 on the exercise of Commission implementing powers (available at: http://www.europarl.europa.eu/, date accessed 28 July 2008). Council of Ministers (1964) ‘Accord interne relatif au financement et á la gestion des aides de la Communauté’, OJ, 11 June 1964, 1493–7. Council of Ministers (1968a) ‘Reglement (CEE) no. 802/1968 du Conseil du 27 juin 1968 relatif a la definition commune de la notion d’origine des merchandises’, OJ L 148, 20 June 1968, 1–5 Council of Ministers (1968b) ‘Reglement (CEE) no. 803/1968 du Conseil du 27 juin 1968 relatif a la valeur en douane des merchandises’, OJ L 148, 20 June 1968, 6–12. Council of Ministers (1987) ‘Council Decision of 13 July 1987 laying down the procedures for the exercise of implementing powers conferred on the Commission’, OJ L 197, 18 July 1987, 33–35. Council of Ministers (1999a) ‘Council decision of 28 June 1999 laying down the procedures for the exercise of implementing powers conferred on the Commission’, OJ L 184, 17 July 1999, 23–26. Council of Ministers (1999b) 2192nd Council meeting – General Affairs – Luxembourg, 21–22 June 1999. PRES99/198 (available at: http://europa.eu/rapid/ searchAction.do, date accessed 30 May 2011). Council of Ministers (2006a) ‘Council decision of 28 June 1999 laying down the procedures for the exercise of implementing powers conferred on the Commission. Decision as amended by Decision 2006/512/EC, OJ L 200, 22 July 2006. Consolidated version’, OJ C 255, 21 October 2006, 4–8. Council of Ministers (2006b) ‘Council decision of 17 June 2006 amending Decision 1999/468/EC laying down the procedures for the exercise of implementing powers conferred on the Commission’, OJ L 200, 22 July 2006, 11–13.

Literature

201

Craig, P. (2008) ‘The Role of the European Parliament under the Lisbon Treaty’, in S. Giller and J. Ziller (ed.), The Lisbon Treaty. EU Constitutionalism without a Constitutional Treaty? (Wien: Springer Verlag). Crombez, C. (1996) ‘Legislative Procedures in the European Community’, British Journal of Political Science, 26, 199–228. Daemen, H. H. F. M. and M. P. C. M. van Schendelen (1998) ‘The Advisory Committee on Safety, Hygiene and Health Protection at Work’, in M. P. C. M. van Schendelen (ed.), EU Committees as Influential Policymakers (Aldershot: Ashgate). Danish Foreign Ministry (2005) Grundnotat af 4. november 2005 om Kommissionens forslag til Rådets afgørelse om ændring af afgørelse 1999/468/ER om fastsættelse af de nærmere vilkår for udøvelsen af de gennemførelsesbeføjelser, der tillægges Kommissionen. Submitted to the Danish Parliament’s EU-committe on 10 November 2005 (available at: http://www.euo.dk/upload/application/pdf/ 4b8b29b6/20040324_01.pdf, date accessed 30 May 2011). Danish Foreign Ministry (2006) Samlenotat af 31. marts 2006 vedr. Rådsmøde (almindelige anliggender og eksterne forbindelser) den 10–11. april 2006. Submitted to the Danish Parliament’s EU-committe on 31 March 2006 (available at: www.ft.dk, date accessed 30 May 2011). de Visscher, C., O. Maiscocq and F. Varone (2008) ‘The Lamfalussy Reform in the EU Securities Markets: Fiduciary Relationships, Policy Effectiveness and Balance of Power’, Journal of Public Policy, 28, 19–47. Dehousse, R. (2003) ‘Comitology: Who Watches the Watchmen?’, Journal of European Public Policy, 10, 798–813. Demmke, C. (2000) ‘Comitology in the Environmental Sector’, in M. Andenas and A. Türk (eds.), Delegated Legislation and the Role of Committees in the EC (The Hague: Kluwer Law International). Demmke, C., E. Eberharter, G. F. Schaefer and A. Türk (1996) ‘The History of Comitology’, in R. H. Pedler and G. F. Schaefer (eds.), Shaping European Law and Policy. The Role of Committees and Comitology in the Political Process (Maastricht: European Centre for Public Affairs). Dogan, R. (1997) ‘Comitology: Little Procedures with Big Implications’, West European Politics, 20, 31–60. Dogan, R. (2000) ‘A Cross-Sectoral View of Comitology: Incidence, Issues and Implications’, in T. Christiansen and E. Kirchner (eds.), Committee Governance in the European Union (Manchester: Manchester University Press). Döring, H. (2007) ‘The Composition of the College of Commissioners. Patterns of Delegation’, European Union Politics, 8, 207–28. ECJ Case C-295/06. Kingdom of Denmark v. Commission of the European Communities – Action for annulment under Article 230 EC, brought on 9 January 2006 (available at: http://curia.europa.eu/, date accessed 30 May 2011). EC-Bulletin (various years and numbers) Bulletin of the European Communities (Brussels: The Commission of the European Communities). EEC-Bulletin (various years and numbers) Bulletin of the European Economic Community (Brussels: The Commission of the European Economic Community). EEC-Commission (1959) ‘La politique agricole commune’, EEC-Bulletin, 5/1959, 5–29. EEC-Commission (1960) Third General Report on the Activities of the Community (21 March 1959 – 15 May 1960) (Bruxelles: European Economic Community).

