Designing the New European Union, Volume 279 (Contributions to Economic Analysis) (Contributions to Economic Analysis) (Contributions to Economic Analysis)

DESIGNING THE NEW EUROPEAN UNION CONTRIBUTIONS TO ECONOMIC ANALYSIS 279 Honorary Editors: D.W. JORGENSON J. TINBERGE...

Author: Helge Berger | Thomas Moutos

13 downloads 780 Views 3MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

DESIGNING THE NEW EUROPEAN UNION

CONTRIBUTIONS TO ECONOMIC ANALYSIS 279

Honorary Editors: D.W. JORGENSON J. TINBERGENy Editors: B. BALTAGI E. SADKA D. WILDASIN

Amsterdam Boston Heidelberg London New York Oxford San Diego San Francisco Singapore Sydney Tokyo

Paris

DESIGNING THE NEW EUROPEAN UNION

Edited by Helge Berger Free University Berlin, Germany Thomas Moutos Athens University of Economics and Business, Greece

Amsterdam Boston Heidelberg London New York Oxford San Diego San Francisco Singapore Sydney Tokyo

Paris

Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK First edition 2007 Copyright r 2007 Elsevier B.V. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: [email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent veriﬁcation of diagnoses and drug dosages should be made Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN-13: 978-0-444-52968-8 ISBN-10: 0-444-52968-3 ISBN: 0573-8555

For information on all Elsevier publications visit our website at books.elsevier.com

Printed and bound in The Netherlands 07 08 09 10 11

10 9 8 7 6 5 4 3 2 1

This page intentionally left blank

Introduction to the Series This series consists of a number of hitherto unpublished studies, which are introduced by the editors in the belief that they represent fresh contributions to economic science. The term ‘economic analysis’ as used in the title of the series has been adopted because it covers both the activities of the theoretical economist and the research worker. Although the analytical method used by the various contributors are not the same, they are nevertheless conditioned by the common origin of their studies, namely theoretical problems encountered in practical research. Since for this reason, business cycle research and national accounting, research work on behalf of economic policy, and problems of planning are the main sources of the subjects dealt with, they necessarily determine the manner of approach adopted by the authors. Their methods tend to be ‘practical’ in the sense of not being too far remote from application to actual economic conditions. In addition, they are quantitative. It is the hope of the editors that the publication of these studies will help to stimulate the exchange of scientiﬁc information and to reinforce international cooperation in the ﬁeld of economics. The Editors

Contents PREFACE

xiii

LIST OF CONTRIBUTORS PART I: CHAPTER 1

1 2 3 4 5 6

CHAPTER 2

4

DESIGNING EUROPE – A SURVEY OF TASKS AHEAD Helge Berger, Thomas Moutos

Introduction Europe’s institutions: the constitution Europe’s institutions: other aspects of EU organization Fiscal and ﬁnancial policies One money, one economy? Concluding remarks References

PART II:

1 2 3

INTRODUCTION

EUROPE’S INSTITUTIONS IN PRAISE OF THE EUROPEAN CONSTITUTION: A POLITICAL ECONOMICS PERSPECTIVE Ge´rard Roland

Introduction A primer on the constitutional treaty Evaluating the constitution 3.1. A cleaned up intergovernmental institution or a parliamentary confederation? 3.2. Which public goods will the EU deliver with the Constitution? Conclusion Acknowledgement References

Comment: Lambros Pechlivanos

xv 1

3

3 5 8 16 20 25 26 31

33

33 35 40 40 47 51 52 52 53

viii

CHAPTER 3

1 2

3

4

INSTITUTIONAL ASPECTS OF EU ORGANIZATION: AN ECONOMIC ANALYSIS Massimo Bordignon

Introduction On enhanced cooperation 2.1. The problem 2.2. What are ECAs? 2.3. ECAs: the debate 2.4. The analysis 2.5. Policy implications The allocation of powers inside the EU and the ‘‘open coordination method’’ 3.1. The problem 3.2. The analysis 3.3. Results 3.4. Policy implications for the EU Concluding remarks Acknowledgement References

Comment: Stefan Voigt CHAPTER 4

1 2 3 4

5

6

WELFARE POLICY INTEGRATION INCONSISTENCIES Giuseppe Bertola

Introduction Welfare policy in an integrating world 2.1. Choices and changes Economic policy in the EU Social policy and the EU 4.1. Principles and policies 4.2. Incoherence Enlargement, delays, and reforms 5.1. Integration, delayed? 5.2. Redesigning and modernizing welfare delivery Summary and prospects Acknowledgments References Appendix

Comment: Laszlo Goerke

57

57 59 59 61 63 65 72 75 75 77 78 80 80 81 82 85

91

91 92 93 96 100 101 104 105 107 110 114 117 117 119 121

ix

CHAPTER 5

1 2

3

4 5

THE EU BUDGET: HOW MUCH SCOPE FOR INSTITUTIONAL REFORM? Henrik Enderlein, Johannes Lindner, Oscar CalvoGonzales, Raymond Ritter

Introduction Some theoretical considerations 2.1. The EU’s institutional framework between supranational and intergovernmental decision modes 2.2. Budgetary procedures - a balance between efﬁciency and legitimacy 2.3. The EU budgetary procedure under the constraint of limited political integration Reforming the EU budgetary procedure: an assessment 3.1. Proposals for reform at the general level 3.2. Proposals for reform at the level of multi-annual planning 3.3. Proposals for reform at the level of the annual procedure The state of play following the intergovernmental conference Conclusion Acknowledgement References

129

129 131 131 134 135 137 140 143 148 153 156 157 157

Comment: Margarita Katsimi

161

PART III:

165

CHAPTER 6

1 2

3

4

FISCAL AND FINANCIAL POLICIES FROM THE STABILITY AND GROWTH PACT TO A SUSTAINABILITY COUNCIL FOR EMU Ju¨rgen von Hagen

Introduction: Europe’s ﬁscal framework under stress Fiscal discipline in the EMU 2.1. The Excessive deﬁcit procedure and the stability and growth pact 2.2. Proposals for SGP reform Fiscal performance under the EDP and the SGP 3.1. Experiences in the 1990s 3.2. Fiscal policy stance 1998-2003 3.3. Patterns of ﬁscal adjustment in the EMU A sustainability council for the EMU 4.1. Mandate 4.2. Method of operation 4.3. Enforcement

167

167 168 170 172 173 173 175 177 182 183 184 185

x

5

4.4. Independence, accountability, and transparency 4.5. Appointments, composition, and resources Conclusions References Appendix: calculation of ﬁscal stance

Comment: Wolfgang Eggert

CHAPTER 7

1 2 3

4 5

6

SUPERVISION OF THE EUROPEAN BANKING MARKET Martin Schu¨ler

Introduction The rationale for integration of banking supervision in Europe Banking supervision in the EU 3.1. The national arrangements 3.2. The internationalization of banking supervision 3.3. Proposed reform of the supervisory arrangements in the EU Evaluation of the supervisory arrangements Open issues in crisis management 5.1. Lender of last resort and central bank involvement 5.2. The European situation - do we need a European LOLR? 5.3. A European observatory of systemic risk Conclusions References Appendix

186 187 188 188 189 191

195

195 197 201 201 202 205 209 210 210 212 214 215 216 220

Comment: Tuomas Takalo

223

PART IV:

229

CHAPTER 8

1 2 3 4 5

ONE MONEY, ONE ECONOMY? REASONS AND IMPLICATIONS OF INFLATION DIFFERENTIALS WITHIN THE EUROPEAN MONETARY UNION Ignazio Angeloni, Michael Ehrmann

Introduction Stylised evidence and interpretations A model of inﬂation differentials in EMU The role of inﬂation persistence Effect of changes in the monetary policy rule

231

231 233 241 244 245

xi

6

Conclusions Acknowledgement References

251 252 252

Comment: Andrew Hughes Hallett

255

CHAPTER 9

1 2

3 4 5 6 7

BLACK TIE REQUIRED? HOW TO ENTER A CURRENCY UNION? Volker Nitsch

Introduction Implementing a monetary union: two case studies 2.1. Greece 2.2. Ecuador Background and literature Data Empirical results Robustness Conclusions Acknowledgements References

263

263 264 264 265 266 267 269 280 281 282 282

Comment: Sarantis Kalyvitis

285

SUBJECT INDEX

289

This page intentionally left blank

Preface This volume is based on papers delivered at the second CESifo-Delphi Conferences (Munich, November 2003, and Delphi, June 2004), organized by CESifo and the Department of International and European Economic Studies of the Athens University of Economics and Business. The CESifo-Delphi Conferences are organized every two years and involve a two-stage process. Following an initial call for abstracts, a number of authors are selected to present their papers at a workshop meeting in Munich. After further refereeing, some of them are invited to present (possibly revised) versions of their papers at the ﬁnal conference meeting in Delphi. For more information on the CESifo-Delphi Conferences and the current call for papers, see http://www.cesifo.de.

This page intentionally left blank

List of Contributors Ignazio Angeloni

Italian Finance Ministry-International Financial Relations, Via XX Settembre 97, 00187 Rome, Italy

Helge Berger

Free University Berlin, Department of Economics, Rooms 251-254, Boltzmannstr. 20, D-14195 Berlin, Germany

Giuseppe Bertola

Dipartimento di Economia, Universita` degli Studi di Torino, Via Po 53, 10124 Torino, Italy

Massimo Bordignon

Universita Cattolica, Largo A. Gemelli, 1, 20123 Milano, Italy

Wolfgang Eggert

Fakulta¨t fu¨r Wirtschaftswissenschaften, Universita¨t Paderborn, Paderborn, Germany

Michael Ehrmann

European Central Bank, Kaiserstrasse 29, D-60311 Frankfurt am Main, Germany

Henrik Enderlein

Hertie School of Governance, Schlossplatz 1, 10178 Berlin, Germany

Laszlo Goerke

Department of Economics, Eberhard Karls University Tu¨bingen, Melanchthonstr. 30, D-72074 Tu¨bingen, Germany

Oscar Calvo-Gonzalez

The World Bank, 1818 H Street, NW, Washington, DC 204 33, USA

Ju¨rgen von Hagen

Department of Economics, University of Bonn, Lennestrasse 37, 53113 Bonn, Germany

Andrew Hughes Hallett

Department of Economics, Vanderbilt University, VU Station B #351819, 2301 Vanderbilt Place, Nashville, TN 37235-1819, USA

Sarantis Kalyvitis

Athens University of Economics and Business, Patision Str. 76, Athens, 10434, Greece

xvi

Margarita Katsimi

Athens University of Economics and Business, Patision Str. 76, Athens, 10434, Greece

Johannes Lindner

European Central Bank, Kaiserstrasse 29, D-60311 Frankfurt am Main, Germany

Thomas Moutos

Athens University of Economics and Business, Patision Str. 76, Athens, 10434, Greece

Volker Nitsch

Free University Berlin, Department of Economics, Boltzmannstr. 20, 14195 Berlin, Germany

Lambros Pechlivanos

Athens University of Economics and Business, Patision Str. 76, Athens, 10434, Greece

Raymond Ritter

European Central Bank, Kaiserstrasse 29, D-60311 Frankfurt am Main, Germany

Ge´rard Roland

University of California, Berkeley, 549 Evans Hall, 3880, Berkeley, CA 94720-3880, USA

Martin Schu¨ler

Zentrum fu¨r Europa¨ische Wirtschaftsforschung GmbH, L 7, 1 D-68161 Mannheim, Postfach 103443 D-68034 Mannheim, Germany

Tuomas Takalo

Monetary Policy and Research Department, Bank of Finland, Rauhankatu 16, FIN-00101 Helsinki, Finland

Stefan Voigt

Philipps University Marburg, Barfu¨ßertor 2, D-35032 Marburg, Germany

Part I: Introduction

This page intentionally left blank

CHAPTER 1

Designing Europe – A Survey of Tasks Ahead Helge Berger and Thomas Moutos

1. Introduction Following the accession of 10 new member states on May 1, 2004, the European Union (EU) faces the difﬁcult task of consolidating its achievements and developing an effective system for the management of interdependence among its member states. The EU’s organizational achievements are without precedent in history: past efforts at regional integration among sovereign states pale when measured against the institutions for pan-European governance which have been developed by the EU, the functions they perform, and the political ambitions enshrined in the treaties signed by the member states. Perhaps, the most important milestones of this process are the abandonment of their national currencies for 12 of the member states and their replacement by a single currency, the Euro, and the drafting of the Treaty Establishing a Constitution for Europe, or, as it is widely known, the Constitutional Treaty, or, the European Constitution. Both of these actions are of high substantive and symbolic value, and are indicative of the desire to proceed with further political integration. Although the euro, and its managing institution, the European Central Bank, are by now well entrenched in the lives of ordinary Europeans and are an important ingredient of the international monetary system, there is wide uncertainty about the fate of the Constitutional Treaty. The Treaty Establishing a Constitution for Europe was signed on October 29, 2004, in Rome. The original plan was for the European Constitution to come into effect on November 1, 2006, following its ratiﬁcation by either the national parliament or through plebiscites in each of the 25 EU countries. After the resounding no-vote by the French and

Corresponding author. CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79001-0

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

4

Helge Berger and Thomas Moutos

the Dutch voters, the EU’s leaders decided during the Luxembourg Summit on June 16–17, 2005, that the ratiﬁcation process by the rest of the countries should be delayed in order for EU ofﬁcials and the governments of the member-countries to ‘‘explain’’ to EU citizens the beneﬁts which the European Constitution will afford them.1 At the moment of writing, the earliest at which the Constitutional Treaty is expected to come into effect, if it ever comes into effect in its present form, is the middle of 2007.2 Appropriately enough, the present volume starts with an analysis of the European Constitution in its entirety (Chapter 2), and then proceeds with the analysis of particular issues that are also important determinants of the framework within which the behavior of states, corporations and individuals will unfold.3 These issues concern the optimal assignment of functions to different levels of government within the unique EU system of governance (Chapters 3 and 4) and the reform of the EU budgetary procedures (Chapter 5). Further expanding on the theme of institutional design, the volume turns to a in-depth discussion of the problems of coordination and cooperation in the areas of ﬁscal policy (Chapter 6) and ﬁnancial supervision (Chapter 7)–arguably among the more pressing issues facing Europe, both ‘‘old’’ and ‘‘new’’, today. Finally, and no less urgent from the perspective of European policy makers, there is a discussion on the importance of economic convergence in the euro area. Two important questions arise in this respect: The ﬁrst is whether the introduction of the euro has lead to a reduction in the differences in the inﬂation among the member countries, and whether the adjustment mechanism operates effectively in the absence of separate monetary policies (Chapter 8). The second is whether economic convergence before the establishment of a monetary union inﬂuences the economic performance of the participating countries ex post (Chapter 9).

1

Another difﬁculty currently faced by the EU (which may also be a side-effect of the no-vote by the French and Dutch voters) is the failure of EU leaders, at the June 2005 Luxembourg Summit, to agree on the EU’s budgetary framework for the period 2007–13. 2 In case the Constitutional Treaty fails to be ratiﬁed, there is no publicly known ‘‘Plan B’’. A substantive renegotiation of the Treaty appears unrealistic since the present form of the Constitution emerged from a complex process of consensus-building involving the Convention (which brought together, among other, representatives from the national governments, the national parliaments, the European Parliament and the Commission, observers from the Committee of the Regions and the Social and Economic Committee), and two intergovernmental conferences. The present text of the Constitution is considered a ﬁnely balanced compromise, with any change in it probably necessitating many others, thus making a revision of the Treaty highly unlikely. 3 As one might expect, the more technical (or less of a ‘‘high politics’’ status) an issue was purported to be, the easier it has been to reach a consensus among the participating actors in the drafting of the Constitutional Treaty (for example, the near absence among member states of any disagreement regarding the objectives of the European Central Bank with the threats issued by some countries to block the signing of the Constitutional Treaty unless the voting rules are amended to their favor).

Designing Europe – A Survey of Tasks Ahead

5

2. Europe’s institutions: the constitution The avowed purpose of the European Constitution is to ‘‘reconstruct’’ the European Union, replacing the various Treaties (e.g., Rome, Maastricht, Nice) with a fundamental and comprehensive document that deﬁnes the Union’s values and objectives, and organizes its structure, functioning and policies. During its 50-year history, the EU has been governed by several treaties that have been revised to reﬂect not only the increase in the number of member-states, but also, and perhaps more importantly, the desire to enshrine into law the intensiﬁcation of integration experienced by and wished for by the member-states. The three original Treaties that founded the European Communities were the Treaty establishing the European Coal and Steel Community (1951), the Treaty establishing the European Community (1957), and the Treaty establishing the Atomic Energy Community (1957). The Treaty of Maastricht (1992) created a new entity, the European Union, with a three-pillar structure: the community pillar (corresponding to the three original Treaties), the Common Foreign and Security Policy (CFSP) pillar, and the Justice and Home Affairs (JHA) pillar. The Constitutional Treaty merges the three pillars, and gives the EU a single legal personality under domestic and international law.4 The founding treaties (which include the original three plus Maastricht) have been amended several times. These amendments include the Merger Treaty (1965), the Single European Act (1986), the Treaty of Amsterdam (1997), and the Treaty of Nice (2001), which has been in force since February 1, 2003. The institutional framework deﬁned by the Treaty of Nice has been widely criticized as too complicated to deliver the decision-making efﬁciency needed to handle the challenges posed by the accession of 10 new member states to the EU, and the challenges posed by globalization (Baldwin, 2004) and the growing unease among citizens that the EU was becoming an elitist project that did not have widespread public support (Tsoukalis, 2003). In an effort to increase the legitimacy of the EU, the Laeken Declaration of 2001 set up a Convention charged with establishing the Convention for the Future of Europe. The Convention, unlike previous conferences, abandoned the traditional instrument of intergovernmental negotiations (in which the only participants are country representatives giving primal attention to their country’s interests) and brought together government representatives (of the 15 member states and 13 candidate countries), representatives of national parliaments, the European Parliament and the Commission, observers from the Committee of Regions and the Social and Economic Committee, and representatives of the European social partners.