202 Literature EEC-Commission (1961a) Fourth General Report on the Activities of the Community (16 May 1960–30 April 1961) (Bruxelles: European Economic Community). EEC-Commission (1961b) ‘Draft regulation providing for the institution of a system of levies and for the gradual establishment of a common market organization in the grain sector’, EEC-Bulletin, 7-8/1961, Annex. EEC-Commission (1961c) ‘Draft proposal on a common policy for rice’, EECBulletin, 11/1961, Annex, 5–27. EEC-Commission (1962) Fifth General Report on the Activities of the Community (1 May 1961–30 April 1962) (Bruxelles: European Economic Community). EEC-Commission (1965a) ‘Common definition of the origin of goods’, EECBulletin, 3/1964, 18–21. EEC-Commission (1965b) ‘Proposal for a Council regulation relating to a common definition of the origin of goods’, EEC-Bulletin, 3/1964, Supplement, 2–11. Egeberg, M. (1999) ‘Transcending Intergovernmentalism? Identity and Role Perceptions of National Officials in EU Decision-Making’, Journal of European Public Policy, 6, 456–74. Egeberg, M. (2006) ‘Executive Politics as Usual: Role Behaviour and Conflict Dimensions in the College of European Commissioners’, Journal of European Public Policy, 13, 1–15. Egeberg, M., G. F. Schaefer and J. Trondal (2003) ‘The Many Faces of EU Committee Governance’, West European Politics, 26, 19–40. Elgström, O., B. Bjurulf, J. Johansson and A. Sannerstedt (2001) ‘Coalitions in European Union Negotiations’, Scandinavian Political Studies, 24, 111–28. Epstein, D. and S. O’Halloran (1994) ‘Administrative Procedures, Information, and Agency Discretion’, American Journal of Political Science, 38, 697–722. Epstein, D. and S. O’Halloran (1996) ‘Divided Government and the Design of Administrative Procedures: A Formal Model and Empirical Test’, The Journal of Politics, 58, 373–97. Epstein, D. and S. O’Halloran (1999) Delegating Powers. A Transaction Cost Politics Approach to Policy Making under Separate Powers (Cambridge: Cambridge University Press). European Convention (2002). Final report of Working Group IX on Simplification. CONV 424/02 (available at: http://european-convention.eu.int/, date accessed 30 May 2011). European Parliament (1962) ‘Resolution concerning the powers of the European Commission in the implementation of the common agricultural policy’, EECBulletin 2/1962: Annex 1. European Parliament (2001) ‘European Parliament resolution of 15 March 2001 on the final report of the Committee of Wise Men on the regulation of European securities markets’, OJ C 343, 5 December 2001, 265–7. European Parliament (2002) ‘European Parliament resolution on the implementation of financial services legislation (2001/2247(INI))’, OJ C 284E, 21 November 2002, 115–17. European Parliament (2003a) Report on the proposal for a Council decision on amending Decision 1999/468/EC laying down the procedures for the exercise of implementing powers conferred on the Commission. Committee on Constitutional Affairs (rapporteur: Richard Corbett). EP doc A5-0128/2003. European Parliament (2003b) Second Report on the proposal for a Council decision on amending Decision 1999/468/EC laying down the procedures for the exercise of

Literature

203

implementing powers conferred on the Commission. Committee on Constitutional Affairs (rapporteur: Richard Corbett). EP doc A5-0266/2003. European Parliament (2006a) Report on the draft Council Decision amending Decision 1999/468/EC laying down the procedures for the exercise of implementing powers conferred on the Commission. Committee on Constitutional Affairs (rapporteur: Richard Corbett). EP doc A6-0236/2006. European Parliament (2006b) Report on the conclusion of an interinstitutional agreement taking the form of a joint statement concerning the draft for a Council Decision amending Decision 1999/468/EC laying down the procedures for the exercise of implementing powers conferred on the Commission (new regulatory procedure with scrutiny). Committee on Constitutional Affairs (rapporteur: Richard Corbett). EP doc A6-0237/2006. European Parliament and Council (2011). ‘Regulation (EU) No 182/2011 of the European Parliament and of the Council of 16 February 2011 laying down the rules and general principles concerning mechanisms for control by Member States of the Commission’s exercise of implementing powers’, OJ L 55, 28 February 2011, 13–19. European Parliament, Council and Commission (2006) ‘Statement by the European Parliament, the Council and the Commission concerning the Council Decision of 17 July 2006 amending Decision 1999/468/EC laying down the procedures for the exercise of implementing powers conferred on the Commission’, OJ C 255, 21 October 2006, 1–3. European Parliament, Council and Commission (2011). Common Understanding on practical arrangements for the use of delegated acts. Draft sent by Jerzy Buzek, President of the European Parliament, to Janos Martonyi, President of the Council of the European Union, on 30 March 2011. Available as Council document 8640/11. Falke, J. (1996) ‘Comitology and other Committees: A Preliminary Empirical Assessment’, in R. H. Pedler and G. F. Schaefer (ed.), Shaping European Law and Policy. The Role of Committees and Comitology in the Political Process (Maastricht: European Centre for Public Affairs). Falke, J. (2000) ‘Comitology: From Small Councils to Complex Networks’, in M. Andenas and A. Türk (ed.), Delegated Legislation and the Role of Committees in the EC (The Hague: Kluwer Law International). Flynn, B. (2000) ‘Postcards from the Edge of Integration? The Role of Committees in EU Environment Policy-making’, in T. Christiansen and E. Kirchner (eds.), Committee Governance in the European Union (Manchester: Manchester University Press). Follesdal, A. and S. Hix (2006) ‘Why There is a Democratic Deficit in the EU: A Response to Majone and Moravcsik’, Journal of Common Market Studies, 44, 533–62. Fouilleux, E., J. De Maillard and A. Smith (2005) ‘Technical or Political? The Working Groups of the EU Council of Ministers’, Journal of European Public Policy, 12, 609–23. Franchino, F. (2000a) ‘Control of the Commission’s Executive Functions’, European Union Politics, 1, 63–92. Franchino, F. (2000b) ‘The Commission’s Executive Discretion, Information and Comitology’, Journal of Theoretical Politics, 12, 155–81. Franchino. F. (2004) ‘Delegating Powers in the European Community’, British Journal of Political Science, 34, 269–93.

204 Literature Franchino, F. (2007) The Powers of the Union. Delegation in the EU (Cambridge: Cambridge University Press). French Foreign Minister Couve de Murville (1966) Aide-memoire presented at the extraordinary meeting of the Council of Ministers in Luxembourg 17–18 January 1966, EEC-Bulletin 3/1966, 6–7. Garrett, G. and B. R. Weingast (1993) ‘Ideas, Interests, and Institutions: Constructing the European Community’s Internal Market’, in J. Goldstein and R. O. Keohane (ed.), Ideas and Foreign Policy (Ithaca: Cornell University Press). Garrett, G. and G. Tsebelis (1996) ‘An Institutional Critique of Intergovernmentalism’, International Organization, 50, 269–99. Gatt, A. (2009) ‘Comitology in the Field of GMO Authorisation: Implications for Legitimacy’, in T. Christiansen, J. M. Oettel and B. Vaccari (ed.), 21st Century Comitology. Implementing Committees in the Enlarged European Union (Maastricht: European Institute of Public Administration). Gehring, T. (1999) ‘Bargaining, Arguing and Functional Differentiation of Decision-making: The Role of Committees in European Environmental Process Regulation’, in C. Joerges and E. Vos (ed.), EU Committees: Social Regulation, Law and Politics (Oxford: Hart Publishing). George, A. L. and A. Bennett (2005) Case Studies and Theory Development in the Social Sciences (Cambridge, MA: MIT Press). Gornitzka, Å. and U. Sverdrup (2008) ‘Who Consults? The Configuration of Expert Groups in the European Union’, West European Politics, 31, 725–50. Groeben, H. v. d. (1987) The European Community: The Formative Years (Brussels: Commission of the European Communities). Guiraudon, V. (2003) ‘The Constitution of a European Immigration Policy Domain: A Political Sociology Approach’, Journal of European Public Policy, 10, 263–82. Hagemann, S. (2007) ‘Applying Ideal Point Estimation Methods to the Council of Ministers’, European Union Politics, 8, 279–96. Haibach, G. (1999) ‘Council Decision 1999/468 – A New Comitology Decision for the 21st Century!?’, Eipascope, 99/3, 10–17. Haibach, G. (2000) ‘The History of Comitology’, in M. Andenas and A. Türk (ed.), Delegated Legislation and the Role of Committees in the EC (The Hague: Kluwer Law International). Hallstein, W. (1962) United Europe. Challenge and Opportunity (Cambridge, MA: Harvard University Press). Hayes-Renshaw, F., W. Van Aken and H. Wallace (2006) ‘When and Why the EU Council of Ministers Votes Explicitly’, Journal of Common Market Studies, 44, 161–94. Heritier, A. and C. Moury (2009) ‘Contested Delegation: The Impact of Codecision on Comitology’, EUI Working Paper RSCAS 2009/64. European University Institute, Florence. Hix, S. (2000) ‘Parliamentary Oversight of Executive Power: What Role for the European Parliament in Comitology?’, in T. Christiansen and E. Kirchner (ed.), Committee Governance in the European Union (Manchester: Manchester University Press). Hix, S. (2002) ‘Parliamentary Behavior with Two Principals: Preferences, Parties, and Voting in the European Parliament’, American Journal of Political Science, 46, 688–98.

Literature

205

Hix, S. (2005) The Political System of the European Union. Houndmills: Palgrave. Hix, S., T. Raunio and R. Scully (2003) ‘Fifty Years On: Research on the European Parliament’, Journal of Common Market Studies, 41, 191–202. Hix, S., A. Noury and G. Roland (2006) ‘Dimensions of Politics in the European Parliament’, American Journal of Political Science, 50, 494–511. Hoffmann, H. (2009) ‘Legislation, Delegation and Implementation under the Treaty of Lisbon: Typology Meets Reality’, European Law Journal, 15, 482–505. Holmes, P. and J. Kempton (1996) ‘EU Anti-Dumping Policy: A Regulatory Perspective’, Journal of European Public Policy, 3, 647–64. Hooghe, L. (ed.) (1996) Cohesion Policy and European Integration. Building Multilevel Governance (Oxford: Oxford University Press). Hooghe, L. (2001) The European Commission and the Integration of Europe. Images of Governance (Cambridge: Cambridge University Press). Hooghe, L. and G. Marks (2003) ‘Unraveling the Central State, but How? Types of Multi-level Governance’, American Political Science Review, 97, 233–43. Huber, J. D. and C. R. Shipan (2000). ‘The Costs of Control: Legislators, Agencies, and Transaction Costs’, Legislative Studies Quarterly, 25, 25–52. Huber, J. D. and C. R. Shipan (2002). Deliberate Discretion? The Institutional Foundations of Bureaucratic Autonomy (Cambridge: Cambridge University Press). Huber, J. D. and C. R. Shipan (2006) ‘Politics, Delegation, and Bureaucracy’, in B. R. Weingast and D. A. Wittman (ed.), The Oxford Handbook of Political Economy (Oxford: Oxford University Press). Hug, S. (2003) ‘Endogeneous Preferences and Delegation in the European Union’, Comparative Political Studies, 35, 41–74. Häge, F. M. (2007) ‘Committee Decision-Making in the Council of the European Union’, European Union Politics, 8, 299–328. IGC (2004) Declaration on Article I-36 to be annexed to the Final Act of the Intergovernmental Conference and the Final Act of 25 October 2004. CIG 87/04 ADD 2 REV 2 (available at: http://www.consilium.europa.eu/igcpdf/en/04/ cg00/cg00087-ad02re02.en04.pdf, date accessed 30 May 2011). Joerges, C. (2002) ‘ “Deliberative Supranationalism” – Two Defences’, European Law Review, 8, 133–51. Joerges, C. (2006) ‘ “Deliberative Political Processes” Revisited: What Have we Learnt about the Legitimacy of Supranational Decision-Making?’, Journal of Common Market Studies, 44, 779–802. Joerges, C. and J. Neyer (1997a) ‘From Intergovernmental Bargaining to Deliberative Political Processes: The Constitutionalisation of Comitology’, European Law Journal, 3, 273–99. Joerges, C. and J. Neyer (1997b) ‘Transforming Strategic Interaction into Deliberative Problem-Solving: European Comitology in the Foodstuffs Sector’, Journal of European Public Policy, 4, 609–25. Joerges, C. and E. Vos (ed.) (1999) EU Committees: Social Regulation, Law and Politics (Oxford: Hart Publishing). Kaeding, M. and A. Hardacre (2010) ‘The Execution of Delegated Powers after Lisbon. A Timely Analysis of the Regulatory Procedure with Scrutiny and its Lessons for Delegated Acts’, EUI Working Papers RSCAS 2010/85, European University Institute, Florence. Kassim, H. (2000) ‘Conclusion: The National Coordination of EU Policy: Confronting the Challenge’, in H. Kassim, B. G. Peters and V. Wright (ed.),