4

The Euratom Treaty remains in existence as a separate legal entity while sharing the same institutions.

6

Helge Berger and Thomas Moutos

The Convention’s efforts culminated in a Draft Constitution, which was submitted to the European Council of Thessaloniki in June 2003. The Council decided that the Draft formed a ‘‘good’’ base for discussion at the intergovernmental conference (comprising of the prime ministers and foreign ministers of the 25 countries), which started in Rome, in October 2003. After a brief pause of the conference’s deliberations due to disagreements (between December 2003 and April 2004), a compromise was reached at the European Council in Brussels (June 2004), and the Constitutional Treaty was formally adopted by the EU leaders on October 29, 2004, in Rome. The main innovations introduced in the Constitutional Treaty are as follows: First, the values and objectives of the European Union (or Union, for short) are enshrined, as are the rights of the European citizens, thanks to the incorporation into the Constitution of the European Charter of Fundamental Rights. Second, the Union is accorded a single legal personality under domestic and international law. Third, the allocation of competences (exclusive, shared and supporting) between the member states and the Union are clearly deﬁned. Fourth, economic coordination between the countries that have adopted the euro is to be improved, and the informal role of the Euro Group is to be recognized. Fifth, the European Council, which will be chaired by a President elected for two-and-ahalf years, is formally institutionalized, and the rotating Presidency of the European Council is discarded. Sixth, the President of the Commission is to be elected by the European Parliament based on a proposal from the European Council. Seventh, a new qualiﬁed majority system is established, under which 55% of the member states representing 65% of the population will constitute a qualiﬁed majority. Eighth, countries wishing to proceed with enhanced cooperation on matters which do not fall within the Union’s exclusive competences, can do so if their number is greater than one-third of the member-states, and subject to the European Council’s approval by qualiﬁed majority voting. The public’s reaction to the European Constitution has been mixed. Although there is hardly any disagreement that the Constitution is a deﬁnite improvement on the Nice Treaty, both economic liberals and left-wing ideologues have been arguing their case against the Constitution. The ﬁrst group claims that the presence in the Constitution of terms such as social justice, social cohesion, sustainable development, and free and fair trade, is indicative of the danger that EU governance will shift in an interventionist direction. However, these claims are farfetched. The Constitution mentions these terms in the same way as the US Constitution mentions ‘‘life, liberty, and pursuit of happiness’’. These are aspirations and are not enforceable by courts of law. In fact, the Constitution does not give the EU more competence on economic issues; for example, decisions on taxation and social security remain subject to unanimity. Moreover, in Article I-2 of the Constitution, solidarity is relegated to the condition of second-rank principle, whereas Article I-3 tilts the balance in favor of economic freedoms when determining the objectives to be aimed at by the Union. For example, in Article I-3, the Union is requested to strive at ‘‘price stability’’ and at ‘‘a highly competitive

Designing Europe – A Survey of Tasks Ahead

7

market economy’’. For some scholars (e.g., Fossum and Menendez, 2005) these references point at a shift in the Union’s balance in favor of the market dimension, to the detriment of the political and social dimensions. Presumably, this is the reason why the ‘‘Nouveau Monde’’ and the ‘‘Nouveau Parti Socialiste’’ factions of the French Socialist Party oppose the Constitution, claiming that the Constitution creates an offensive against public services. Their case has recently received support from a judge at Germany’s constitutional court, who claimed that the principle of competition, enshrined in the EU Constitution, contravenes the German constitution’s principles of solidarity and fairness (Grant, 2003). It is, of course, unrealistic to imagine that any draft of a pan-European constitution would receive support across the political, social and economic spectrum. This arises not only because there are differences in economic interests, but also because there are differences in how the EU is conceived. Some scholars (e.g., Moravcsik, 1998) conceive the EU as a functional organization whose purpose is to resolve problems which the member states cannot solve when acting independently. This implies that the EU must be empowered with an efﬁcient law-making process, which is possible if its legal order is grounded on a set of legal norms enshrined in a constitutional treaty. Other scholars (Habermas, 1996) prefer to regard the EU as a political community that is founded on the citizens’ mutual acknowledgment of their rights and duties. A constitutionally established democratic decision-making process is then seen as the best way to guarantee the stability of this arrangement. The European Constitution certainly includes elements of both the problemsolving (e.g., the allocation of competencies between the Union and the member states with respect to economic issues) and the rights-based (e.g., by making the Charter of Fundamental Rights in the EU an explicit part of the primary law) conceptions of the European Union. The presence of both conceptions of the EU in the Constitution implies that the Treaty has moved beyond the issues mentioned in the Laeken Declaration regarding the future of Europe; although the Convention was directed to evaluate alternative proposals for the ‘‘grand issues’’ of European integration, the draft produced by the Convention was a comprehensive document which is akin to national constitutions in the sense that it ensures a modicum of autonomy to its legal order, and it regulates the production of legal norms within speciﬁcally delineated realms of action. The paper by Gerard Roland (Chapter 2) notes that the autonomy of legal order which the Constitution provides to the EU, in conjunction with sufﬁcient ambiguities in the interpretation of some articles (which most likely reﬂect the complexity of negotiations involved due to existing political constraints), is conducive to some uncertainty regarding future developments.5 Nevertheless,

5

This probably partly explains the rejection of the Treaty in the French and Dutch referendums, and why both economic liberals and those wishing to see a more social Europe are opposed to it.

8

Helge Berger and Thomas Moutos

Roland’s evaluation is largely positive. He uses the framework developed by Persson et al. (1997, 2000), who emphasize the difference between Parliamentary and Presidential systems in terms of accountability and public good provision, and tries to predict the evolution of the EU’s policies on the basis of the constellation of forces produced by the Constitution. From a methodological point of view this implies that instead of simply analyzing the allocation of competences within Europe, it starts with an analysis of the allocation of powers within the institutions, in order to predict how the allocation of powers will affect the future constellation of competences and the efﬁciency of decision making. Persson et al. conclude that, given the political constraints, the Constitutional Treaty has stricken a reasonable compromise in most of the cases, thus helping to provide a framework which will both stand the test of time and enhance the efﬁciency of problem-solving (by providing the necessary public goods) at the European level. One fruitful way to think of the making of constitutions, is to perceive that they are made under a ‘‘veil of ignorance’’ with respect to the future needs and allocation of power among the interested stakeholders; this also implies that constitutions are like ‘‘incomplete contracts’’, leaving some agents with some power in the form of residual control rights (see, Persson and Tabellini, 2003). In this sense, Roland’s observations regarding the uncertainty of interpretation of certain articles in the Constitution reﬂect the impossibility of getting all the major stakeholders to agree to an unambiguous formulation of all matters dealt with in the Constitutional Treaty. Nevertheless, Roland observes that the Treaty provides considerable ﬂexibility for changing both the competences of the Union and the voting procedures. By requiring unanimity for such changes, the Constitution has managed to provide mechanisms that guarantee the further deepening of European integration and the efﬁciency of decision-making, while protecting national sovereignty. This is further strengthened by the provisions enshrined in the Constitution regarding enhanced cooperation. 3. Europe’s institutions: other aspects of EU organization The 50-year project of European integration has been characterized by continuous expansion in terms of both geography and policy functions, and from the need to create an institutional design that can strike a delicate balance between reaping the economic beneﬁts of a large internal market and retaining some degree of national sovereignty. This process of continuously widening and deepening integration has fundamentally altered the ways in which states, institutions, and individuals interact. European integration started with the design of economic instruments for the management of the customs union created by the original six members, and still the EU remains largely an economic organization. Nevertheless, much of the process of European integration has involved the development of institutions which mediate between the state and the market, and whose inﬂuence varies signiﬁcantly between one policy area to another. For

Designing Europe – A Survey of Tasks Ahead

9

example, in the cases of monetary, competition, and agricultural policy there is a high degree of centralization of the decision-making, whereas in the cases of ﬁscal, social, and redistributive policy the division of powers between the Union and national policy-makers gives more power to the latter. This highly complex institutional set-up is coupled with an extensive network of organized economic and political interests, and a continuing and increasing stream of Union-wide laws and regulations, thus creating the semblance of a transnational federation. Yet, the EU possesses a political-economic structure unlike any other encountered in a European or non-European nation-state (Alesina et al., 2005a). First, whereas the degree of inter-jurisdictional labor mobility is an important aspect of nations having a federal structure and of the theory of ﬁscal federalism (Sinn, 2003), there is a limited amount of inter-national labor mobility within the EU. Second, the size of ﬁscal transfers within the member states of the EU is far larger than the ﬁscal transfers across countries. Third, especially after enlargement, the heterogeneity of tastes is probably much larger across EU countries than across federal districts within a nation. Fourth, the EU does not have its own genuine resources through taxes directly collected by the Union, but only has what is an euphemism for own resources, which are in fact transfers by each member state to the Union. Fifth, even though the Union’s budget is very small in comparison to existing federations (i.e., bounded to be less than 1.27% of Union Gross National Income (GNI)), the Union is not allowed to borrow as its budget must be balanced at all times. These differences should alert scholars from relying excessively on knowledge acquired from the analysis of other federations in order to analyze the design of EU institutions. As a ﬁrst step in the analysis of EU organizational structures, one must deﬁne the objectives of the ‘‘European project’’. A widely held view among social scientists (not least economists) is that European integration can best be understood as a series of rational choices made by national leaders (Moravcsik, 1998). These choices were shaped by constraints and opportunities arising from the (mainly) economic interests of powerful groups within each country and the relative power of each state in the international system. The assumption that states act rationally or instrumentally in pursuit of their self-interest in the European arena implies that governments make internal decisions ‘‘as if’’ they were efﬁciently pursuing a weighted, stable set of underlying preferences given a constrained choice of means. This is the assumption also made in recent politicoeconomic analyses that examine the inﬂuence of lobbies in the design of policy in both domestic and international settings (see, for e.g., Grossman and Helpman, 2001). Theories of political economy explain international cooperation as an effort to arrange mutually beneﬁcial policy coordination among countries whose domestic policies have an impact on one another. If markets render the preferred policies of each country incompatible or permit a unilateral policy to be adjusted without costs to achieve a particular goal, a zero-sum game exists with little

10

Helge Berger and Thomas Moutos

incentive for cooperation. If, on the other hand, mutual policy adjustments can create beneﬁcial policy externalities or eliminate negative ones more effectively than unilateral responses, governments ﬁnd it in their interest to coordinate their activities (Bagwell and Staiger, 2002). The political economy explanation differs from a purely ‘‘economic’’ explanation in that it takes into account both the efﬁciency as well as the ‘‘special interest politics’’ consequences of policy coordination.6 Externalities-reducing international policy coordination very often also creates winners and losers among domestic groups of political actors, who may have the power to intervene and change the policy outcomes that would have been achieved on the basis of comparison of only economic costs and beneﬁts. Domestic producer groups (in contrast to consumers, taxpayers and rest-of-the-world producers) usually are the most able to intervene and alter the design of economic policy. This is probably most evident in the design of trade policy (see, Rodrik, 1995), and it reﬂects the ability of producer groups to not only organize efﬁciently, but also their ability to collectively wield some structural power, especially when governments intent on promoting growth have to encourage business investment by satisfying some broad business demands (Olson, 1965). A central issue with which the EU had to struggle with since its inception is the heterogeneity of preferences regarding policy outcomes among its member states. The Treaty of Nice was supposed to be able to deliver efﬁcient decision–making rules for a Union of 25 countries with very different priorities, aspirations and ability to harness the forces unleashed by global and European integration. Nevertheless, it was quickly realized that this Treaty made it even more difﬁcult for the European Council to reach decisions even in those areas formally under qualiﬁed majority rule (Baldwin, 2004). However, both the Treaty of Nice and the Constitutional Treaty provide a way out of the near impossibility of ﬁnding EU policies that could beneﬁt all, or most of the member countries. This is related to the possibility that at least a third of the member countries to go on with further integration (i.e., to form sub-unions within the EU) on some particular issue, following ex-ante agreed upon decision and governance rules, without the ability of other member countries to veto the implementation of such ‘‘Enhanced Cooperation Agreements’’ (ECAs), subject to the Council’s approval.7 This is a ‘‘No Veto-No Exclusion’’ agreement in the sense that neither can the initially non-interested countries block other countries from forming an ECA, nor the countries which have initially opted out of the ECA

6

A purely economic explanation would not go very far in explaining the formation of the EEC, since a unilateral move to free trade by the six countries that established the customs union 50 years ago may well have been the more efﬁcient policy. Subsequent policy decisions (e.g., the common agricultural policy, the eastern enlargement) are also hard to justify on purely efﬁciency considerations alone (see, Adam and Moutos, 2004) for a political economy analysis of the eastern enlargement). 7 The Council decision is taken via Qualiﬁed Majority Voting.

Designing Europe – A Survey of Tasks Ahead

11

can be excluded from joining at a later stage if they so wish on the condition that they comply with the decisions already taken by the ECA. Among economists the verdict on ECAs is rather positive (Baldwin et al., 2001, Alesina et al., 2005b), whereas political science and law scholars are ambivalent. In his contribution to the volume, Massimo Bordignon (Chapter 3), provides a political economy analysis of the issue. His main point is that the environment under which policy decisions are taken is both uncertain (or stochastic) and dynamic. It is uncertain in the sense that political contingencies change in a way that cannot be predicted at the present time, and dynamic in the sense that a country’s decision not to join an ECA initially may have an impact on the rules it will have to accept if it decides to join the ECA at a later point in time. In other words, what happens initially (when the ECA is formed) may well affect the degree and/or the type of integration which will be achieved at a later date. Bordignon ﬁnds that on efﬁciency grounds, larger heterogeneity among countries and lack of compensating transfers8 makes ECAs desirable; on the other hand, lack of commitment with respect to the policy rules which a non (initially) participating country may have to adopt when it later decides to join, makes ECAs less appropriate. An important policy implication of this analysis is that both the Treaty of Nice and the Constitutional Treaty make it more likely for countries not willing to see some policy harmonization at the present time9 to nevertheless agree to join an ECA in order to avoid having to join at a later stage and having to adopt standards and policies which they may have had a chance to inﬂuence to their beneﬁt had they initially joined the ECA. This may well also imply that the Constitutional Treaty’s rules regarding ECA formation may make it less likely to see the development of ECAs, and more likely to force reluctant countries to accept more harmonization or centralization of policy. Bordignon also tackles the related issue of the proper allocation of functions (or competencies) between the EU institutions and national governments. According to Oates (1972), functions with more (less) externalities and less (more) heterogeneity of preferences across jurisdictions should be centralized (decentralized). However, this decision rule is derived under the assumption of welfare maximizing governments, an assumption which seems at variance with the widespread demands that the EU should work towards eliminating its ‘‘democratic deﬁcit’’. Bordignon’s analysis is based on a political economy model in which policy decisions are inﬂuenced by the actions of private lobbies. He ﬁnds that if national lobbies have a common interest on an EU policy, then it is better to decentralize the making of the policy, since decentralization induces a sort of competition across lobbies when in fact there is none; on the other hand, if each national lobby has conﬂicting interests with other national lobbies (e.g., in the

8

Compensating transfers can be used by a central planner to impose an efﬁcient, EU-wide policy which could leave some countries worse-off. 9 Uncertainty about the effects of the policy may induce countries to adopt a wait-and-see attitude.

12

Helge Berger and Thomas Moutos

case of production subsidies to domestic ﬁrms), then it is better to centralize policy making, as national governments do not take into account the losses that their policy (e.g., domestic subsidies) have on the proﬁts of foreign ﬁrms. Still, another important implication of Bordignon’s analysis is that in the presence of negative externalities, lobbies are more likely to have conﬂicting interests and hence centralization is preferable, while when there are positive spillovers, lobbies’ interests are more aligned and hence decentralizing policy making becomes preferable. The issue of the proper allocation of functions with regard to welfare (or, social) policy is also dealt with in Guiseppe Bertola’s paper (Chapter 4). This is one of the issues for which the EU’s principle of subsidiarity has been consistently applied, and it reﬂects the historical impermeability of Europe’s diverse welfare states. Within the EU, the citizenship-based Scandinavian model of universal social protection coexists with the ‘‘Bismarckian’’ employment-based model of Northern-European countries; with the Anglo-Saxon model of residual social protection; and with the more idiosyncratic, less mature system of social protection prevailing in the Southern EU countries (Esping-Andersen, 1999). Social policies across EU countries remain uncoordinated, sometimes with conﬂicting objectives, and often they are far from simple to understand or implement. EU countries differ in the relative importance attached to redistributive goals, in the instruments used, in the way in which social policy interacts with other economic policies, and in the effectiveness with which the stated social policies achieve their aims. This state of affairs remains despite the obvious interactions between economic integration and social policy: integration reduces the effectiveness of social policies by increasing the range of opportunities to both households and ﬁrms to escape from mandatory re-distributive schemes, thus generating the possibility of a race to the bottom in the provision of social policy. Indeed, there is some evidence that in response to the greater integration introduced by the Single Market Programme, many EU countries have been changing their structure of social policy by, among other things, introducing stricter eligibility requirements for social protection programs, increasing the ﬁnancial incentives for ﬁnding (and keeping) a job, and providing incentives for greater reliance on the private provision of social-policy ‘‘goods’’ (Kalisch et al., 1998). The pressures for reform of the decentralized and fragmented national systems of welfare policy are becoming stronger in the enlarged EU, since it is feared that migration of workers from the relative poor Central and Eastern European Countries (CEECs) to the more developed EU countries, as well as the migration of capital in the opposite direction, is bound to test the resilience of national social policy arrangements. For some economists this is a challenge that should be welcomed, since the ensuing competition among systems will act as a disciplinary device, thus leading – much as the competition between private economic entities – to more efﬁcient outcomes (Oates and Schwab, 1988). Yet,

Designing Europe – A Survey of Tasks Ahead

13

as forcefully argued by Sinn (1997, 2003), the competition between systems may well suffer from the same problems that have lead societies to develop institutions to deal with the undesirable consequences of the unfettered operation of markets. By and large, governments have taken over all those activities which the private sector has been unable to carry out efﬁciently. This implies that the reintroduction of ‘‘markets’’ through systems competition cannot always be relied upon to deliver the desired outcomes. If the welfare states of the more advanced EU countries are the result of deliberative policy making reﬂecting the preferences of their citizens, the introduction of systems competition may well lead to a sub-optimal erosion of the welfare state as each country tries to offset the drainage of its elastic tax base to lower taxing jurisdictions. But what should in this case be the appropriate institutional framework for dealing with the ensuing competition between national welfare systems in wellintegrated economies? The possible failure of systems competition to produce desirable outcomes does not necessarily imply the indiscriminate harmonization of welfare policies across the EU.10 Much as in the case of German uniﬁcation, the price of harmonization may be that of mass unemployment in lower productivity countries. Richter (2002) and Sinn and Ochel (2003) have proposed an intermediate solution whereby some time elapses before the welfare system from which a migrant draws welfare beneﬁts switches from the original to the destination constituency (delayed integration). A justiﬁcation for this solution is that migration11 may, in the short-run, be excessive as it can be misguided by welfareshopping considerations; the number of migrants in the long-run will be smaller as inward capital ﬂows would raise productivity and real wages in initially poor countries thus reversing the initial migration ﬂows. Nevertheless, Bertola expresses some doubts about the empirical importance of welfare-shopping migration and argues that both the design and implementation of delayed integration schemes are problematic, whereas their adverse effects on (efﬁcient) labor mobility should not be ignored. Bertola ‘s preference is for a system that differentiates between social policies whose purpose is to correct market failures and the absence of private insurance markets covering employment risks (the insurance motive), and social policies motivated by feelings of solidarity and aimed at preventing social unrest (the redistribution motive).12 Regarding social policies pertaining to the ﬁrst motive,

10

Sinn (2003) presents a model in which a national government’s choice of national standards regarding the quality of the workplace is efﬁcient from a supra-national perspective, thus nullifying the need for supra-national harmonization of working conditions. 11 It is assumed that the act of migration involves substantial psychic costs for those concerned since most people prefer to stay in their own countries even when wage differentials are large. 12 The distinction between insurance and redistribution motives is not always clear in actual policy frameworks as, ex post, every re-distributive contract implies a redistribution from the fortunate to the unfortunate, and ex ante, most of the re-distributive activities undertaken by governments can be interpreted as insurance (Sinn, 2003).

14

Helge Berger and Thomas Moutos

Bertola proposes that there should neither be coordinated nor funded at an EU level, since social insurance need not be subject to downward pressure from ﬁscal competition when there is a clear link between contributions and beneﬁts at a national level (see also, Atkinson, 1998). Regarding the redistribution motive, he suggests that in order to ensure the long-run sustainability of social protection and free mobility of persons within the EU, the social rights of European citizens should be deﬁned in the form of minimum welfare standards at the EU level (see also, Bertola et al., 2001). Moreover, and in order to avoid the prospect of mass unemployment, minimum assistance levels could be speciﬁed on a relative basis, and should be adjusted to reﬂect local earnings and price levels. Finally, these minimum assistance levels should be co-ﬁnanced by speciﬁc budget line items of the EU’s budget. The size of the relevant budget line should not be larger than that of the current Structural, Cohesion or Common Agricultural Policy funds, which if used for that purpose could lift all EU citizens out of extreme poverty. The issue of EU budget reforms is dealt with by Henrik Enderlein, Johannes Lindner, Oscar Calvo-Gonzalez, and Raymond Ritter in Chapter 5. Given the failure of EU leaders to reach an agreement on the 2007-13 Financial Perspective (the multi-annual budgetary framework), at the June 2005 Luxembourg Summit, this issue has again resurfaced as a top priority item for EU policy-makers. The failure to agree on the EU budget is commonly attributed to two issues. The ﬁrst concerns the Common Agricultural Policy (CAP), and the second the socalled ‘‘UK rebate’’.13 However, both of these issues have arisen as a consequence of the Eastern Enlargement in 2004. The reason is that the new member states are on average much poorer and more agricultural than the incumbents. Yet, according to the agreement reached at the 2002 EU15 summit in Copenhagen (when the newcomers had no voting rights), the newcomers were given an average annual allocation of CAP and Structural and Cohesion Funds (SCFs) that is not only substantially lower than what is allocated to the poorer of the EU15 members, but also lower compared to the EU15 as a whole. According to Baldwin (2005a), if the rules for the disbursement of SCFs applying to the incumbent members were applied to the newcomers, each of these countries should have received much more than the amount promised to them. Baldwin calculates that if Poland (the more populous of the newcomers) had received ‘‘equal treatment’’, it would have been allocated h380 per person from the SCFs, instead of the h214 that it got, whereas Lithuania (the poorest of the 10 countries) should receive h411 per person, instead of the h69 it was actually

13

The UK rebate arose out of Margaret Thatcher’s efforts to reduce the large (according to the GNP per capita criterion) amount of net contributions that the UK was making to the EU budget. According to the agreement reached in 1984, the rebate is roughly equivalent to two-thirds of the net contribution that the UK would have made without the rebate. This amount is added to the net contributions of all other member states – a fact that creates frictions every year.

Designing Europe – A Survey of Tasks Ahead

15

allocated.14 The newcomers will also have some misgivings about the agreement reached (among the EU15) regarding the allocation of CAP support, which has been ﬁxed at h172 per farm for the 10 new member states, whereas the relevant number for the EU15 is h5,000. Since this difference in CAP support cannot be explained on the basis of ‘‘objective’’ criteria (for example, average farm size for the EU15 is only twice as large as for the 10 new members – 18.7 and 9 hectares, respectively), it could only be expected that the acquisition of voting power by the new member states would create difﬁculties in reaching agreement on budget issues. It would, however, be far-fetched to believe that without the recent enlargement there would be no demands for reforming the EU budget. The PadoaSchioppa Report (Padoa-Schioppa et al., 1987) had already expressed its misgivings about the Community’s efforts to execute ‘‘distributive policies at the level of individual persons or small enterprises’’ through its budget. And in 2003, the Sapir Report (Sapir et al., 2003) suggested a radical overhaul of the EU budget by shifting away from traditional expenditure items (such as the CAP), and regrouping spending into three new instruments (or funds): the ‘‘growth fund’’, aimed at fostering growth within the EU by subsidizing R&D activities, education, training and infrastructure; the ‘‘convergence fund’’, aiming to help low-income countries catch-up with the rest; and the ‘‘restructuring fund’’, targeted at smoothing the process of resource reallocation necessary for deeper and wider economic integration (by giving aid to a shrinking agricultural sector and providing income support to displaced workers in general). With respect to the revenue side of the EU budget, the Sapir Report suggests that in order to avoid the overt character of national contributions, which create the tendency for national governments to focus the debate on the net contributions (or, juste retour) issue, thus preventing an efﬁcient allocation of the EU budget, the EU should ﬁnd sources of revenue for its budget that have either a clear EU dimension or have a very mobile tax base within the EU (e.g., seignorage earned from issuing euro banknotes, capital income taxes, and stock exchange taxes). In their contribution to the volume, Enderlein, Lindner, Calvo-Gonzalez, and Ritter review the main criticisms of the EU budgetary procedures and assess the main reform proposals that have been suggested thus far. The discussion is informed by the constitutional choice literature (Buchanan and Tullock, 1962), which allows for both a normative and positive account of the origins of constitutional orders and their legitimacy and efﬁciency, and a comparison with the budgetary procedures prevailing in the United States. The focal point of the

14

The comparison between Lithuania and Poland (whose population and number of votes in the European Council are much larger than Lithuania’s) provides some evidence for the hypothesis that the higher the political power of a country, the higher is – ceteris paribus – its ability to get an agreement providing it with more funds than it would have received on the basis of ‘‘objective’’ criteria.

16

Helge Berger and Thomas Moutos

constitutional choice literature is that (economic) policy-making generates conﬂicts among different groups of voters (citizens), among different groups of politicians, and between voters and politicians (Persson and Tabellini, 2003). The resolution of these conﬂicts – and thus the policy outcomes we observe – can be understood as the result of the pooling of individual citizen’s sovereignty and the delegation by the citizens, who act as principals, to the elected representatives (the agents), the functions of administering the public institutions and policy-making. For some scholars (e.g., Alesina et al., 2005a), the same analogy can be used to explain institution building at the supra-national level, with the states regarded as principals, who may decide to pool their sovereignty in selected areas. This may be done in two ways: by intergovernmental cooperation (whereby states agree to take decisions under the principle of unanimity, thus preserving a high degree of sovereignty), and by supranational governance (whereby states ﬁnd it preferable to delegate, as principals, certain actions to an independent institution). The higher the weight given to supranational governance, the higher is the degree of political integration. The EU is currently governed by a combination of both these means of decision-making, and this limited degree of political integration acts as a constraint on the political feasibility of budgetary reforms. In this context, Enderlein, Lindner, Calvo-Gonzalez, and Ritter argue that the present state of political integration in the EU does not allow the implementation of reform proposals that could increase both the efﬁciency and the legitimacy of the EU budget. In their view, the ‘‘embeddedness’’ of issues relating to the budget in the overall state of European integration gives to the current institutional design, by-and-large, the properties of an equilibrium (or, as some people may describe it, a stalemate), whereby any welcome change in it is conditioned on further political integration. As an example of the application of their analysis, consider the proposals for modifying the expenditure structure and the creation of an ‘‘EU tax’’ mentioned earlier. Among the alleged beneﬁts of such reforms would be greater efﬁciency due to the provision of more public goods and the conveyance of a strong symbolic message of a ‘‘United Europe’’. Nevertheless, the passage of such reforms requires the willingness, among member states, to allocate further tasks to a supranational authority, something that the member states are currently unwilling to do. 4. Fiscal and financial policies A much-observed litmus test of the political will of the EU, or more speciﬁcally, the euro area member countries, to act cooperatively is the Stability and Growth Pact (SGP) – and to some extent this is also true for the way ﬁnancial supervision is organized within Europe. The European Economic and Monetary Union (EMU) changed the rules of the game in both areas, introducing the possibility of large-scale negative externalities from inappropriate national policies. Regarding ﬁscal policy, Uhlig (2002) and others have stressed the free rider

Designing Europe – A Survey of Tasks Ahead

17

problem that occurs when national policy makers use deﬁcits to compensate the actions of the European Central Bank (ECB) at the European level. In a more general sense, the absence of bilateral exchange rates within the euro area brought to the fore the possibility of excessive national debt levels threatening the common European good price stability, for instance through effects on the foreign value of euro or the threat of bail-outs. Similarly, regulation failures at the national level are more likely to produce negative cross-border effects as ﬁnancial markets become increasingly integrated within the EMU area, making effective coordination and cooperation of national authorities ever more important. The call for prudent ﬁscal policies is an old one, and – at least on paper – the existing legal framework leaves little room for complaints in this regard. The Treaty on the EU in its Article 4(3) deﬁnes ‘‘sound public ﬁnances’’ as one of the guiding principles of economic policy in the EU, reﬂecting demands of the 1989 Delors Report to introduce institutional safeguards for ﬁscal discipline. The legal framework, including the no-bail-out clause and the excessive deficit procedure as concretized by the original SGP, introduced a detailed monitoring process for the public ﬁnances to ensure sustainability and to keep ﬁnancial liabilities at the national level. However, by most accounts the SGP failed on enforcement, in particular with regard to larger member states (see, among others, Berger et al., 2004). The incentive to force other countries to play according to the rules is low, in part because the externalities of not adhering to the SGP might be small in the short run and because of ‘‘horse-trading’’ between actual or potential large-country SGP-offenders. The lack of an impartial surveillance and enforcement mechanism rendered the Pact largely ineffective. The reform of the SGP in 2005 might do little to heal the implementation problem.15 In short, the amended SGP combines a more ﬂexible assessment of ﬁscal deﬁcits with the possibility of direct ‘‘early policy advise’’ by the Commission and an agreement of the required ‘‘minimum ﬁscal adjustment’’ in cases where the ﬁscal stance is considered excessive. While, in principle, the reform kept the rule-based character of the old SGP intact, even the Commission seems to agree that it is far from clear whether the reformed Pact will be more effective. Or, as the Commission’s Regling (2005) puts it: ‘‘The rules are good. What ultimately matters is their effective implementation.’’ In Chapter 6, Ju¨rgen von Hagen argues that a more fundamental change in the framework of ﬁscal coordination and cooperation is needed to ensure sustainability. He argues that increasing ﬂexibility, while providing relief in some

15

The European Commission in April 2005 launched the legislative procedure to amend the SGP following the agreement on the reform of the Pact at the European Council in March. The Council remains responsible for the adoption of the reform, consulting the European Parliament and the European Central Bank.

18

Helge Berger and Thomas Moutos

situations, will do little to avoid the basic conﬂict between long-run budget principles and government claims that the constraints are too rigid in the short term. What is more, the demise of the old SGP has shown that there is little reason to believe that governments can be ‘‘credible judges’’ of their own policies – and nor can the Commission, which has to rely on the cooperation of member states in achieving a large variety of other goals. The solution, argues von Hagen, might be the introduction of an independent Sustainability Council for EMU. The Sustainability Council, legally to be set up by the European Parliament, would consist of experts in the ﬁeld of public ﬁnance considered to be independent from governments. Its main task, helped by a small number of staff and with a guarantee for full access to national information, would be impartial surveillance, monitoring, and ultimately evaluation of national ﬁscal developments. Other than the governments themselves in the present framework, the Stability Council could have the means and the independent political standing to weigh the necessities of long-run ﬁscal sustainability and the need for short-run ﬂexibility without compromising the credibility of the SGP framework. Operationally, the von Hagen-proposal would have national governments submit their annual and medium-term budget plans to the Stability Council, which would then –in a transparent and highly visible way – determine their compatibility with long-term ﬁscal sustainability and, in cases of conﬂict, issue ‘‘unambiguous limits’’ for the change in national public debts. Importantly, the Stability Council would make its evaluation and the underlying reasoning public, for instance, in a written report to the European Parliament, including speciﬁc suggestions for necessary adjustments in national ﬁscal policies. Being independent of the Commission and the Council of Ministers, appealing to public opinion would be the Stability Council’s main (and only) enforcement channel. However weak this might sound, or so argues von Hagen, it might still achieve more than the SGP’s thoroughly discredited mechanism of self-enforcement. Financial supervision might be the other major area of governance where designing the new EU could mean changing the traditional ways in which policies are conducted in Europe. Most observers agree that the introduction of the euro has contributed to an ongoing process of ﬁnancial integration within EMU member countries (see, among others, Galati and Tsatsaronis, 2003). Fratzscher (2002) shows that this also extends to equity markets. Another important factor was the completion of the EU’s Financial Services Action Plan in 2004, aimed at removing regulatory obstacles to the free ﬂow of capital and ﬁnancial services within the union. On the beneﬁt side, increasing market integration promises efﬁciency gains. On the other hand, conventional wisdom has it that closer links between ﬁnancial markets also increase the potential for systemic risk (i.e., the risk that failure of a single ﬁnancial institution may lead to the failure of other institutions and, ultimately, the ﬁnancial system in its entirety) in the euro area. While failing banks or ﬁnancial institutions were mostly a national problem prior to the euro’s introduction, they might be mostly a European one today.