206 Literature The National Coordination of EU Policy. The Domestic Level (Oxford: Oxford University Press). Kelemen, D. (2002) ‘The Politics of “Eurocratic” Structure and the New European Agencies’, West European Politics, 25, 93–118. Keohane, R. O., S. Macedo and A. Moravcsik (2009) ‘Democracy-Enhancing Multilateralism’, International Organization, 63, 1–31. Kiewiet, D. R. and M. D. McCubbins (1991) The Logic of Delegation (Chicago: The University of Chicago Press). Lascher, E. L. (1996) ‘Assessing Legislative Deliberation: A Preface to Empirical Analysis’, Legislative Studies Quarterly, 21, 501–19. Lewis, D. (2008) The Politics of Presidential Appointments. Political Control and Bureaucratic Performance (Princeton, NJ: Princeton University Press). Lewis, J. (2002) ‘National Interests: Coreper’, in J. Peterson and M. Shackleton (ed.), The Institutions of the European Union (Oxford: Oxford University Press). Lieberman, E. S. (2005) ‘Nested Analysis as a Mixed-Method Strategy for Comparative Research’, American Political Science Review, 99, 435–52. Lindberg, L. (1963) The Political Dynamics of European Economic Integration (Stanford: Stanford University Press). Lintner, P. and B. Vaccari (2007) ‘Comitology and the EP’s Scrutiny of Commission Implementing Acts: Real Parliamentary Control?’, in T. Christiansen and T. Larsson (ed.), The Role of Committees in the Policy-Process of the European Union. Legislation, Implementation and Deliberation (Cheltenham: Edward Elgar). Lupia, A. (2003). ‘Delegation and its Perils’, in K. Strøm, W. C. Müller and T. Bergman (ed.), Delegation and Accountability in Parliamentary Democracies (Oxford: Oxford University Press). Mahoney, C. (2004) ‘The Power of Institutions. State and Interest Group Activity in the European Union’, European Union Politics, 5, 441–66. Majone, G. (1998) ‘Europe’s “Democratic Deficit”: The Question of Standards’, European Law Journal, 4, 5–28. Majone, G. (2009) Europe as the Would-be World Power. The EU at Fifty (Cambridge: Cambridge University Press). Marks, G. (1992) ‘Structural Policy in the European Community’, in A. M. Sbragia (ed.), Europolitics (Washington, D.C.: The Brookings Institution). Marks, G. (1996) ‘Exploring and Explaining Variation in EU Cohesion Policy’, in L. Hooghe (ed.), Cohesion Policy and European Integration. Building Multilevel Governance (Oxford: Oxford University Press). Marks, G., L. Hooghe and K. Blank (1996) ‘European Integration from the 1980s: State-Centric vs. Multi-level Governance’, Journal of Common Market Studies, 34, 341–79. Mattila, M. (2008) ‘Voting and Coalition in the Council after the Enlargement’, in D. Naurin and H. Wallace (ed.), Unveiling the Council of the European Union (Houndsmills: Palgrave). Maxianova, K. and T. M. Rusche (2006) ‘Restriction of Hazardous Substances: On the Need for and the Limits of Comitology’, Review of European Community & International Environmental Law, 15, 202–10. McCubbins, M. D. (1985) ‘The Legislative Design of Regulatory Structure’, American Journal of Political Science, 29, 721–48.