Designing Europe – A Survey of Tasks Ahead

19

Supervisory coordination of Europe’s security markets is governed by the socalled Lamfalussy model, and current policies call for an extension of this approach to banks and ﬁnancial institutions (ECB, 2004).16 The Lamfalussy model is based on a consecutive four-level procedure that divides the legislation process toward a European supervisory arrangement into high-level framework provisions and implementing measures: at Level 1, the Council and the European Parliament decide on the broader legislative framework; at Level 2, a committee comprising the Commission and national supervisors speciﬁes and implements the Level 1 legislation; at Level 3, committees of national supervisors in consultation with ﬁnancial market participants work on the convergence of actual regulatory practices; ﬁnally, Level 4 envisages the enforcement of all measures in EU member countries. This approach leaves room for tension between a still mostly regional supervision structure and an increasingly integrated market for ﬁnancial services within the EU and euro area. To be sure, European supervisors already cooperate both formally and informally along many dimensions. For instance, the ‘‘Memorandum of Understanding on cooperation between the EU banking supervisors and central banks in crisis situations’’, signed in 2003 (the new member states signed in 2004) does determine basic principles of coordination and information exchange; and similar agreements exists between Nordic authorities. Moreover, the Level 3 committees of the Lamfalussy approach will further foster supervisory convergence and coordination. That said, the Lamfalussy approach stops short of supervisory centralization. And despite these efforts, most supervisory action – from information gathering to closure –still remains (and is scheduled to remain) at the national level. In his contribution to the volume (Chapter 7), Martin Schu¨ler argues that the existing level of coordination and cooperation between European supervisors is not sufﬁcient for safeguarding ﬁnancial stability and should be strengthened by the introduction of a uniﬁed European Observatory of Systemic Risk. To ﬁll the gap between ﬁnancial internationalization and supervisory regionalism, and following Aglietta (2000), he demands that the new institution be given sufﬁcient staff and resources as well as unrestricted access to national information to effectively monitor ﬁnancial sector risk at the European level, issue warnings about looming troubles, and facilitate appropriate regulatory and supervisory

16

In 2002, the Council of Economic and Finance Ministers opted for extending the Lamfalussy model to banking sector legislation as well as to the insurance and pensions sectors based on the EFC (2002) report. The European Banking Committee and the European Insurance and Occupational Pensions Committee replicate the European Securities Committee at Level 2 of the process. At Level 3, the Committee of Banking Supervisors, the Committee of European Insurance and Occupational Pensions Supervisors will fulﬁll the same functions as Committee of European Securities Regulators in securities market regulation.

20

Helge Berger and Thomas Moutos

reactions. Moreover, the Observatory could help ensure common supervisory standards, supporting the Lamfalussy approach. Schu¨ler is careful to point out that the need for a more encompassing European perspective is not restricted to prudential supervision to prevent ﬁnancial turmoil, it is also a necessity when crises occur. The argument for a more centralized approach to providing Lender of Last Resort functions and coping with failure or bankruptcy of ﬁnancial institutions within the EU and euro area mirrors the one for centralized supervision: when banks or other ﬁnancial institutions operate on a European level, the decision to support or close an institution also takes on a European dimension. This holds true for a number of reasons. First, within the euro area, obviously only the ECB can provide immediate liquidity support for troubled ﬁnancial institutions. Second, in case of insolvency with cross-country ownership and cross-country market activities, the cost and beneﬁts of bailouts or restructuring will be spread over a number of countries, creating a strong need for cooperation at the European level. Third, a European perspective is crucial to accumulate and evaluate the information necessary to determine the appropriate cause of action in either case. Whereas many observers already view the ECB as the natural focus point for efforts to organize crisis management at the EU level, Schu¨ler argues for more transparency in this respect – in essence demanding a more explicit role for another (albeit already functioning) highly centralized European institution.

5. One money, one economy? The euro and the ECB may be the most prominent beacons on the road to an ever more tightly integrated Europe – but the common currency also sheds light on some of the problems inherent in the attempt to integrate an economically diverse set of countries. There seems to be a (perhaps growing) sense of persistent heterogeneities within the euro area – ranging from differences in trend growth and business behavior to important idiosyncrasies in nominal variables –all of which have the potential to call into question the very basis underlying the currency union (see, e.g., Fels, 2004). And indeed, if economic developments still differ starkly between member countries and regions after more than ﬁve years into the EMU, the disadvantages of having given up ﬂexible bilateral exchange rates might outweigh the beneﬁts to be reaped from increased market integration (by all members) and price stability (by some). Going a step further, Italy’s Welfare Minister Roberto Maroni, in June 2005, even demanded that the country should hold a referendum to bring back the lira (Financial Times, 2005). With its national currency back in place, or so the argument goes, a country unhappy with the ‘‘one size ﬁts all’’ monetary policy conducted by the ECB might better be able to steer economic policy according to its own needs.

Designing Europe – A Survey of Tasks Ahead

21

However, the fact that EMU might not qualify as an optimal currency area (OCA) in the sense of Mundell, must not necessarily imply that it never will. Famously, Frankel and Rose (1998) and Rose (2000) argue that trade integration (and, as a consequence, real convergence in general) might very well be a consequence rather than a precondition for real convergence.17 These results have not gone unchallenged (see, Baldwin’s (2005b) discussion). A prominent example is Persson (2001), who points out that Rose’s empirical speciﬁcation might be inappropriate since countries that belong to currency unions are systematically different from the rest of the sample on which the analysis is based. On the other hand, Micco et al. (2003) provide evidence suggesting that the introduction of the euro lead to increases in bilateral trade ﬂows within the euro area already in the late 1990s compared to a control group of other industrialized countries. A common problem shared by the endogeneity literature, however, is that it might underestimate the role of institutions other than the common currency. For instance, Berger and Nitsch (2005) argue that the creation of EMU was a continuation (or culmination) of a series of previous policy changes – from the Marshall Plan that helped reconstructing the pre-war European division of labor to the EU’s ‘‘single market’’ initiative – that have led over the last ﬁve decades to greater economic integration among the countries that now constitute the euro area. They argue that there is strong evidence of a gradual increase in trade intensity between European countries rather than a euro-induced growth spurt after 1999. Glick and Rose’s (2002) result that the effect of currency union entries/exits may be subject to extremely long lags (estimated at about three decades) seems to point in a similar direction. The potential endogeneity of OCA criteria and their likely interaction with other dimensions of institutional integration emphasize the importance of learning more about the determinants of the economic heterogeneity observed in the euro area today – and this is where Ignazio Angeloni and Michael Ehrmann contribute to the debate (Chapter 8). They employ a stylized aggregate supplyand-demand model of the euro area that can be used to analyze differences in business cycle and inﬂation developments. Their particular attention is on euro area differences in inﬂation – a very worrying aspect of euro area heterogeneity for monetary policy – and the question they ask is what are the driving factors behind the idiosyncrasies observed.

17

Similar effects might be at work when it comes to ﬁnancial integration (see above). Corsetti and Pesenti (2002) stress that currency unions might be ‘‘self-validating’’ even without endogenous real or ﬁnancial convergence if ﬁrms adjust their pricing behavior in line with the monetary regime. Alesina et al. (2002) point to the fact that, when inﬂation is rooted in credibility problems, it is the act of joining a (stable) monetary union and/or of forgoing monetary policy independence that brings about convergence of inﬂation levels.

22

Helge Berger and Thomas Moutos

Empirically, the evidence points at a number of explanations. One factor stressed by Angeloni and Ehrmann is inﬂation persistence. While differences in inﬂation rates within the euro area have nudged up in recent years and tend to exceed those in the U.S., a number of statistical problems, including differences in size of the underlying regions, make a direct comparison to other currency areas difﬁcult. At the same time, however, the observed inﬂation differentials tend to be more persistent in the euro area compared to the U.S. This could be driven by cross-country differences in institutional environments such as the wage-setting framework, giving rise to varying degrees of inﬂation persistence. In such a setting even small national shocks – for instance, from national ﬁscal policies – could lead to persistent medium-term differences in inﬂation rates. In addition, there is evidence of other factors being at work as well. These include differences in the degree to which exporters are pricing-to-market, real convergence processes along the lines of the Balassa-Samuelson effect, and differences in the way common shocks (e.g., unexpected changes in the euro exchange rate or monetary surprises) translate into inﬂation at the regional level (see, among others, Cecchetti, 1999, Honohan and Lane, 2003). The latter argument regarding differences in transmission mechanisms has received some attention recently. Benigno (2004), for instance, has argued that it might be optimal to take these differences into account when making area-wide monetary policy decisions. Using a simple AD-AD model based on a hybrid Phillips curve allowing for country-speciﬁc pass-through of the nominal exchange rate of the euro to inﬂation and a backward-looking demand equation, Angeloni and Ehrmann are able to nest these alternative explanations of observed inﬂation differentials in a uniﬁed simulation approach. Their results indicate that any ‘‘satisfactory’’ account of euro area inﬂation differences needs to include a combination of shocks (both national supply and demand shocks, and external shocks) and price level convergence. Importantly, they also note that varying degrees of inﬂation persistence in individual countries go a long way in explaining the size and duration of inﬂation differences in the euro area. In fact, in the limit, inﬂation differentials can be shown to increase inﬁnitely as persistence increases – not a desirable outcome from the perspective of European policy makers. One policy implication is that to bring the euro area closer to Mundell’s idea of an optimum currency areas (OCA), measures are needed to limit inﬂation persistence at the national level and monetary policy should be aimed at stabilizing aggregate inﬂation. The discussion so far is of crucial importance to the ongoing process of EMU enlargement: If important, OCA criteria are endogenous and can (or should) be inﬂuenced by economic policy, what is the reason behind enforcing ex ante convergence for new EMU entries? The so-called Maastricht criteria demand convergence of inﬂation rates within reach of the euro area’s best performers, in addition to, among other things, the well-known upper limits on deﬁcits and

Designing Europe – A Survey of Tasks Ahead

23

debt levels.18 In light of the above, these conditions seem at least debatable. Hausmann and Powell (1999), for instance, maintain that there are, in general, little or no pre-conditions to introducing a foreign currency as national tender. Artis (2003), on the other hand, maintains that the Maastricht criteria remain a helpful framework of reference –notwithstanding their possible endogeneity. Entering this debate, Volker Nitsch, in the volume’s concluding Chapter 9, takes a helpful closer look at the performance of countries after entering a currency union, concluding that convergence prior to joining a currency union rarely matters. He notes, ﬁrst, that not all entrants into currency unions are treated equally. While, for instance, the 11 founding members of the EMU went through a lengthy adjustment process that brought national inﬂation and interest rates in line with each other before 1999, and the 12th member, Greece, entered in 2001 only after an additional waiting period dedicated mostly to nominal and (some – as we know today) ﬁscal adjustment. Ecuador’s decision in 2000 to unilaterally introduce the US-dollar as legal tender was all but thoughtout and hardly any of the pre-conditions usually associated with an optimum currency areas (OCA) were met. Nevertheless, or so argues Nitsch, both countries have fared well in the aftermath of their decision. A problem common to empirical studies of this kind is a severe lack of observations. For instance, the extensive Glick and Rose (2002) data set on transitions of monetary regimes involving currency unions contains only some 15 entries. Excluding observations that lack sufﬁcient data to allow statistical analysis and entries based on political uniﬁcations (i.e., Germany in 1990), there are even less observations. However, by carefully pairing the performance of joining countries with existing member countries as a control group, Nitsch manages to obtain a number of interesting results and stylized facts. For instance, in line with the related literature (e.g., Edwards, 2001), it appears that, compared to current members, entering countries tend to be small, burdened with higher inﬂation, and have worse ﬁscal performance. There is also some evidence of convergence of export performance between entering countries and

18

The Maastricht convergence criteria described in Article 121 of the EU Treaty include: (1) the achievement of a high degree of price stability (i.e., an average rate of inﬂation, observed over a period of one year before the examination, that does not exceed by more than 112 percentage points that of, at most, the three best performing Member States in terms of price stability); (2) the sustainability of the government financial position (i.e., a ratio of the planned or actual government deﬁcit to GDP below 3% and a ratio of government debt to GDP below 60%); (3) the observance of the normal fluctuation margins provided for by the exchange rate mechanism of the European Monetary System for at least two years, without devaluing against the currency of any other Member State (emphasis is placed on the exchange rate being close to its central rate against the euro); and (4) the durability of convergence y being reflected in long-term interest-rate levels (i.e., the average nominal long-term interest rate should not exceed by more than two percentage points that of, at most, the three best performing Member States in terms of price stability).

24

Helge Berger and Thomas Moutos

the currency area they are about to join – which could be a sign that trade liberalization or other forms of institutional integration precede the decision to enter the currency union. But perhaps the most surprising results stemming from Nitsch’s exercise is that there is only very limited evidence that economic convergence prior to joining a currency union inﬂuences performance ex post. Increasing trade intensity ex ante is mostly irrelevant for a country’s growth performance (both relative to the pre-entry period and compared to the currency union’s older members) and only has a modest impact on business cycle correlation. The same result holds for convergence of inﬂation rates and exchange rate stability, which are important elements of the Maastricht criteria for joining the euro area. As for ﬁscal policy, there is some evidence that lower debt levels and deﬁcits prior to joining a currency union are associated with somewhat better relative growth performance ex post, but the results seem not to be very robust and driven mostly by the euro area countries and sample outliers. So, are the Maastricht criteria obsolete and should the EMU abolish its current approach to expanding the euro area? Summing up, there are two main arguments that suggest a ‘‘yes’’ for an answer. The empirical ﬁndings of Chapter 9 show that based on the (admittedly scarce) historical evidence there is little or no reason to expect ex ante convergence to inﬂuence ex post performance. In addition, there is the endogeneity of the criteria to consider. Frankel (2004), referring to the evidence discussed earlier in this section, puts the traditional OCA literature on its head when he suggests that the new EU members may better qualify for the optimum currency area criteria after they have joined the euro area. And indeed, if currency unions are good for trade, and trade fosters growth and business cycle synchronicity, their adopting the euro earlier – even when not (yet) fulﬁlling the Maastricht criteria – is sensible policy advice. But it might be too soon to close the book on pre-accession convergence. First, ex post convergence might not be totally independent of initial conditions. For instance, Hughes-Hallett and Piscitelli (2001) argue that while introducing a common currency tends to foster business cycle synchronicity, this endogeneity effect requires a certain level of initial symmetry in national shocks and institutional structure across countries.19 This is in line with Krugman (1993), who argues that, from a theoretical point of view, there is no clear theoretical argument that trade liberalization (or, for that matter, trade integration) must foster convergence. Trade liberalization might also lead to increasing specialization in line with comparative advantage and, thus, to more divergent business cycles in a currency area. Second, as discussed above, the empirical evidence on endogeneity is still subject to discussion. Especially in the European case there is reason to believe that the positive impact of the euro introduction on trade is not

19

Compare the discussion in Fidrmuc (2004).

Designing Europe – A Survey of Tasks Ahead

25

independent from institutional integration (Berger and Nitsch, 2005). Since institutional integration requires both time and incentives, starting the process early and making its ultimate goal, accession to the euro area, dependent on its success, would have advantages. Finally, given the intense debate about the SGP (see, Chapter 6), current euro area members might defend the Maastricht criteria, in particular the ﬁscal area, as a screening-device that tests the commitment of potential entrants to run prudent ﬁscal policies. 6. Concluding remarks The European project has faced and continues to face many challenges. But, as this survey shows and the chapters contained in this volume vividly illustrate, today’s EU continues to be a dynamic political and economic system that develops along a number of dimensions. The ongoing, sometimes even heated, discussion of the Constitution as well as other aspects of the EU’s organization – from the reform of the budgeting process to welfare policy – proves to be a useful engine of change, pointing towards ways to make the existing decision making more efﬁcient and, perhaps, more democratic. What is true at the constitutional level, rings even truer for some of the more pressing operational issues that increasing economic integration has brought to the fore. Problems that deserve particular attention are ﬁscal policy coordination (the infamous Stability and Growth Pact) and the organization of ﬁnancial supervision, but other issues loom – for instance, the overhaul of the EU’s agricultural policies demanded by many. Finally, the European Central Bank rightfully worries about some of the surprisingly persistent heterogeneities of economic development in the euro area, and it seems worthwhile to explore the implications these might have for the operationalization of the Maastricht convergence criteria as the euro area membership expands. A ﬁnal question is whether the crisis that continues to revolve around the European Constitution poses a serious risk to the European project as such. After all, it is very easy to develop gloomy scenarios for the future: the combination of slow economic growth and widening regional disparities along with unfavorable demographics and a growing impotence to affect events beyond Europe’s borders could gradually lead to the re-nationalization of European politics and the unraveling of many achievements of the past. Although such a scenario cannot – a priori – be ruled out, we regard it as unlikely. On the contrary, if the discussion surrounding the Constitution triggers a debate about the meaning and direction of the EU, the crisis could prove to be a blessing in disguise. Such a debate could help limit the elitist character of European integration so far, by creating attention and a sense of stakeholding and ownership among a still rather diverse populace that so far has decided to treat the European project as an event outside their interest and control. And anchoring the evolving ‘‘New EU’’ more solidly on a consensus among Europeans seems like a good idea – even when this means crossing some uncharted waters.

26

Helge Berger and Thomas Moutos

References Adam, A. and T. Moutos (2004), ‘‘The Political Economy of EU Enlargement: Or, Why Japan is not a Candidate Country’’, in: H. Berger and T. Moutos, editors, Managing European Union Enlargement, Cambridge: MA: MIT Press. Aglietta, M. (2000), ‘‘A Lender of Last Resort for Europe’’, pp. 31–67 in: C. Goodhart, editor, Which Lender of Last Resort for Europe?, London: Central Bank Publication. Alesina, A., I. Angeloni and L. Schuhknecht (2005a), ‘‘What Does the European Union Do?’’, Public Choice, Vol. 123, pp. 275–319. Alesina, A., I. Angheloni and F. Etro (2005b), ‘‘International Unions’’, American Economic Review, Vol. 95, pp. 602–615. Alesina, A., R. Barro and S. Tenreyro (2002), ‘‘Optimal currency areas’’, NBER Working Paper No. 9072. Artis, M. (2003), ‘‘Reﬂections on the optimal currency area criteria in the light of EMU’’, International Journal of Finance & Economics, Vol. 8(4), pp. 297–307. Atkinson, A. (1998), Poverty in Europe, Oxford: Basil Blackwell. Bagwell, K. and R. Staiger (2002), The Economics of the World Trading System, Cambridge, MA: MIT Press. Baldwin, R., E. Berglof, F. Giavazzi and M. Widgren (2001), Nice Try: Should the Treaty of Nice be Ratified?, London: Monitoring European Integration 11, CEPR. Baldwin, R. (2004), ‘‘Managing EU enlargement’’, in: H. Berger and T. Moutos, editors, Managing European Union Enlargement, Cambridge, MA: MIT Press. Baldwin, R. (2005a), ‘‘The real budget battle: Une crise peut en cacher une autre’’, CEPS Policy Brief No. 75. Baldwin, R. (2005b), ‘‘The euro’s trade effects’’, Typescript, University of Geneva. Benigno, P. (2004), ‘‘Optimal monetary policy in a currency area’’, Journal of International Economics, Vol. 63(2), pp. 293–320. Berger, H. and V. Nitsch (2005), ‘‘Zooming out: the trade effect of the euro in historical perspective’’, CESifo Working Paper, No. 1435. Berger, H., J. de Haan and D.J. Jansen (2004), ‘‘Why did the stability and growth pact fail?’’, International Finance, Vol. 7(2), pp. 235–260. Bertola, G., J.F. Jimeno, R. Marimon and C. Pissarides (2001), ‘‘EU welfare systems and labour markets: diverse in the past, integrated in the future?’’, in: G. Bertola, T. Boeri and G. Nicoletti, editors, , Welfare and Employment in a United Europe, Cambridge, MA: MIT Press. Buchanan, J.M. and G. Tullock (1962), The Calculus of Consent: Logical Foundations of Constitutional Democracy, Ann Arbor: University of Michigan Press.

Designing Europe – A Survey of Tasks Ahead

27

Cecchetti, S. (1999), ‘‘Legal structure, ﬁnancial structure and the monetary policy transmission mechanism’’, pp. 170–194 in: Bundesbank, editor, The Monetary Transmission Process, Recent Developments and Lessons for Europe, London, UK: Palgrave. Corsetti, G. and P. Pesenti (2002), ‘‘Self-validating optimal currency areas’’, Typescript, University of Rome. Economic and Financial Committee [EFC] (2002), ‘‘EFC report on ﬁnancial regulation, supervision, and stability’’. Edwards, S. (2001), ‘‘Dollarization: myths and realities’’, Journal of Policy Modeling, Vol. 23, pp. 249–265. Esping-Andersen, G. (1999), Social Foundation of Postindustrial Economies, Oxford: Oxford University Press. European Central Bank (2004), ‘‘Developments in the EU framework for ﬁnancial regulation, supervision and stability’’, Monthly Report, November, pp. 81–93. Fels, J. (2004), ‘‘Euro wreckage?’’, European Economics, Morgan Stanley Europe, 22 January. Fidrmuc, J. (2004), ‘‘The endogeneity of the optimum currency area criteria, intra-industry trade, and EMU enlargement’’, Contemporary Economic Policy, Vol. 22(1), pp. 1–12. Financial Times (2005), ‘‘Italian minister moots return of the lira’’, 3 June. Fossum, J. and A.J. Menendez (2005), ‘‘The draft constitutional treaty: between problem-solving treaty and rights-based constitution’’, European Institute of Public Administration, Working Paper No. 2005/W/04. Frankel, J. (2004), ‘‘Real convergence and euro adoption in central and eastern Europe: trade and business cycle correlations as endogenous criteria for joining EMU’’, KSG Working Paper No. RWP04-039. Frankel, J. and A. Rose (1998), ‘‘The endogeneity of the optimum currency area criterion’’, The Economic Journal, Vol. 108, pp. 1009–1025. Fratzscher, M. (2002), ‘‘Financial market integration in Europe: on the effects of EMU on stock markets’’, International Journal of Finance & Economics, Vol. 7(3), pp. 165–193. Galati, G. and K. Tsatsaronis (2003), ‘‘The impact of the Euro on Europe’s ﬁnancial markets’’, Financial Markets, Institutions & Money, Vol. 12, pp. 165–221. Grant, C. (2003), ‘‘The liberal case for the constitution’’, Wall Street Journal, 30 October. Grossman, G. and E. Helpman (2001), Special Interest Politics, Cambridge, MA: MIT Press. Glick, R. and A. Rose (2002), ‘‘Does a currency union affect trade? The timeseries evidence’’, European Economic Review, Vol. 46, pp. 1125–1151. Habermas, J. (1996), Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy, Cambridge, MA: MIT Press.

28

Helge Berger and Thomas Moutos

Hausmann, R. and A. Powell (1999), ‘‘Dollarization: issues of implementation’’, Typescript, Inter-American Development Bank. Honohan, P. and P. Lane (2003), ‘‘Divergent inﬂation rates in EMU’’, Economic Policy, Vol. 18, pp. 358–394. Hughes-Hallett, A. and L. Piscitelli (2001), ‘‘The endogenous optimal currency area hypothesis: Will a single currency induce convergence in Europe?’’, Typescript, University of Strathclyde. Kalisch, D., T. Aman and L.A. Buchele (1998), ‘‘Social and health policies in OECD countries: a survey of current programmes and recent developments’’, OECD Labor Market and Social Policy Papers No. 33. Krugman, P. (1993), ‘‘Lessons of Massachusetts for EMU’’, pp. 241–261 in: F. Giavazzi and F. Torres, editors, The Transition to Economic and Monetary Union in Europe, New York: Cambridge University Press. Micco, A., E. Stein and G. Ordon˜ez (2003), ‘‘The currency union effect on trade: early evidence from EMU’’, Economic Policy, Vol. 37, pp. 315–356. Moravcsik, A. (1998), The Choice for Europe: Social Purpose and State Power from Messina to Maastricht, London: UCL Press. Oates, W. (1972), Fiscal Federalism, New York: Harcourt-Brace. Oates, W.E. and R.M. Schwab (1988), ‘‘Economic competition among jurisdictions: Efﬁciency enhancing or distortion inducing?’’, Journal of Public Economics, Vol. 35, pp. 333–354. Olson, M. (1965), The Logic of Collective Action, Cambridge, MA: Harvard University Press. Padoa-Schioppa, T., M. Emerson, M. King, J.C. Milleron, J.H.P. Paelinck, L.D. Papademos, A. Pastor and F.W. Scharpf (1987), Efficiency, Stability and Equity: A Strategy for the Evolution of the Economic System of the European Community, Oxford: Oxford University Press. Persson, T. (2001), ‘‘Currency unions and trade: How large is the treatment effect?’’, Economy Policy, Vol. 33, pp. 433–448. Persson, T., G. Roland and G. Tabellini (1997), ‘‘Separation of powers and political accountability’’, Quarterly Journal of Economics, Vol. 112, pp. 310–327. Persson, T., G. Roland and G. Tabellini (2000), ‘‘Comparative politics and public ﬁnance’’, Journal of Political Economy, Vol. 108, pp. 1121–1161. Persson, T. and G. Tabellini (2003), The Economic Effects of Constitutions, Cambridge, MA: MIT Press. Regling, K. (2005). ‘‘The reform of the stability and growth pact: view from the commission’’, Paper presented at The ECB and its Watchers VII conference, June 3, Frankfurt (http://www.ifk-cfs.de/papers/Panel II, Klaus Regling, DG ECFIN.pdf). Richter, W. (2002), ‘‘Social security and taxation of labor subject to subsidiarity and freedom of movement’’, Swedish Economic Policy Review, Vol. 9, pp. 75–78.

Designing Europe – A Survey of Tasks Ahead

29

Rodrik, D. (1995), ‘‘The political economy of trade policy’’, in: G. Grossman and K. Rogoff, editors, Handbook of International Economics, Vol. III, Amsterdam: Elsevier. Rose, A. (2000), ‘‘One money, one market: estimating the effect of common currencies on trade’’, Economic Policy, Vol. 30, pp. 9–45. Sapir, A., P. Aghion, G. Bertola, M. Hellwig, J. Pisani-Ferry, D. Rosati, J. Vinals and H. Wallace (2003), An Agenda for a Growing Europe – The Sapir Report, Oxford: Oxford University Press. Sinn, H.W. (1997), ‘‘The selection principle and market failure in systems competition’’, Journal of Public Economics, Vol. 66, pp. 247–274. Sinn, H.W. (2003), The New Systems Competition, Oxford: Blackwell. Sinn, H.W. and W. Ochel (2003), ‘‘Social union, convergence, and migration’’, Journal of Common Market Studies, Vol. 41, pp. 869–896. Tsoukalis, L. (2003), What Kind of Europe, Oxford: Oxford University Press. Uhlig, H. (2002), ‘‘One money, but many ﬁscal policies in Europe: What are the consequences?’’, CentER Discussion Paper, pp. 2002–2032.

This page intentionally left blank

Part II: Europe’s Institutions

This page intentionally left blank

CHAPTER 2

In Praise of the European Constitution: A Political Economics Perspective Ge´rard Roland

1. Introduction The European Constitution in June 2004 got the clearance of the European Council after the failure of the Rome summit in December 2003. It is, however, not clear how the Constitution will be ratiﬁed in the different Member States of the European Union. Referenda are planned in various countries. I have been following the process starting from the Laeken declaration to the ﬁnal text of the Constitutional Treaty ((CT): I will refer to it as the European Constitution in the text) and have written with some colleagues a report on the challenges facing Europe and the desirable features a European Constitution should have (Berglo¨f et al., 2003). However, the adopted Constitution is very different from the one I had dreamt of. Especially, I thought that it would be better to have a presidential system rather than a parliamentary system for Europe. The Convention clearly chose for the parliamentary path, as I will explain in this chapter. However, while reading the Constitution, I was positively surprised to ﬁnd that there were quite a few safeguards in place against the drawbacks of a parliamentary Europe. On the whole, I think that the Convention delegates have done a very good job in designing this Constitution, given the constraints, especially the various political constraints to further European integration in various Member States. The version of the Constitution adopted by the European Council is incomplete in many respects (including in some key areas such as qualiﬁed majority), than the one drawn up by the Convention and has generated some disappointment; but it is not drastically different from the Convention draft and there are reasons to be quite optimistic about Europe’s future if the Constitution is adopted. CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79002-2

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

34

Ge´rard Roland

In hindsight, the simple existence of a European Constitution appears to be a very unlikely event. The prospect of the increased membership of the European Union to 25 had raised fears that the EU would simply become a large free-trade area and that the post-World War II European project of peace on the continent via a closer political integration may never materialize. The pessimists claimed that Europe had lost its historical opportunity after the breaking down of the Berlin Wall in 1989. The Nice Treaty had made decisions on voting weights in the European Council and on the number of seats in the European Parliament after the EU increased in size. It seemed inevitable that decision-making would become much more difﬁcult in the enlarged EU (Baldwin et al., 2001). However, a seemingly unimportant event, the institution of a Convention to establish the Charter of Fundamental Rights, was to have far-reaching consequences. The Belgian presidency prepared a declaration called the Laeken declaration, which was destined to renew the impetus for reform and to move farther ahead than the bland Nice Treaty. The most revolutionary affect of the declaration would prove to be the abandonment of the traditional instrument of the intergovernmental conference and the decision instead to mandate a Convention to prepare these reforms. Intergovernmental conferences are always composed of country representatives who only have the interests of their country in mind. This often usually leads to wasted discussions. The Convention was to be composed not only of representatives of national governments but also of members of the European Parliament and members of national parliaments. Even more interestingly, it included representatives from the newly joined countries. Members of the Convention were to act not as country representatives but as conventioneers jointly trying to prepare a draft Constitution. Despite a very strict deadline and sometimes idiosyncratic presidency by the aging former French president Valery Giscard d’Estaing, the Convention fulﬁlled its task. The Intergovernmental Conference (IGC) that took place in the fall of 2003 could not make any progress over the work of the Convention. The important questions I would like to address are: What is the historical importance of this Constitution? Will it really create institutional stability? Will it provide an effective and democratic decision-making mechanism at the European level? Will the institutions stemming from the Constitution deliver the public goods needed at the European level? My general answer to these questions is: Yes, but with a great deal of uncertainty. I will argue that the Constitution has all the necessary ingredients for historical success similar to that of the US Constitution. However, I will also argue that it does not contain guarantees of success. The reason is not that a good Constitution is never a guaranteed condition of success but that the Constitution contains sufﬁcient ambiguities so as to allow for developments in different directions, including good and not so good scenarios. These ambiguities are the reﬂection of the political constraints in Europe. In that sense, it would be futile to criticize them as ill designed. One must however be aware that these ambiguities leave a lot of room for uncertainty and that there will most likely be many battles over the

In Praise of the European Constitution

35

interpretation of the Constitution after its adoption. Nevertheless, even the worse-case scenarios one can imagine are still better than the status quo of the Nice Treaty. While it is difﬁcult to deﬁne with certainty what is an optimal Constitution for Europe, my overall assessment is that the Convention has done a very good job. More importantly, I think that the current Constitution has the potential to last the test of time, which is the most important test for any Constitution. From a methodological point of view, I will take the political economics perspective when analyzing the Constitution. Instead of analyzing the allocation of competences within Europe, I go one step further and look instead at the allocation of powers within the institutions, trying to predict how this allocation of powers will affect the allocation of competences. This is why most of this paper is devoted to analyzing the allocation of powers resulting from the Constitution. In Section 2, I will go over the main elements of the Constitution. In Section 3, I will comment on how the Constitution is likely to work in the long run. When doing so, I will go over the many ambiguities in the Constitution and how they might evolve. I will also argue that it is deﬁnitely an improvement over the current status quo. In Section 4, I will use the methodology of political economics to analyze to what extent the Constitution will help provide the necessary public goods at the European level. 2. A primer on the constitutional treaty The Constitution is composed of four parts. Part I though short is the most important. It deﬁnes the objectives of the Union, clariﬁes its competences, deﬁnes the powers of the main institutions, and the main rules of functioning. Part II is the charter of fundamental rights that was proclaimed at the Nice summit but did not have the force of Law. The inclusion of the Charter is a natural step given the importance of fundamental rights in democratic Constitutions. Part III not only contains a lengthy catalogue of the Union’s competences but also contains more detailed provisions than Part I regarding the functioning of the Institutions and the European budget. Part IV contains various protocols including important points on the application of the subsidiarity principle and on the role of national parliaments in the EU. I will only review the main elements of the Constitution that are of importance. I have based these elements on the version adopted by the European Council in June 2004.1 Article I-2 deﬁnes the values of the Union: human dignity, liberty, democracy, equality, the rule of law, respect for human rights,

1

It will need some cleaning up in the numbering of the articles as, e.g., some articles have been scrapped but the numbering has remained unchanged.

36

Ge´rard Roland

pluralism, non-discrimination, tolerance, justice, solidarity, and equality between women and men. These values are not only the basis for admission of new members but also the basis for possible exclusion of Member States. Among the Union’s objectives in article I-3, there is the mention of a social market economy, social progress and explicit reference to equality between men and women,and to the protection of children’s rights. Article I-4 explicitly mentions the prohibition of discrimination on grounds of nationality in the application of the Constitution; this article will certainly have widespread repercussions in the future. Article I-5a declares the primacy of EU law over national laws and the obligation of Member States to fulﬁll their obligations resulting from the Constitution or the Union Institutions’ acts. The latter clause is expected to be used in future legal disputes between the EU and Member States. Article I-6 gives the Union a legal personality, through which it will be able to sign Treaties with other nations. This also paves the way for direct representation of the Union in international organizations, a step that would undoubtedly give it more weight at the international level. Article I-9 deﬁnes the principle of subsidiarity: ‘‘in areas which do not fall within its exclusive competence the Union shall act only if and insofar as the objectives of the intended action cannot be sufﬁciently achieved by the Member States (y) but can rather, by reason of the scale or effects of the proposed action, be better achieved at the Union level. The principle of proportionality is also deﬁned: ‘‘the content and form of Union action shall not exceed what is necessary to achieve the objectives of the Constitution’’. Article I-11 deﬁnes three categories of competence: (1) exclusive competence, (2) shared competence under the primacy of EU law, implying that EU law suppresses national competence to legislate, and (3) supportive measures, ‘‘carrying out actions to support, coordinate or supplement the actions of the Member States, without thereby superseding their competence in these areas’’. No harmonization of Laws is allowed in the latter category. The promotion and coordination of economic and employment policies is explicitly mentioned as the ‘‘competence to deﬁne and implement a common foreign and security policy, including the progressive framing of a common defense policy’’. A very important article is Article I-17, the Flexibility Clause that allows an extension of EU powers proposed by the Commission within the objectives of the Constitution by unanimous decision of the Council and consent of the European Parliament. It should be noted that under the subsidiarity principle, national parliaments would be consulted on all such changes. Article I-18 deﬁnes a single Institutional Framework for the EU, thereby abolishing the ‘‘three pillar system’’ of the Amsterdam Treaty. Article I-19 deﬁnes the role of the European Parliament (EP): It shall colegislate with the Council and elect the President of the European Commission. It shall not exceed 750 members, with a minimum of 6 and a maximum of 96 seats for any Member State. The decision on the speciﬁc composition of this

In Praise of the European Constitution

37

body is left to a unanimous decision by the Council on the initiative of the European Parliament. Article I-20 on the European Council states that its decisions shall be taken by consensus instead of unanimity. Article I-21 deﬁnes the role of the president of the European Council: elected by qualiﬁed majority of the Council (and recallable by the same procedure) for two and a half years, renewable once. The president’s main role is not only that of chairing Council meetings and driving forward its work but also externally representing the Union ‘‘without prejudice to the responsibilities of the Union Minister for Foreign Affairs’’. Initially, both French President Jacques Chirac and UK Prime Minister Tony Blair had insisted on giving executive power to the Council chair. Former German chancellor Gerhard Schroeder had even agreed to the proposal. The Constitution clearly circumscribes the power of the European Council chair to being currently similar to that of the German president rather than that of the French president. However, the 6-month rotating presidency of the European Council has been abolished as the smaller countries opposed it. The outcome is thus a compromise between both positions. I will argue in Section 3 that this is an effective compromise. Article I-22 deﬁnes the role of the Council of Ministers and speciﬁes that the general decision rule shall be qualiﬁed majority. Article I-23 deﬁnes the General Affairs Council, the Foreign Affairs Council and other Council conﬁgurations. The COREPER (Committee of Permanent Representatives of Member States responsible for preparing the agenda for Council meetings) has also been deﬁned. Deliberations and voting on legislative acts in the Council shall be in open meetings. The European Council shall decide on the presidency of Council conﬁgurations on the basis of a principle of equal rotation, while the Foreign Affairs Council shall be chaired by the Union Foreign Minister (Article I-27.3). Article I-24 deﬁnes the rule for qualiﬁed majority: 55% of Member States (at least 15 countries) and 65% of the population.2 However, a proposal can be blocked by four Council Members. This was the most controversial article that initially provoked opposition from Spain and Poland, and led to a breakdown of the negotiations under the Italian presidency in 2003. The current article, with the higher majority threshold but also the possibility of a blocking minority, is the reﬂection of a compromise in the European Council. A very important ‘‘deepening clause’’ in the Constitution (previously I-24 and now IV-7a) is that the European Council can decide by unanimity (and after a ‘‘cooling off’’ period of 6 months and consultation of the European Parliament and informing the national parliaments) to use the ordinary legislative procedure

2

Proposals not emanating from the Commission or the Union Foreign Minister require a supermajority of 72% of Member States and 65% of the population.