Literature

207

McCubbins, M. D. and T. Schwartz (1984) ‘Congressional Oversight Overlooked: Police Patrols Versus Fire Alarms’, American Journal of Political Science, 28, 165–79. McCubbins, M. D. and E. Page (1987) ‘A Theory of Congressional Delegation’, in M. D. McCubbins and T. Sullivan (ed.), Congress: Structure and Policy (Cambridge: Cambridge University Press). McCubbins, M. D., R. G. Noll and B. R. Weingast (1987) ‘Administrative Procedures as Instruments of Political Control’, Journal of Law, Economics, and Organization, 3, 243–77. McCubbins, M. D., R. G. Noll and B. R. Weingast (1989) ‘Structure and Process, Politics and Policy: Administrative Arrangements and the Political Control of Agencies’, Virginia Law Review, 75, 431–82. McKelvey, R. D. (1976) ‘Intransivities in Multidimensional Voting Models and Some Implications for Agenda Control’, Journal of Economic Theory, 12, 472–82. Ministers of five EU countries and Norway (2005) Letter of 16 June 2005 from the Environmental Ministers of Belgium, Denmark, Finland, Norway, Portugal, and Sweden to Commissioner for the Environment Mr. Stavros Dimas on the Exemption of the Flame retardant decaBDE in the RoHS directive. 10192/05 (available at: http://register.consilium.europa.eu/pdf/en/05/st10/st10192.en05.pdf, date accessed 30 May 2011). Moe, T. M. (1990a) ‘The Politics of Structural Choice: Towards a Theory of Public Bureaucracy’, in O. E. Williamson (ed.), Organization Theory: From Chester Barnard to the Present and Beyond (New York: Oxford University Press). Moe, T. M. (1990b) ‘Political Institutions: The Neglected Side of the Story’, Journal of Law, Economics, and Organization, 6, 213–53. Moravcsik, A. (1998) The Choice for Europe (London: Routledge). Moravcsik, A. (2002) ‘In Defence of the “Democratic Deficit”: Reassessing Legitimacy in the European Union’, Journal of Common Market Studies, 40, 603–24. Morgan, D. and G. Goh (2004) ‘Genetically Modified Food Labelling and the WTO Agreements’, Review of European Community & International Environmental Law, 13, 306–19. Moser, P. (1997) ‘A Theory of the Conditional Influence of the European Parliament in the Cooperation Procedure’, Public Choice, 91, 333–50. Naurin, D. (2009) ‘Most Common When Least Important: Deliberation in the European Union Council of Ministers’, British Journal of Political Science, 40, 31–50. Neyer, J. (1998) ‘The Standing Committee for Foodstuffs: Arguing and Bargaining in Comitology’, in M. P. C. M. van Schendelen (ed.), EU Committees as Influential Policymakers (Aldershot: Ashgate). Neyer, J. (2003) ‘Discourse and Order in the EU: A Deliberative Approach to MultiLevel Governance’, Journal of Common Market Studies, 41, 687–706. Neyer, J. (2006) ‘The Deliberative Turn in Integration Theory’, Journal of European Public Policy, 13, 779–91. Nugent, N. (2001) The European Commission (Houndmills: Palgrave). Nugent, N. (2003) The Government and Politics of the European Union. (Durham: Duke University Press). Ogul, M. S. and B. A. Rockman (1990) ‘Overseeing Oversight: New Departures and Old Problems’, Legislative Studies Quarterly, 15, 5–24.

208 Literature Ostrom, E. (1986) ‘An Agenda for the Study of Institutions’, Public Choice, 48, 3–25. Page, E. (1997) People Who Run Europe (Oxford: Clarendon Press). Pedler, R. H. and G. F. Schaefer (ed.) (1996) Shaping European Law and Policy. The Role of Committees and Comitology in the Political Process (Maastricht: European Centre for Public Affairs). PEP (1958) Agricultural Policy in the European Economic Community. Occasional paper no. 1 (London: Political and Economic Planning). PEP (1959) Agricultural Policies in Western Europe. Occasional paper no. 3 (London: Political and Economic Planning). PEP (1961) Agriculture, the Commonwealth and EEC. Occasional paper no. 14 (London: Political and Economic Planning). Peters, B. G. (1995) The Politics of Bureaucracy (White Plains, NY: Longman). Peters, B. G. and J. Pierre (2004) ‘Multi-level Governance and Democracy: A Faustian Bargain?’, in I. Bache and M. Flinders (ed.), Multi-Level Governance (Oxford: Oxford University Press). Philip, A. B. (1998) ‘The Eco-Label Regulatory Committee’, in M. P. C. M. van Schendelen (ed.), EU Committees as Influential Policymakers (Aldershot: Ashgate). Piedrafita, S. (2009) ‘The Impact of Enlargement on the Functioning of Comitology Committees’, in T. Christiansen, J. M. Oettel and B. Vaccari (ed.), 21st Century Comitology. Implementing Committees in the Enlarged European Union (Maastricht: European Institute of Public Administration). Piattoni, S. (2009) ‘Multi-level Governance: A Historical and Conceptual Analysis’, Journal of European Integration, 31, 163–80. Pierson, P. (2000) ‘The Limits of Design: Explaining Institutional Origins and Change’, Governance, 13, 475–99. Pollack, M. A. (1995) ‘Regional Actors in an Intergovernmental Play: The Making and Implementation of EC Structural Policy’, in C. Rhodes and S. Mazey (ed.), Building a European Polity? The State of the European Union, vol. 3 (Harlow/Boulder, CO.: Longman/Lynne Rienner). Pollack, M. A. (2001) ‘International Relations Theory and European Integration’, Journal of Common Market Studies, 39, 221–44. Pollack, M. A. (2003a) The Engines of European Integration. Delegation, Agency, and Agenda Setting in the EU (Oxford: Oxford University Press). Pollack, M. A. (2003b) ‘Control Mechanism or Deliberative Democracy? Two Images of Comitology’, Comparative Political Studies, 36, 125–55. Pollack, M. A. and G. C. Shaffer (2008) ‘Risk Regulation, GMOs and the Limits of Deliberation’, in D. Naurin and H. Wallace (ed.), Unveiling the Council of the European Union. (London: Palgrave). Ponzano, P. (2008) ‘ “Executive” and “Delegated” Acts: The Situation after the Lisbon Treaty’, in S. Giller and J. Ziller (ed.), The Lisbon Treaty. EU Constitutionalism without a Constitutional Treaty? (Wien: Springer Verlag). Putnam, R. D. (1988) ‘Diplomacy and Domestic Politics: The Logic of Two-Level Games’, International Organization, 42, 427–60. Quaglia, L., F. D. Francesco and C. Radaelli (2008) ‘Committee Governance and Socialization in the European Union’, Journal of European Public Policy, 15, 155–66. Rey, J. (1968) ‘Extracts from the statement by M. Jean Rey, President of the Commission, on the First General Report of the Commission of the European Communities’, EC-Bulletin, 4/1968, 8–15.