38

Ge´rard Roland

when the Constitution requires a special procedure and to use the normal qualiﬁed majority in areas where the Constitution speciﬁes the unanimity rule. This procedure will not require ratiﬁcation or referenda. The Constitution thus gives powers to the European Council to change by unanimity the legislative procedure and voting rule on a given topic. As we will argue below, this will be a very important tool that will allow the EU to evolve gradually without having to change the Constitution. This clause has nevertheless been somewhat toned down by the European Council. Article IV-7a states that defense and decisions with military implications are excluded from the clause and also that any national parliament can veto a decision (within 6 months) taken under that clause. Article I-25 deﬁnes the power of the European Commission. The Convention proposed a limit of 15 Commissioners: the Commission president, the Union Foreign Minister and 13 Commissioners based on a system of equal rotation between the Member States with non-voting Commissioners from other states. The current draft statesthat after 2014 there shall be a limit of two- thirds of the number of Member States (16 with 25 Member States), unless the European Council decides to alter this ﬁgure by unanimity. The Commission shall represent the demographic and geographic range of the EU and there will be equal rotation between Member States in such a way that the difference between the total number of terms between any pair of countries is never more than one. According to article I-26, the president of the Commission will be proposed by the Council, and decided by qualiﬁed majority taking into account the elections to the European Parliament. The Parliament must elect the president by majority. The Council shall appoint the other members of the Commission, jointly with the president-elect on the basis of the proposals of the Member States. In other words, Member States shall choose their commissioner. This is a step back from the Convention draft that stated that the President of the Commission shall choose each Commissioner from a list of three persons put together by each Member State. However, the Constitution draft still gives the Commission President the right to ask any Commissioner, including the Union Foreign Minister, to resign. Article I-27 deﬁnes the role of the Union Foreign Minister. Hes appointed by the Council with the agreement of the President of the Commission and can also be ﬁred by the Council. He will make proposals for foreign, security and defense policies for which he will receive mandates from the European Council. He will be one of the Vice-presidents of the Commission. Article I-28 deﬁnes the Court of Justice, which is composed of the Court of Justice, the High Court and specialized Courts. Both the Court of Justice and the High Court shall have at least one judge per Member State. Article I-32 clariﬁes the different legal acts in an exhaustive way. A European Law shall be binding and of general application. A framework law shall have binding results but the implementation is left to national authorities. A European regulation is a non-legislative act for implementation of legislation and may be binding in its entirety or only in its results. A decision is a non-legislative act

In Praise of the European Constitution

39

that is binding but only to whom it is addressed. Recommendations and opinions have no binding force. The general legislative procedure, deﬁned in article I-33 is the following: the Commission proposes the European Parliament votes by simple majority and the Council by qualiﬁed majority. Article I-39 identiﬁes procedures for foreign policy. Proposals can come from a Member State or the Union Foreign Minister. Decisions will generally be taken by unanimity in the Council and implemented by the Union Foreign Minister. Laws and framework laws are excluded. Here also, the Council can decide to switch to qualiﬁed majority. Article I-40 deﬁnes procedures for defense policy. Only national capabilities can be used, and unanimity is required for decisions. There is a clause of obligatory aid and assistance if a Member State is the victim of an aggression. For Justice and Home Affairs, article I-41, Member States have the right of initiative. Article I-42 contains a solidarity clause in case of a terrorist attack or a natural disaster. Military resources made available by Member States may be used in such cases. Article I-43 deﬁnes the rules for enhanced cooperation. One-third of Member States must participate in an initiative with the authorization of the European Council. Only these participating states vote on decisions within an enhanced cooperation. The usual qualiﬁed majority rules apply except that supermajority is required if the proposal does not come from the Commission or the Union Foreign Minister. Decisions under enhanced cooperation are not part of the acquis communautaire. Articles I-52 to -55 are related to the budget and state that the budget shall be balanced. The Union’s budget is to be ﬁnanced from its own resources. A law, to be approved by all Member States, shall limit these resources. There will be a multi-annual framework and annual ceilings for each expenditure category. The annual budget will have to be approved by the European Parliament and the Council of Ministers (here also a unanimity decision may allow the Council to decide by qualiﬁed majority). Article I-58 deﬁnes conditions for suspension of membership rights. The Council must ﬁrst decide by unanimity that a Member State is seriously breaching the values of the Union. The decision to suspend a Member State of voting rights is by qualiﬁed majority and so is the decision to revoke a suspension. The European Parliament must approve of this decision by a majority of two-thirds. A suspended member continues to be bound by its obligations to the Union. Article I-59 deﬁnes the right to exit the EU. Part II includes the Charter of fundamental rights of the EU, and Part III includes a lengthy catalogue of the existing competences of the EU. It is mainly a codiﬁcation of the status quo, but articles 166–168 of Part III mention an integrated management of borders, a common asylum policy and a common immigration policy.

40

Ge´rard Roland

Title VI of Part III elaborates on the functioning of the EU. Article III-232 establishes that elections to the European Parliament should follow a uniform procedure. Following article III-234, the Parliament may request by absolute majority an initiative from the Commission. Article III-243 sets the rules for a motion of censure of the Commission as a whole. It requires a two-third majority of the Parliament representing a majority of members. There is no censure of individual commissioners by the Parliament. The appointment of judges to the European Court of Justice will be slightly less politicized than it is currently. The appointment of judges will be staggered with a partial replacement of judges every 3 years. Judges will be appointed for a period of 6 years renewable. Governments will still appoint the judges but a panel of seven experts, six chosen by the president of the Court and one by the European Parliament, will review their suitability (Article III-262). The proﬁle of the panel members would be former members of the Court of Justice, members of national supreme courts and recognized experts. Decisions by the High Court may be reviewed by the European Court of Justice (III-263). The latter will review the legality of EU legal acts and will have the power to interpret the Constitution (III-274). Article III-289 relates to the European Central Bank (ECB). The Governing Council of the ECB will be composed of the governors of Central banks of Euroland (countries that have adopted the Euro) together with the six members of the ECB. The Constitution, however, does not stipulate voting rules within the governing council. 3. Evaluating the constitution The ﬁrst question I want to ask is whether the Constitution is s imply a codiﬁcation, and obviously a welcome simpliﬁcation, of existing Treaties or does it go further than that and represent an improvement over the current functioning of EU institutions? Taking the status quo as the benchmark is a more efﬁcient way of evaluating the Constitution as compared to any idealized benchmark. In fact, we do not have any solid theories that tell us how to compare and thus rank the suitability of constitutions for Europe. A comparison with the status quo is easier and more convincing to carry out than doing the same with a dubious ‘‘optimal’’ constitution. 3.1. A cleaned up intergovernmental institution or a parliamentary confederation? A ﬁrst careful read of the Constitution gives the impression that it is nothing more than a welcome codiﬁcation of the existing consensus at the European level. Though it simpliﬁes the multiple Treaties in place currently, it does not really propose any drastic changes. There is also no real radical change in the catalogue of competences of the Union and a careful status quo has been maintained. There is no question of eliminating the common agricultural policy

In Praise of the European Constitution

41

or of any sweeping step forward on defense. There has not been any change regarding the unanimity rule in controversial areas either. This should however not come as a big surprise. The process by which Member States agree to cease partial or total sovereignty to the European Union has been a gradual process that has always required unanimity. There is currently no unanimity on tax harmonization or defense and the Convention was not going to create it either. And, a bolder proposal may have backﬁred all likelihood. It can thus be seen that the Constitution provides for a status quo on competences and on areas where unanimity is required. However, the important, and very interesting decision, is the adoption of procedures that make it possible to change both the competences of the Union and the voting procedure. The Constitution rightly recommends unanimity for such changes and this makes up an adequate protection of national sovereignty. However, without these ﬂexibility clauses, any minor change to the EU competences or to the voting rules would have needed a lengthy procedure of Constitutional change, required a Convention to be established and a lengthy process of ratiﬁcation on a country by country basis. This would also involve an extremely high transaction cost. Within a bit more than 10 years, the EU has gone through three new Treaties (Maastricht in 1992, Amsterdam in 1997 and Nice in 2000). Each of these Treaty changes has required extensive preparations. Problems such as the Danish rejection of the Maastricht Treaty and the Irish rejection of the Nice Treaty were resolved through ratiﬁcation. . Moreover, these Treaties, especially Amsterdam and Nice, were more of an improvised patchwork and the result of nightly marathons between heads of state rather than being the products of a serene and consistent elaboration. It is clear that with a EU of 25 countries, these transaction costs would have further increased in an exponential way and possibly jeopardized all future options for further European integration. Not only has the Convention brought great improvements in terms of depth and consistency of its draft, but it has also managed to reduce future transaction costs within the framework of the same Constitution. The Convention has achieved this without sacriﬁcing national sovereignty. The changes brought by the Council to the Convention draft (giving one national parliament a veto on measures decided under the ‘‘deepening’’ clause plus the exclusion of defense from the clause) further protect national sovereignty. Nevertheless, they still keep the transaction costs low compared to the status quo. Given the evolutionary nature of European integration, this ﬂexibility of the Constitution is a key element. If the Constitution is adopted, transaction costs for further integration would be drastically reduced. Given the protection provided by the unanimity requirement, there is no good argument for maintaining these unnecessarily high transaction costs. If anything, I would claim that this is one of the most important achievements of the Constitution: it has provided the mechanisms for further deepening of European competences while protecting national sovereignty.

42

Ge´rard Roland

The provisions in the Constitution on enhanced cooperation further strengthen this evaluation. They do make it much easier for a subset of countries to initiate joint competences. However, the Council still needs to approve such initiatives by a qualiﬁed majority. The Constitution also rightly stresses that enhanced cooperation cannot threaten or weaken the single market. Enhanced cooperation, according to the Constitution, shall be done with the existing institutions. Facilitated enhanced cooperation is, in fact, likely to affect the integration process. In effect, it allows a subset of countries to experiment with a new dimension of integration while allowing other countries not to participate, but also denies the latter a veto power over the experiment. The inclusion of enhanced cooperation is probably a constitutional innovation. I am not aware of similar clauses in democratic constitutions. We will now review other aspects related to decision-making within the EU. The ﬁrst issue concerns the European executive. The most important one is obviously the question of the head of the Executive. Initially, both Chirac and Blair, then backed by former prime minister of Spain Jose Manuel Aznar, made a proposal focusing mainly on replacing the 6-month rotating presidency of the Council with a president elected from within itself by qualiﬁed majority for a period of at least two years, and which could be renewed. This proposal was criticized on several grounds. It was rejected by the smaller countries that thought they would lose their turn to chair the European Council, as they feared that an elected president would be from a bigger country. The proposal was also criticized for strengthening the power of the Council at the expense of the Commission. It was argued that this could weaken European institutions as the Council represents countries, and thus mainly national interests, whereas the Commission represents pan-European interests. It was further argued that strengthening the Council at the expense of the Commission could weaken European integration altogether. The alternative favored by European federalists and representatives of the smaller countries was to strengthen the Commission by making it more democratically legitimate. The German government was in favor of such an alternative and argued in favor of the election of the president of the Commission by the European Parliament after the elections to the latter. A French-German compromise, the so-called Chirac-Schroeder compromise, was proposed in January 2003 by the two heads of state. It crudely combined the two proposals: a Council president elected by the European heads of state and a president of the Commission elected by the European Parliament. The proposal was unanimously criticized in the Convention and throughout Europe for putting forward a two-headed executive. This was seen as a setting for sharp and unnecessary conﬂicts between the Commission and the Council that would most probably create dysfunction and discredit for the European institutions. The fact that both presidents would claim executive responsibility would not be the only reason for a potential conﬂict. It would be further heightened by the likelihood of both the presidents coming from different ideologies. Hix Noury and Roland. (2004) have shown that because elections to the European Parliament are the

In Praise of the European Constitution

43

equivalent of US midterm elections, this usually leads to the Parliament being right wing with the Council leaning towards the left and vice versa. In other words, under the Chirac-Schroeder proposal, the president of the Council and of the Commission would most likely be of opposing ideologies, which would further increase their rivalry. German Foreign Minister Joschka Fischer proposed a smart way to ﬁx the problem by proposing the ‘‘double hat’’ idea, whereby the president of the Commission would chair the meetings of the European Council.3 This, however, made the election process for such a president ambiguous, as it is not clear whether the person would be elected by the Council or by the European Parliament. I would argue that the Convention had found a satisfactory solution to this problem. Indeed, the main danger of executive duplication has been prevented. The Council president’s main role is to chair its meetings. The president will also have a role of external representation, but any mention of executive responsibility has been dropped. The president of the Commission will retain powers and receive additional legitimacy from being elected by the European Parliament. I think this is a good and efﬁcient compromise for several reasons. First, the role of the president of the Council will be closer to that of the German president who is the head of state and has functions of representation but no real power.4 Second, abolishing the rotating presidency will lead to better focus and continuity in European Council meetings. Third, nothing prevents the double hat solution, as the same person can be chosen for both jobs. This practice may emerge either from the Constitution or evolve otherwise. In both cases, there is not much of a ﬂipside. A second issue concerns the selection of the president of the Commission, who is the most powerful person in the EU system. It is no secret that heads of government in larger countries like France and the UK had been reluctant to give more democratic legitimacy to the president of the Commission because this would give the latter more power. Strong opposition was thus voiced in some quarters, particularly to the idea of the Commission’s president being elected by

3

This should not be confused with the other ‘‘double hat’’ idea of having one ‘‘Foreign Minister’’ of the European Union holding de facto the post of Commissioner for external Affairs, currently held by Chris Patten, and that of the High Representative for Common Foreign and Security Policy, currently held by Javier Solana, that was effectively adopted by the Convention. 4 The mention of external representation may give the impression that the role of the president of the Council will be analogous to that of the French president. However, this is only a notion. The French president has powers to dissolve the parliament and call for new elections. The president of the Council will not have these powers. Moreover, the French president has strict prerogatives in foreign and defense policy, whereas, the president of the Council will only have external representation responsibilities. Second, the French president is less powerful than was initially thought when the Constitution of the Fifth republic was designed. Whenever there is ‘‘cohabitation’’ due to the majority in parliament and the government being from a different ideology than the president, the latter has little power.

44

Ge´rard Roland

the majority emerging from elections to the European Parliament. The wording in the Constitution is quite clearly ambiguous on this front, because on the one hand, it states the European Council will propose the president of the Commission, but on the other hand, it says a majority in the Parliament must approve the proposal. This sounds like a continuation of the current status quo, where the European Council de facto chooses the president of the Commission. The past few presidents of the Commission like Santer and Prodi had to be approved the European Parliament before assuming ofﬁce. However, the Constitution also includes the words ‘‘Taking into account the results of elections to the European Parliament’’ in mentioning the proposal powers of the Council. A ﬁrst interpretation is that this does not hold much ground and that the Council will propose whomever it wants as Commission president, as is the case currently. A second interpretation is that the Council will act as heads of state in parliamentary systems that choose a formateur, who is usually from the party that has won the elections. This would mean that the president of the Commission is mainly determined by the majority that emerges from the elections to the European Parliament and not by the European Council. As can be seen, these are quite different interpretations. The wording of the Constitution thus contains ground for conﬂict. My evaluation is that while the wording of the Constitution seems non binding to the European Council, in practice, the Council’s role will most likely evolve gradually to resemble that of current heads of state in parliamentary systems. The reason is that the need to secure a parliamentary majority makes it possible for a determined majority coalition in the European Parliament to impose its will on the European Council. However, this is made possible by the Constitution and not by how well parties are structured in the European Parliament. In my opinion, it is only in case of disagreements within Euro parties that win the European elections that the European Council can impose a president of the Commission of its choice. Note that the selection of the new president of the Commission, Jose Manuel Barroso, after the approval of the Constitution draft), in 2004, was done by the European Council and that the European Parliament kept a relatively low proﬁle in the process. However, it was successful in blocking Barroso’s initial choice for the composition of the Commission, because of a controversial choice of some commissioners, in particular the Italian commissioner, and a majority that would vote against the proposed Commission was found. Barroso was therefore forced to withdraw and revise the Commission’s composition. This is a clear indication that the European Parliament has matured and can muster strength if necessary. It remains to be seen which of the two bodies, the Council or the European Parliament, will play a more active in the choice of the Commission president in the future. Another very important issue is that of the Union Foreign Minister. Here the solution adopted is indeed that of a ‘‘double hat’’, as an individual will have to

In Praise of the European Constitution

45

combine the responsibilities of the Commissioner for external Affairs and of the High Representative for Common Foreign and Security Policy (CFSP). This is a good solution for two reasons. First, the ‘‘double hat’’ will eliminate possible conﬂicts between the Commission and the Council of CFSP. Second, keeping the Union Foreign Minister in close contact with the Council is very important to ensure the support of Member States for mandates in their area. I want to mention that the ambiguities created in the writing of the Constitution will automatically give a more important role to the European Court of Justice (ECJ). In fact, these ambiguities will create conﬂicting interpretations, and force each side in a conﬂict to request the arbitrage of the Court. The ECJ will thus have considerable power in solving such arbitrations. This leads me to make some remarks about the role of the ECJ. As stated above, the process of selecting the judges would remain the same as it is currently. Member States will propose candidates with the proviso that a panel of experts recognizes their expertise. It remains to be seen how this process will work. Member States will have strong incentives to propose judges who are close to the political establishment of their country so as to have some inﬂuence in the ECJ. I, however, do not think there is a danger of the ECJ not being independent enough. In a EU of 25 countries, no single executive of a country or even of a few big countries will ever be strong enough to signiﬁcantly inﬂuence the Court. The real danger is when Member States choose candidates that are politically well connected but may lack competence. The panel of experts will in all likelihood help screen candidates who are not competent enough, even though the panel will not have the power to choose the best from the pool. In my opinion, there is another reason why the ECJ will have an important role. Despite the streamlining in legislative decision-making provided by the Constitution, legislative decision-making will still obviously be characterized by more hurdles than usual legislative decision-making in parliamentary democracies. This means that in areas where legislation is ambiguous, and where this ambiguity cannot be removed by reﬁning the existing legislation due to legislative hurdles, the role of interpreting legislation will be left to the Court. Overall, while not providing drastic changes to the functioning of the EU, I would argue that the Constitution provides for a signiﬁcant increase in efﬁciency because: (1) There has been a clear simpliﬁcation in the number of legal tools, (2)The transaction costs to modify the competences of the EU have been signiﬁcantly reduced, (3) The role of the European Parliament has been increased, and (4) Co-decision will now be the normal legislative method. This will provide a good balance between pan-European interests represented in the European Parliament and national interests represented in the European Council (Article III-234). Moreover, the Parliament can now request the Commission to make legislative proposals. The president of the Commission will be elected by the European Parliament after elections to the latter, and will thus derive more

46

Ge´rard Roland

legitimacy from that selection. These are fundamental aspects of a parliamentary system that are being introduced. There will also be a Foreign Minister present both in the Commission and the Council, another clear improvement.5 The increased efﬁciency provided by the Constitution will give more power to all three bodies (the Commission, the Parliament and the Council) and will not increase the power of some bodies at the expense of others. This is not a zerosum game but a positive-sum game. Even though one may think that the relative power of the European Council will comparatively shrink more than that of the two other bodies, one should note that the generalization of qualiﬁed majority voting and the lower threshold for achieving a majority strongly increases the decision-making power of the Council, and of the legislative Councils (even though it remains to be seen how often the ‘‘blocking minority’’ instrument will be used). There are undoubtedly quite a number of weak spots in the Constitution. One of the weaker spots in my view is the rule for the choice of the Commissioners. National governments will continue to appoint the Members of the Commission, and I do not see any good reason for that rule, as a balance of national representations will be present in the Commission in any case. The president of the Commission should have the power to choose the team according to the needs. This will lead to a more cohesive team. Right now, the president can only force them to resign. The current rule creates possibilities for unnecessary hurdles in the formation of the Commission.However, the rule could prove to be harmless over time if countries cater to the Commission’s president in the choice of candidates, but I do not see that happening in the medium run. A big question looming behind the whole discussion on the Constitution is whether it is likely to be ratiﬁed. The answer to that question depends to a great deal on how one views the likely outcome in case one or two countries refuse to ratify the Constitution. For example, it is indeed quite likely, though not inevitable, that there will be a negative referendum outcome in the UK and maybe in another country like Denmark. One view is that the whole project would fail if any one country refuses to ratify the Constitution. The other view is that the Constitution will be adopted by the countries ratifying it and that a special solution will have to be found for the countries that did not ratify it, as long as their number is not too large. It is not my intention here to discuss the legal implications of these two possible outcomes. Politically, I think it is unlikely that the Constitution will be blocked if only one or two countries refuse to ratify it. In the Convention draft, it was explicitly stated that if four–ﬁfth of the countries

5

One should, however, not expect revolutionary improvements in the short term in European Foreign Policy. No mandate will be given to the Foreign Minister for foreign policy initiatives for which there is no unanimity. For example, the Constitution would not have changed anything about the European division on the Iraq war in 2003.

In Praise of the European Constitution

47

ratiﬁed the Constitution, the matter would then be referred to the European Council, which indicates that a political solution would have to be found. In the draft approved by the Council, this article has been scrapped and replaced by one referring to ratiﬁcations of amendments to the Constitutional Treaty. In terms of the dynamics of ratiﬁcation, it is better if this is ﬁrst decided in the countries that are in favor of the Constitution, and leave the decisions by the most eurosceptic countries for the end. Non–ratiﬁcation would then be seen more as a possible exit route from the EU rather than as a way to kill the Constitution. Such a dynamic was present in the ratiﬁcation of the US Constitution by the early American states. The Constitution represents a move towards a parliamentary Europe, as the Parliament will elect the president of the Commission and have the power to censure the Commission, two of the most important characteristics of a parliamentary regime. The move towards a parliamentary Europe is likely to signiﬁcantly increase the accountability of the Commission relative to the status quo. Elections for the Parliament will thus become elections for the Commission, just like parliamentary elections are elections for the government in Member States. Hence, elections to the Parliament will gradually cease to be a sum of ‘‘national protest votes’’ as has been the case in the past. In fact, the European electorate will ﬁnd after a few elections that it has the power to determine the political orientation of the coalition. The electorate will also have the power to punish those parties that act irresponsibly while in power. Giving more power to the European Parliament could increase the cohesion of ‘‘Europarties’’ inside the parliament as suggested by the econometric evidence (Noury and Roland, 2002), which shows that voting unity tended to be stronger in votes where the Parliament wielded more power (co-decision as opposed to consultation). Representation of European socio-economic groups, capital, labor, middle class, etc. would thus be better assured, and electoral campaigns would put forward EU-wide issues of interest to the broadest categories of voters. 3.2. Which public goods will the EU deliver with the Constitution? What can we expect from the institutions emerging from the Constitution in terms of providing public goods? In order to answer that question, we take the political economics perspective (Persson and Tabellini, 2000). Instead of looking at the task allocation deﬁned in the Constitution and analyzing it from the viewpoint of the classical theory of public economics and ﬁscal federalism, we look at the institutions for decisionmaking and predict the kind of economic policy equilibria these institutions may produce. In other words, we look at how given political institutions will tend to systematically deviate from a ﬁrst best–allocated optimum. I will take the theory of parliamentary and presidential democracy expressed by Persson et al. (2000), as the starting point.. In a parliamentary democracy, the

48

Ge´rard Roland

government has executive powers and acts as the agenda setter, initiates all major legislations, and drafts the budget. However, its power depends on continued support in the legislature. A vote of conﬁdence can bring down the government between two elections. In a presidential democracy with separation of powers like in theUS,6 the president has full executive powers, but the government’s agenda setting powers are smaller; the president has a veto right, but for domestic policies, the parliament usually has the power to propose. Those who hold executive powers (the president) and agenda-setting powers (Congressional committees) typically keep them throughout the legislature. Thus, presidential and parliamentary regimes apply checks and balances on elected ofﬁcials in very different ways. In a parliamentary regime, a coalition of representatives (the government) has strong and comprehensive powers. But this coalition is subject to the constant threat of losing these powers if parliamentary support is lost. In a presidential regime with separation of powers, in contrast, no single ofﬁce is invested with very comprehensive powers: the presidential and legislative branches are powerful in different and much more limited policy dimensions. But these powers are assigned for the duration of the legislature. These institutional differences have implications for policymaking. The separation of powers of presidential systems implies more checks and balances on elected ofﬁcials. This is likely to limit corruption and abuse of power. But, in a parliamentary system, politicians in the legislative majority have an incentive to collude in order to prolong the life of the government. This collusion could then be exploited at the expense of the voters. Moreover, in a presidential system, ofﬁce bearers are separately and directly accountable to the voters, while accountability is more indirect in parliamentary regimes. This is a strong incentive for parties to ensure that the voters are kept happy in a presidential system. Accountability is thus stronger in a presidential system. Electoral campaigns by the candidates would be held all over Europe, which would encourage candidates to emphasize pan-European issues and contribute to the creation of a genuine European public opinion. However, presidential regimes have a downside as well. In a parliamentary democracy, policy has to be jointly optimal for a majority coalition; otherwise the government will lose the parliament’s support and fall. This leads to ‘‘legislative cohesion’’ with stable majorities between coalition parties (assuming each party cares about a multitude of issues). It also creates an incentive for the majority coalition to spend time and money on broad redistributive programs or general public goods that beneﬁt many voters. In a presidential regime, different ofﬁce bearers are responsible for various dimensions of the policy and are accountable to different constituencies, leaving them with only weak

6

I have not considered presidential regimes like Russia or many of the Latin American countries where the president concentrates large executive powers.

In Praise of the European Constitution

49

institutional incentives to come to an agreement. Narrowly targeted redistribution is the most efﬁcient policy instrument for these ofﬁcials to achieve their goals. Broad redistributive programs and general public goods are considered a waste, as they provide beneﬁts to many more voters than the individual politician cares about. Hence, in a presidential system, redistribution is likely to take the form of narrowly targeted and selective programs or local public goods, with more limited provision of general public goods and broad redistributive programs. Presidential systems are thus generally less likely than parliamentary systems at providing general public goods that would beneﬁt a large number of voters. In summary, the choice between the presidential and parliamentary form of government involves a trade-off between accountability and public goods. The presidential system fares better on accountability (the agency problem between voters and politicians). But, the parliamentary system is better in terms of the provision of public goods and on conﬂict resolution among the voters. In the context of Europe, there are further trade-offs to consider. A big advantage of the presidential model is that the executive would be better able to react swiftly in times of crisis without facing the danger of a government crisis or inefﬁcient wars of attrition within the executive (see Alesina and Drazen, 1991). Checks of the executive by the Council and the Parliament also ensure representation indirectly via veto powers of these bodies. This is a very important advantage in the context of the challenges facing Europe. The big challenges in the post 9/11 world will be internal and external security. A presidential model is better equipped to face these challenges than a parliamentary model, especially one with coalition governments. The latter argument is especially strong in the context of the EU. Even if the commission has increased powers, the constraint of having more or less one commissioner per country (less than one on average with a system of rotation) would be difﬁcult to escape in the absence of direct elections of the European executive, as it would have negative effects. The commission would be a very ‘‘proportional’’ government. Even more importantly, given the heterogeneity of voter preferences across countries, any coalition in the commission would probably require two, three or four parties. Smaller parties would carry considerable holdup power within the Commission, as would country representatives who threaten to resign. Such threats can be effective because they jeopardize the survival of the incumbent coalition and lead to a government crisis similar to those observed in parliamentary governments with large coalition governments in the past (the French fourth republic, the Italian ﬁrst republic and post-war Belgium). One commissioner from a small party in a small country could thus hold up the Commission. A resignation (when the Commission does not give in) would necessarily lead to multiple cabinet reshufﬂes in order to maintain the fragile balances, and would undermine any executive powers the Commission may be given over areas such as foreign and internal policy, where the ability to respond swiftly in crisis situations is critical.

50

Ge´rard Roland

Given these trade-offs, in Berglo¨f et al. (2003), we argued that a presidential model might be better suited for the European Union because of its advantages in terms of executive swiftness and the disadvantages of coalition governments. A presidential regime with a directly elected president would also be better in terms of accountability. It should be noted however that the likelihood of a crisis in a coalition government in a parliamentary system depends on the rules for no conﬁdence voting to a certain extent. Here, research tends to show that the German style ‘‘constructive vote of non conﬁdence’’ is more desirable (Diermeier et al., 2003). This rule speciﬁes that a government can only be brought down if there is an alternative majority coalition available that can govern. This has two advantages. First, this avoids protracted periods of government crisis (and lengthy negotiations for government formation with caretaker governments), which are often observed in countries with coalition governments. Second, it makes it tougher to bring down a government as agreeing on an alternative majority is often very difﬁcult. Germany has had few government crises thanks to this mechanism. Belgium, traditionally known for the short duration of its governments, has had no government crisis since this rule was introduced in the early nineties. The Constitution does not include a constructive vote of conﬁdence but the rule for a motion of censure is a two third majority. I think this considerably weakens the danger of a coalition crisis as I described above. In the long run, it makes it possible for the president of the Commission to use his executive powers swiftly in case of a security crisis. On the other hand, this also somehow weakens the incentive for legislative cohesion. On the whole, I would nevertheless say that a ﬁne balance has been struck by the Convention. Even though I would have preferred a presidential solution (for better accountability and executive efﬁciency), the advantage of a parliamentary system in terms of public good provision should nevertheless be considered good. In the European context, this means progress on foreign policy, immigration policy and defense. European legislation is less likely to resemble pork barrel politics as practiced in the US. From that point of view, an increase in the powers of the European Parliament should lead to a decrease in lobbying by European institutions. In budgetary matters, amendments in the European Parliament tend to favor general rather than particularistic public goods (Noury and Roland, 2002). A lesson I draw from this is the importance of traditions in institution building. The political tradition in Europe is deﬁnitely one of parliamentary democracy. However, there is no presidential democracy in Western Europe. In the debates of the Convention, advocates of the presidential model were always fewer in number than those in favor of the parliamentary model. The conversations I had with delegates in the Convention showed a clear risk aversion to try out institutions with which they were less familiar, and a strong conﬁdence in institutions that have worked relatively well in the national context. This is a general lesson, I believe, for the political economy of reforms. It seems easier to

In Praise of the European Constitution

51

change existing institutions, with which people are familiar, towards a direction that is desirable from the viewpoint of the objectives of reform, rather than to import institutions from abroad, let alone institutions that exist only on paper. An important test of the Constitution will be to what extent the EU will be able to formulate policies for growth and answer the challenges put forward in the Sapir report. Europe’s growth is lagging behind the US and important initiatives are urgently needed not only in the areas of research, high technology and infrastructure, but also in terms of reforms that will enhance factor mobility within Europe. My prediction is that the Constitution will better equip Europe to adopt such policies than the current institutions. 4. Conclusion To conclude, I think the European Constitution has a very good chance to pass the test of time. First, it will not only strike a nice balance between pan-European interests and the national interest of Member States but it will also signiﬁcantly increase decision-making efﬁciency. This will happen through more reasonable voting thresholds, an increase in the transparency of decision-making by the simpliﬁcation of instruments and procedures, and improved accountability by linking the selection of the president of the Commission to the results of elections in the European Parliament. While one has observed a status quo in the allocation of competences, the Constitution considerably simpliﬁes the rules for such changes in the future. Moreover, the rules for enhanced cooperation represent a constitutional innovation that can further add to this ﬂexibility. European integration is a very gradual and evolutionary process. For now, defense remains a national prerogative and the Constitution does not make much progress in enhancing European defense. However, this is likely to change in the years to come. The same can be said for other areas. Similarly, pressures to eliminate CAP expenditures will continue to mount and it may fade away in the not too distant future. More modestly, the number of areas over which there is unanimity compared to the number of areas for which there is qualiﬁed majority voting in the Council is likely to vary continuously over time. The Constitution allows changes in the allocation of competences to be made without going through extremely costly Treaty changes and ratiﬁcation procedures. European integration can thus continue to proceed in an evolutionary way. The direction in which EU institutions are going is that of a parliamentary model where the executive emanates from a majority in the legislature. This is also the logic behind the election of the president of the Commission by the European Parliament. The parliamentary model is not as good as the presidential model in relation to the direct election of the president in terms of accountability and also in terms of executive effectiveness. Coalition governments are even less effective and prone to instability. They are however better for reducing pork barrel politics and providing general public goods. However, the two-thirds rule for a motion of censure should reduce potential instability. This is a good thing even

52

Ge´rard Roland

though it reduces the accountability. Overall, the choice of the parliamentary model shows the stickiness of traditions in the choice of institutions. I have noted that in several crucial areas (the presidency of the Council, the selection of the president of the Commission) the Constitution contains ambiguous formulations that could be the cause for conﬂicts in the future. However, this ambiguity is itself the result of political constraints. Ambiguous formulations appear as compromises between opposing views, and delay conﬂicts to a future date to ensure that an agreement on the Constitution is reached today. However, such ambiguity is difﬁcult to avoid. It though highlights the fact that Constitutions appear less powerful than institutionalists like me may think. How the Constitution is interpreted will depend on the evolution of mindsets of politicians and voters in various countries. This shows that political institutions are not enough and that the evolution of values and beliefs plays a fundamental underlying role in the interpretation of the Constitution. Institutions cannot fully inﬂuence values and beliefs. Time will tell how the Constitution will be interpreted and used. Acknowledgement I would like to thank Lambros Pechlivanos, Helge Berger and Thomas Moutos and an anonymous referee for comments as well as all participants at the Delphi conference on ‘‘The New EU’’ in June 2004. References Alesina, A. and A. Drazen (1991), ‘‘Why are stabilizations delayed?’’, American Economic Review, Vol. 81(5), pp. 1170–1188. Baldwin, R., E. Berglo¨f, F. Giavazzi and M. Widgre´n (2001), ‘‘Nice try: should the Treaty of Nice be ratiﬁed?’’, in: R.E. Baldwin, editor, Monitoring European Integration, 11, London: CEPR series. Berglo¨f, E., B. Eichengreen, G. Roland, G. Tabellini and C. Wyplosz (2003), ‘‘Built to last. A political architecture for Europe’’, 80 p. in: Monitoring European Integration, n112, CEPR Series, London: CEPR. Diermeier, D., H. Eraslan and A. Merlo (2003), ‘‘A structural model of government formation’’, Econometrica, Vol. 71(January), pp. 27–70. Hix, S., A. Noury and G. Roland (2004), ‘‘A ‘normal’ parliament? Party cohesion and competition in the European Parliament, 1979-2001’’, British Journal of Political Science, Vol. 35(2), pp. 209–234. Noury, A. and G. Roland (2002), ‘‘More power to the European Parliament?’’, Economic Policy, Vol. 17(35), pp. 279–319 London: CEPR. Persson, T., G. Roland and G. Tabellini (2000), ‘‘Comparative politics and public ﬁnance’’, Journal of Political Economy, Vol. 108, pp. 1121–1161. Persson, T. and G. Tabellini (2000), Political Economics - Explaining Economic Policy, Cambridge, MA: MIT Press.