Literature

209

Rhinard, M. (2002) ‘The Democratic Legitimacy of the European Union Committee System’, Governance, 15, 185–210. Rockman, B. A. (1984) ‘Legislative–Executive Relations and Legislative Oversight’, Legislative Studies Quarterly, 9, 387–440. Sannerstedt, A. (2005) ‘Negotiations in European Union Committees’, in O. Elgström and C. Jönsson (ed.), European Union Negotiations. Processes, Networks and Institutions (London: Routledge). Savino, M. (2009) ‘The Role of Committees in the EU Institutional Order: Deliberative or Procedural Supranationalism?’, in T. Christiansen, J. M. Oettel and B. Vaccari (ed.), 21st Century Comitology. Implementing Committees in the Enlarged European Union (Maastricht: European Institute of Public Administration). Schaefer, G. R. and A. Türk (2007) ‘The Role of Implementing Committees’, in T. Christiansen and T. Larsson (ed.), The Role of Committees in the Policy-Process of the European Union. Legislation, Implementation and Deliberation (Cheltenham: Edward Elgar). Scharpf, F. W. (1989) ‘Decision Rules, Decision Styles and Policy Choices’, Journal of Theoretical Politics, 1, 149–76. Scharpf, F. W. (1997) Games Real Actors Play. Actor-Centered Institutionalism in Policy Research (Boulder, CO: Westview Press). Scharpf, Fritz W. (1999) Governing in Europe. Effective and Democratic? (Oxford: Oxford University Press). Schelling, T. (1963) The Strategy of Conflict (New York: Oxford University Press). Schindler, P. (1971) ‘The Problems of Decision-Making by way of the Management Committee Procedure in the European Economic Community’, Common Market Law Review, 8, 184–205. Schmidt, S. K. (1998) ‘Commission Activism: Subsuming Telecommunications and electricity under European Competition Law’, Journal of European Public Policy, 5, 169–84. Schusterschitz, G. and S. Kotz (2007) ‘The Comitology Reform of 2006. Increasing the Powers of the European Parliament without Changing the Treaties’, European Constitutional Law Review, 3, 68–90. Skogstad, G. (2003) ‘Legitimacy and/or Policy Effectiveness?: Network Governance and GMO Regulation in the European Union’, Journal of European Public Policy, 10, 321–38. Skogstad, G. (2006) ‘Regulating Food Safety Risks in the European Union: A Comparative Perspective’, in C. Ansell and D. Vogel (ed.), What’s the Beef? The Contested Governance of European Food Safety (Cambridge, MA: The MIT Press). Steiner, J., A. Bächtiger, M. Spörndli and M. R. Steenbergen (2004) Deliberative Politics in Action. Analyzing Parliamentary Discourse (Cambridge: Cambridge University Press). Steunenberg, B. (1994) ‘Decision Making Under Different Institutional Arrangements: Legislation by the European Community’, Journal of Institutional and Theoretical Economics, 150, 642–69. Steunenberg, B., C. Koboldt and D. Schmidtchen (1996) ‘Policymaking, Comitology, and the Balance of Power in the European Union’, International Review of Law and Economics, 16, 329–44. Steunenberg, B., C. Koboldt and D. Schmidtchen (1997) ‘Beyond Comitology: A Comparative Analysis of Implementation procedures with Parliamentary Involvement’, Aussenwirtschaft, 52, 87–112.

210 Literature Strøm, K. (2003) ‘Parliamentary Democracy and Delegation’, in K. Strøm, W. C. Müller and T. Bergman (ed.), Delegation and Accountability in Parliamentary Democracies (Oxford: Oxford University Press). Strøm, K., W. C. Müller and T. Bergman (2003) (eds), Delegation and Accoutability in Parliamentary Democracies (Oxford: Oxford University Press). Szapiro, M. (2006) ‘Comitologie: Retrospective et prospective après la reforme de 2006’, Revue du Droit de l’Union Europeenne, 501, 545–87. Thomson, R., F. N. Stokman, C. H. Achen and T. König (ed.) (2006) The European Union Decides (Cambridge: Cambridge University Press). Thomson, R. and F. N. Stokman (2006) ‘Research Design: Measuring Actors’ Positions, Saliences and Capabilities’, in R. Thomson, F. N. Stokman, C. H. Achen and T. König (ed.), The European Union Decides (Cambridge: Cambridge University Press). Tsebelis, G. (1996) ‘More on the European Parliament as a Conditional Agenda Setter: Response to Moser’, American Political Science Review, 90, 839–44. Tsebelis, G. (2002) Veto Players. How Political Institutions Work (Princeton, NJ: Princeton University Press). Tsebelis, G. and G. Garrett (2000) ‘Legislative Politics in the European Union’, European Union Politics, 1, 9–36. Tsebelis, G., C. B. Jensen, A. Kalandrakis and A. Kreppel (2001) ‘Legislative Procedures in the European Union: An Empirical Analysis’, British Journal of Political Science, 31, 573–99. Tsioumani, E. (2004) ‘Genetically Modified Organisms in the EU: Public Attitudes and Regulatory Developments’, Review of European Community & International Environmental Law, 13, 279–88. Türk, A. (2007) ‘The Role of the European Court of Justice in the Area of Comitology’, in T. Christiansen and T. Larsson (ed.), The Role of Committees in the Policy-Process of the European Union. Legislation, Implementation and Deliberation (Cheltenham: Edward Elgar). Türk, A. (2009) ‘The Development of Case Law in the Area of Comitology’, in T. Christiansen, J. M. Oettel and B. Vaccari (ed.), 21st Century Comitology. Implementing Committees in the Enlarged European Union (Maastricht: European Institute of Public Administration). Töller, A. E. (1998) ‘The “Article 19 Committee”: The Regulation of the Environmental Management and Audit Scheme’, in M. P. C. M. van Schendelen (ed.), EU Committees as Influential Policymakers (Aldershot: Ashgate). van Schendelen, R. (1996) ‘The In-Sourced Experts’, Journal of Legislative Studies, 8, 27–39. van Schendelen, M. P. C. M. (ed.), (1998) EU Committees as Influential Policymakers (Aldershot: Ashgate). Voigt, S. and E. M. Salzberger (2002) ‘Choosing not to Choose: When Politicians Choose to Delegate Powers’, Kyklos, 55, 289–310. Volden, C. (2002) ‘Delegating Power to Bureaucracies: Evidence from the States’, Journal of Law, Economics & Organization, 28, 187–220. Vos, E. (1997) ‘The Rise of Committees’, European Law Journal, 3, 210–29. Wallström, M. (2003) Statement by Commissioner Wallström in the debates of the European Parliament on 13 May 2003 on the exercise of Commission implementing powers (available at: http://www.europarl.europa.eu/, date accessed 28 July 2008).

Literature

211

Weber, M. (1970) ‘Bureaucracy’, in H. H. Gerth and C. Wright Mills (ed.), From Max Weber: Essays in Sociology (London: Routledge and Kegan Paul). Weingast, B. R. and M. J. Moran (1983) ‘Bureaucratic Discretion or Congressional Control? Regulatory Policymaking by the Federal Trade Commission’, Journal of Political Economy, 91, 765–800. Wessels, W. (1998) ‘Comitology: Fusion in Action. Politico-Administrative Trends in the EU System’, Journal of European Public Policy, 5, 209–34. Wilson, W. (1887) ‘The Study of Administration’, Political Science Quarterly, 2, 197–222. Wonka, A. (2007) ‘Technocratic and Independent? The Appointment of European Commissioners and its Policy Implications’, Journal of European Public Policy, 14, 169–89. Woolcock, S. (2000) ‘European Trade policy. Global Pressures and Domestic Constraints’, in H. Wallace and W. Wallace (eds.), Policy-Making in the European Union (Oxford: Oxford University Press).

Index

Note: Page numbers in italics refer to figures and tables in the text. actor orientation, 6, 49 advisory comitology procedure, 19, 23, 24–5, 31–2, 98–9, 108, 109, 110, 111, 119–20, 131–2, 172 advisory committee, see advisory comitology procedure Agence Europe, 12, 58 agency theory, see delegation theory agenda setting power, 37–42, 70 Amsterdam Treaty, 107 analytic narrative, 57, 71 anticipated reaction, 95, 97, 100–3, 116, 178 appeal committee, 21, 24–5 Bache, Ian, 10 bargaining interaction style, 48, 145, 158, 165–8, 168, 172, 179 Bates, Robert H., 57 Bawn, Kathleen, 45, 183 behavioural disposition, 162–5, 164, 173–4, 179 Bellamy, Richard, 188, 190–1 Bendor, Jonathan, 35, 45 Bennett, Andrew, 75 Bergström, Carl Fredrik, 4, 5, 17, 18, 22, 23, 26, 27, 37, 42, 54, 64, 72, 73, 81, 83, 98, 100, 131, 183, 189 Bradley, Kieran St Clair, 6, 7, 18, 21, 27, 42, 72, 78, 86, 148, 189 Brandsma, Gijs Jan, 3, 7, 16, 19, 74, 131, 174, 175, 189, 190 bureaucratic discretion theory, see delegation theory call-back right, 76, 82, 87, 90–1, 92 Christiansen, Thomas, 9, 18 codecision procedure, 4, 16, 18, 19, 20, 22, 23, 43, 76, 84, 87, 90, 91, 96, 97, 106, 108, 136, 139

comitology 1987 reform, 4, 23, 24–5, 26, 73, 95, 111 1999 reform, 4, 23, 24–5, 26, 73, 97–9, 99–100, 111 2006 reform, 4, 23, 24–5, 26, 72–93, 90–1, 92, 178 2011 reform, 19–21, 23, 24–5, 26, 74 and common agricultural policy, 26, 37, 41, 53, 55–7, 58–64, 99 and common commercial policy, 20, 21, 26 and common customs policy, 67–9 decisions, 3–4, 23–7, 24–5 and development aid, 65–7 guidelines, 26, 91, 99–100, 136, 139 history of, 17–18 institutional set-up of, 21–30 origins of, 5, 8, 36–42, 53–55, 70–71 preferences, 42–46, 72–74, 88–93, 94–6, 116–17 procedures, 23–7, 24–5, 189; see also advisory comitology procedure; examination comitology procedure; management comitology procedure; regulatory comitology procedure; regulatory procedure with scrutiny; safeguard procedure register, 16, 189 regulation, 19–21, 26, 74, 95 representatives, 7, 8, 46–50, 149, 153–6 Commission, European agenda setting power, 37–42, 70 comitology preferences, 44, 73, 89, 94, 116–17 212

Index position in comitology committees, 28, 28–9, 49, 168–73 as rule-maker, 3 Commission expert groups, 9, 148 committee governance, see EU committee governance committee procedure, see comitology common agricultural policy, 26, 37, 41, 53, 55–7, 58–64, 99 Constitutional Treaty, 81–83 consultation procedure, 20, 22, 73, 106, 108, 136, 139 Coombes, David, 169 Council of Ministers A and B agenda points, 127 comitology preferences, 43, 73, 89, 94, 116–17 Council working groups, 8–9, 127, 146, 148 Crombez, Christophe, 101 Customs Code Committee, 160 customs union, 67–9 Danish comitology survey, 149, 149–51 deck stacking, 42, 43, 73 deficit, see democratic deficit Dehousse, Renaud, 7, 47, 146 delegation theory, 7–8, 34–6 deliberative interaction style, 49, 145, 157, 165–8, 168, 171, 172, 179 deliberative supranationalism, 146 Demmke, Christoph, 17, 37, 54, 80, 183 democratic deficit, 185–92 development aid, 65–7 directors’ committee, 59–64 discretion, 44–5, 47–8, 50, 145, 156, 163 measure of, 156, 157, 161 distributive effects, 191 Dogan, Rhys, 5, 27, 30, 37, 96, 106, 107, 111, 118, 123 Egeberg, Morten, 7, 9, 125, 148, 151, 162 empty chair crisis, 64 Epstein, David, 11, 35, 45, 124, 128, 129, 182, 183

213

EU committee governance, 8–9 Eurocontrol, 1 European Commission, see Commission, European European convention, 81–3 European Council, 75, 76, 77 European Court of Justice, 18, 26, 55, 78, 81, 92, 100 European Development Fund (EDF) Committee, 67, 69 European Parliament amendments, 101–2, 106–7, 127 comitology preferences, 43–4, 73, 89, 116–17 examination comitology procedure, 6, 20–1, 23, 24–5, 26, 74, 95, 131 expert groups, see Commission expert groups Falke, Josef, 17, 27, 30, 118, 122 fire alarms, 36, 182, 190 first-tier rule-making, 2, 3, 4 focal point, 40–1, 54, 57, 64, 70–1, 177, 180 Follesdal, Andreas, 185, 186, 187, 191 food safety, 74, 77–81 Franchino, Fabio, 5, 7, 11, 27, 30, 35, 42, 45, 72, 96, 107, 123, 126, 128, 129, 130, 146, 147, 182, 183 Garrett, Geoffrey, 40, 101, 125 genetically modified organisms (GMOs), 77–81, 92, 148 George, Alexander L., 75 Haibach, Georg, 4, 5, 17, 18, 37, 79, 98, 131, 183 Hallstein, Walter, Commission President, 62, 193 Heritier, Adrienne, 96, 100, 106, 107 Hix, Simon, 5, 63, 100, 125, 127, 185, 186, 187, 190, 191 Hooghe, Liesbet, 10, 125, 162, 163, 171 Huber, John D., 35, 45, 125, 128, 182, 183

214 Index input legitimacy, 14, 186–92 institutional conflict, 47, 50, 121, 136, 139 measure of, 124–8, 134 institutional preferences, see comitology, preferences interaction style, 48, 49, 50, 145, 146, 157–8, 165–8, 168, 172 intergovernmental bargaining, 7, 30, 37, 41, 42, 54, 147 issue complexity, 46, 48, 50, 121, 141–2 measure of, 128–31, 134 Joerges, Christian, 7, 9, 30, 47, 146, 147, 169 Kiewiet, Roderick D., 11, 35, 182 Lamfalussy reform, 75–7, 88, 91, 148 least-likely case, 95, 117 legitimacy, 186–92 Lieberman, Evan S., 175, 185 Lindberg, Leon N., 56, 58, 193 Lisbon Treaty and comitology, 3, 4, 18–21, 22, 23, 74, 95 and democratic deficit, 185–92 McCubbins, Mathew D., 11, 35, 36, 42, 45, 46, 73, 182, 183, 190 Majone, Giandomenico, 185, 187, 188 management comitology procedure, 23, 24–5, 26, 31–2, 98–9, 108, 109, 110, 111, 119–20, 131–2, 172 management committee, see management comitology procedure Management Committee for Cereals, 65 Marks, Gary, 10 Moe, Terry M., 35, 36, 42, 73, 183 Moravcsik, Andrew, 37, 147, 166, 167, 185, 187, 188 most-likely case, 74, 90, 94, 117, 178 multi-level governance, 9–11

Neyer, Jürgen, 7, 30, 47, 146, 147, 169 Nice Treaty, 81 Nugent, Neill, 5, 125, 127, 169 Ogul, Morris S., 11, 35, 182 O’Halloran, Sharyn, 11, 35, 45, 124, 128, 129, 182, 183 Origin of Goods Committee, 68 Ostrom, Elinor, 6, 49 output legitimacy, 14, 186–92 Parliament, see European Parliament Pierson, Paul, 5, 37 police patrols, 36, 182, 190 politics of structural choice, 36, 42, 43, 73 Pollack, Mark A., 5, 7, 10, 11, 27, 28, 30, 35, 38, 42, 72, 76, 78, 123, 125, 144, 146, 147, 148, 166, 183 post-Lisbon comitology system, see Lisbon Treaty principal-agent theory, see delegation theory process tracing, 75 Prodi, Romano, Commission President, 15 Putnam, Robert D., 166, 167 reform of comitology, see comitology regulatory comitology procedure, 23, 24–5, 26, 31–2, 98–9, 108, 109, 110, 111, 119–20, 131–2, 172 regulatory committee, see regulatory comitology procedure regulatory procedure with scrutiny, 4, 24–5, 26, 31–2, 74, 87, 91, 93, 98 research agenda, 183 Rockman, Bert A., 11, 35, 182 Rome Treaty, 17, 22, 38, 55, 56, 58, 61, 65, 67 safeguard procedure, 23, 24–5, 31–2, 98–9, 108, 109, 110, 111, 119–20, 131–2 Santer, Jacques, Commission President, 15 Schaefer, Guenther F., 9, 23, 29, 73 Scharpf, Fritz W., 6, 49, 55, 185

Index Schelling, Thomas C., 40, 41 Schwartz, Thomas, 36, 182, 190 second-tier rule-making, 1, 2, 4 Shipan, Charles R., 35, 45, 125, 128, 182, 183 Single European Act, 4, 17, 22, 23, 73, 75, 183, 184 single sky committee, 1, 2, 3, 193 Skogstad, Grace, 78 Special Committee on Agriculture, 60, 61 Standing Committee on the Food Chain and Animal Health, 79, 149–50 Steunenberg, Bernard, 6, 7, 23, 30, 99, 101, 132, 146, 147 Strøm, Kaare, 35, 45 structural choice, see politics of structural choice supranationalism, see deliberative supranationalism

215

technical complexity, see issue complexity Tsebelis, George, 101, 102, 105, 125 Türk, Alexander, 9, 23, 26, 29, 73, 100 Valuation Committee, 68 van Schendelen, Rinus, 7, 9, 27, 122, 147, 169 Volden, Craig, 35, 45, 125, 128 Vos, Ellen, 5, 9, 18, 37, 54 Weber, Max, 11, 34, 34 Weingast, Barry R., 40, 143, 144 white paper on European governance, 15, 83, 92 Wilson, Woodrow, 11, 34, 35 World Trade Organization (WTO), 79, 80