Comment Lambros Pechlivanos

Ge´rard Roland offers an analysis of the proposed Constitutional Treaty. His account of the main innovations of the draft delivered by the president of the European Convention to the European Council meeting in Thessaloniki, Greece, in June 2003, and their repercussions on the efﬁciency of European Union governance is both lucid and insightful. By employing the analytical insights of political economics literature, he evaluates the effectiveness of the political system that will emerge in the EU in terms of the economic policies that are likely to be enacted after the adoption of the Constitution. The author is deﬁnitely not shy and clearly expresses his preferences. He is in favor of a more federal-like, presidential system for Europe. Based on this premise, he evaluates the new draft favorably although he recognizes the draft undertook very reluctant steps in this direction. The built-in ﬂexibility/ambiguity of the draft permits him to be optimistic, as it allows for further changes when the political environment is mature enough for more drastic reform proposals. In this discussion, I will try to use the analytical tools of the incomplete contracts literature to draw some insights on the efﬁciency of some aspects of the draft. A rather fruitful analytical paradigm to analyze constitutions is the theory of incomplete contracts. In this theory, the constitution of an organization is supposed to be written by a benevolent principal, the ‘‘founding father’’, whose intention is to maximize a well-speciﬁed welfare function, given that subsequent decisions are going to be taken in this organization by self-interested agents. This approach has been successful in analyzing the properties of corporate constitutions, where the role of the ‘‘founding father’’ can be attributed to the entrepreneur who starts a ﬁrm as a private company, before going public, and whose goal is to maximize the ﬁrm’s value by allocating decision rights to various claims that the ﬁrm issues. Clearly, things are not as straightforward and easy when one considers political constitutions, and in particular, the European Constitution. Here, there is no founding father to start with – only various CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79010-1

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

54

Lambros Pechlivanos

interested stakeholders –, and moreover there is no agreement over which is the appropriate welfare function that should be maximized. To circumvent this problem a ‘‘veil of ignorance’’ condition can be employed. If the various participants do not know on which side of the distribution of power they will end up, they would be interested in devising such rules of decision making by which they will be constrained when they have to make decisions in the future, so that they will be protected from expropriation, while at the same time efﬁcient outcomed can be reached. Under this condition, participants transcend their vested interests knowing that today’s established positions may not hold much ground in the future. But, again this condition is used to simplify things. Nonetheless, one can try to ﬁnd conditions under which the ‘‘veil of ignorance’’ is more palatable. The ﬁrst of such conditions is the degree of heterogeneity among participants, and the second is the multidimensionality of the policy space over which the established decision rules will be applied in an anonymous fashion. Having presented the analytical framework I have in mind, I will brieﬂy address some issues analyzed by Roland. The ﬁrst one has to do with the establishment of the European Convention per se. As was stressed by the author, the fact that the new draft was not written by an intergovernmental conference, but by an ad hoc convention with wider representation, should make us more conﬁdent in the resulting draft. Members were not representing their countries in a strict sense, and consequently, they may have been less constrained by their country’s vested interests. One of the major innovations in the new draft are the ﬂexibility clauses introduced in the constitution, particularly the one presented in Article I-24 under which the European Council may decide by unanimity that an issue for which the constitution speciﬁes that the decision be based on the unanimity rule can be decided by the qualiﬁed majority rule (after a cooling off period of six months). Essentially, this clause allows for an issue speciﬁc amendment process, which is much easier to implement, as it does not require the usual cumbersome procedures. The author welcomes these ﬂexibility clauses and considers them as one of the main reasons for the draft constitution to be considered a step forward towards more efﬁcient EU governance. Nonetheless, one should keep in mind that ex post efﬁciency may come at a cost. A central result coming from the contract theory, either through a complete contracts full blown mechanism design framework or an incomplete contracts framework, is that there is a trade off between ex ante incentives and ex post efﬁciency. In this case, the question is whether the ability to amend the constitution reduces the incentive to reach a more balanced constitution in the ﬁrst place. The argument is that rules with shorter expected horizon are less likely to be enacted under a ‘‘veil of ignorance’’ condition: Currently, advantaged members may enact those rules that beneﬁt them; knowing they can change the rules if they fear a future reversal in the power distribution. How much stress should one place on this trade off is an open discussion, but a historical anecdote is rather illuminating. When Solon, the founder of the early Athenian Republic

Comment

55

handed in a ‘‘constitution’’, he stated that it is not amendable by anyone but himself, and then he went on self-exile for 10 years. Another point I would like to raise on the effectiveness of the ﬂexibility clause is whether it can really bring about the intended ﬂexibility in practice. Here I would like to stress that the clause will be applied on an issue-by-issue basis, and hence there appears no logrolling-like scheme to facilitate a compromise. In bottom, the question is whether there is an incentive for a member that intends to veto a decision not to veto the enactment of the ﬂexibility clause. Of course, when discussing this, one should take into account the theatrics of the political decision process, in the sense that it may be easier for politicians to sell to their constituency an unpopular decision if they are indirectly responsible for it – and, moreover, via an obscure technical decision. Finally, I would like to address another topic discussed by the author. The topic being whether EU governance will look more like a presidential or a parliamentary democracy. Roland, based on the political economics literature, argues that although he personally would have preferred a more clear cut presidential system as it promotes accountability, the EU system being closer to a parliamentary system will be able to promote the provision of general public goods (something that is important for the legitimization of the EU authorities). Here, I would like to point out that the insights one gets about the beneﬁts of the parliamentary system depend on the fact that multi-issue parties are assumed to be in power. If members in the majority coalition are one-issue parties (which only have to cater to one constituency), then the attention will be shifted from the provision of generalized public goods to targeted ones. An example that ﬁts this case is the current majority coalition between the Likhud party and the ultra-orthodox parties in Israel. The main reason, for which I mention this particular scenario, is that in the context of EU politics, nationally minded parties can be understood as one-issue parties. The extent to which such parties can exercise power as coalition makers is, of course, something that remains to be seen. In conclusion, I would like to stress that my above stated concerns should not be understood as a polemic to the new draft constitution. On the contrary, given the current political state of the European Union, I share the author’s positive view of the draft produced by the European Convention.

This page intentionally left blank

CHAPTER 3

Institutional Aspects of EU Organization: An Economic Analysis Massimo Bordignon

1. Introduction The European Union (EU) is a strange animal. It has similar institutions of other existing federations, but it is not a federal state, not even a somewhat looser version of it (say, a Confederation such as Switzerland). True, the EU has a Central Government (the Commission), a Senate representing the States (the Council), a Parliament representing the People, and even a Constitutional Court. But these European federal institutions perform a very different, and in general, a much more limited role than that of the similar institutions of other consolidated federations. The amount of powers still concentrated in the member states, the quality and the extent of these powers, the small scale of the European budget, the lack of true executive powers by the EU government, the residual role still played by the European Parliament in passing legislation, and more generally, the absence of any true pan-European issue at stake in the day by day politics of the member states makes the EU completely different from any other known federation. For example, to date, the EU does not have a common foreign policy or a common defense policy; the European Parliament cannot propose new legislature, and the European Government is not elected but appointed by the Council.1 On the other hand, the EU is clearly far more

1

More precisely, the President of the European Commission, appointed by the Council, needs a conﬁdence vote by the majority of the European Parliament to be conﬁrmed. However, the acts of the Commission do not need a majority of the Parliament, and the Commission can only be dismissed by the European Parliament with the votes of two-thirds of its members. See Berglof et al. (2003) for further details.

CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79003-4

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

58

Massimo Bordignon

than a simple economic agreement among sovereign states. True, it does share many features with an economic union, such as a common internal market and, for those members who joined the EMU, even a common currency. But the extent of powers transferred to the center in determining, for instance, competition and monetary policy, has no parallel in other existing economic agreements, and makes the EU a far more politically and economically integrated region than, say, NAFTA. These peculiar characteristics should warn scholars from using the knowledge accumulated on other federations to analyze the workings of the EU. One can certainly learn something from these other experiences. But in many cases it seems more advisable to attempt to analyze the European institutions on their own terms, because probably that is the only context in which they gain their meaning.2 Take, for instance, the problem of the optimal attribution of functions to different levels of government, which is a fundamental issue discussed at large in the ﬁscal federalism literature. Economists have since developed a recipe to address this issue (e.g. Oates, 1972). But this recipe seems to be of little use when the ‘‘central government’’ has the peculiar characteristics of the European Commission. Or take the issue of the optimal allocation of taxing powers to subcentral governments. Again, it makes a lot of difference if this problem is discussed in a context where the central government has large taxing powers of its own (as in most existing federations) or in a context where it does not (as in the EU) (Keen, 1998). And the examples could go on. In short, it appears we need an economic analysis, which takes fully into account the peculiar institutional characteristics of the EU. In the following, I am going to present some elements of this analysis. The paper builds upon formal work that I have developed together with other coauthors on speciﬁc issues of European concern. But more than in the results of the single papers, I am mainly interested here in underlining the common methodology used in these papers. We are very far from having a satisfactory understanding of the working of the European institutions, but I believe these works may offer some suggestions on how we could proceed. Focussing ﬁrst on a detailed analysis of the EU institutions for the problem at hand, they help clarifying the issues for the subsequent theoretical research, putting in evidence unexpected policy implications and offering explanations for otherwise hard to explain facts. I will be discussing two issues of relevance for the EU here; the introduction of formal rules to form sub-unions inside the EU (known as Enhanced

2

Indeed, as a referee suggested, this can be taken as an example of the ‘‘theory of second best’’ as applied to institutional design. One always needs to look at the details of the institutional environment, since a speciﬁc rule which may appear meaningless at ﬁrst glance may prove to be a beneﬁcial response to another rule or to an ‘‘institutional incompleteness’’ once the entire picture is taken into account.

Institutional Aspects of EU Organization

59

Cooperation Agreements in the jargon of the European Treaties) and, more brieﬂy as space is limited, the problem of the optimal allocation of functions to the Union and the member states on the supply side of the economy, where the ‘‘open method of coordination’’ now prevails. To emphasize the common methodology used in these more formal works,3 I will present their main insights following the same structure. Thus, I will start by explaining the importance of the problem in the European context and why this problem presents particular features inside the European Union; I will then move to illustrate brieﬂy the model used to address the issues and the main results obtained; and ﬁnally, I will discuss the main insights and policy implications which follow from the analysis. The rest of the paper is organized as follows. Section 2 discusses the problem of sub-unions formation and governance inside the EU. Section 3 analyses the problem of the allocation of functions. Section 4 concludes by brieﬂy suggesting further avenues for research. 2. On enhanced cooperation 2.1. The problem At bird’s-eye, the EU as a working federation clearly has two main problems. First, it is becoming increasingly heterogeneous. The EU was internally differentiated to start with, but the addition of 10 new countries in 2004 has transformed the EU in a region of very heterogeneous countries, which are differentiated by size, population, history, institutions, language, and levels of economic development. The difference in the GDP per capita from the richest to the poorest country skyrocketed after the extension; and the newly included countries obviously require very different policies than the original members, whose market economies are more mature. This means that it is becoming more and more difficult to find common policies that could benefit all, or most of, the members of the EU. Second, the EU’s decision- making is becoming increasingly cumbersome. Unanimously agreeing to efﬁcient decision-making rules for a Union with 25 different members has proved to be extremely difﬁcult, as member countries have resisted surrendering sovereignty to the European institutions. The current version of the Founding Treaty, the Treaty of Nice, which was supposed to streamline the EU institutions to enable them to deliver even with 25 members, has been a disaster on these grounds (Baldwin et al., 2001), making it more difﬁcult for the Council to arrive at decisions even in those areas formally under qualiﬁed majority rule. The recently approved (by the European Council) European Constitution (June 2004), with its modiﬁcation of decision rules for the Council starting with 2009, still has to be ratiﬁed in many countries to become effective, and in any case it is yet to prove that it can work. This

3

See Bordignon and Brusco (2003) and Bordignon et al. ( 2003).

60

Massimo Bordignon

means that not only is it becoming more difficult to find common beneficial policies, but it is also becoming tougher to have them implemented by the EU. In addressing these problems, the EU legislators have possibly found a working solution, one that has no parallel in other existing real world federations. Faced with the dilemma between proposing the integration of a given function at the European level at the risk of hurting some member countries and ﬁnding it therefore impossible to have the proposal approved by the Council, or giving up the integration policies which could instead beneﬁt a large subset of member countries altogether, the EU legislators have opted for an intermediate solution. Namely, the possibility for only subsets of member countries to go on with integration on some particular issue, following ex ante agreed upon decision and governance rules; or, to put it differently, the possibility for member countries to legally form sub-unions inside the larger EU. These sub-unions have been termed ‘‘Enhanced Cooperation Agreements’’ in the evocative language of the Treaty of Nice, or ECAs for short. It is true that the EU has always had to deal with sub-unions of member countries in its history. For example, the Shengen Treaty, initially approved by only a subset of members and then extended to the others, can well be considered as an ante-litteram example of an ECA. In a sense, even the European Monetary Union, which currently covers only 12 countries, or less than half of the current members of the EU, can be thought of as an ECA. There are some important differences, however. First, these previous examples of ECAs were the result of intergovernmental agreements, largely reached outside the formal working of the EU institutions. This is not the case with the arrangements of the current ECAs. As we will shortly see, there are precise rules deﬁned in the Treaty and therefore ex ante approved by all members, according to which subsets of members, if they so wish , can propose and then, if approved, form a sub-union on some issue. Importantly, this implies that at least in principle, no subset of EU member countries can now form an ECA outside these rules, and vice versa, no excluded country can oppose the formation of an ECA once it is approved by the Commission and the Council following the agreed upon rules. Of course, it is yet to be seen if these rules are really binding, as they refer to what are still separate and sovereign countries. But the fact that the countries themselves have signed the Treaties introducing these rules gives them a legal and political force, which no sovereign country can safely ignore. For example, an agreement between a subset of member countries in some of the matters covered by the Union, reached outside these formal proceedings, could well result in a legal case being put forward by the Commission or by the excluded countries against the European Court of Justice. Another important difference is that in these previous examples of ECAs, sub-unions have always been thought of, and presented as, temporarily deviations from a common unifying trend. In the case of EMU, for example, some countries have been allowed ‘‘to opt out’’ from the monetary union for the time being, but the understanding, or at least the rhetoric of the agreement, was that

Institutional Aspects of EU Organization

61

eventually all EU members would join the sub-union. However, with the current arrangements, this will not be the case. The Union may well wish this to happen, and the wordings of the Treaty makes it clear that this is the intention of the legislators, but there is no obligation for any member country to join a subunion, either in the present or in the future, if it does not wish to do so. Hence, the new rules open the way to an EU largely segmented internally, with different sub-groups of members, possibly partly overlapping, that coordinate their policy on different issues; a situation which could go on indeﬁnitely. 2.2. What are ECAs? Before continuing, it is important to be more speciﬁc about the rules for forming an ECA in the current European Union Treaty (the Treaty of Nice) as this is essential to understand the analysis which follows. The current procedure to form an ECA was ﬁrst introduced in the Amsterdam Treaty, signed in 1997. The Treaty of Nice (drafted in 2000, ﬁnally ratiﬁed by all countries in 2003) basically reproduces all these rules, but with an important difference. Whereas under the Amsterdam Treaty, any country could block an ECA proposed by some other group of countries by invoking reasons of paramount ‘‘national interest’’, this is no longer possible under the Treaty of Nice. That is, the latter Treaty has effectively revoked the veto power that the previous Treaty still guaranteed to each member country, thus making ECAs much easier to implement. On these grounds, the Convention’s proposal (whose amended draft was approved by the European Council in June 2004) changes very little, except that it tends to make ECAs even easier, and to further reduce the difference in the rules for the formation of ECAs between the different ‘‘pillars’’ in which the functions of the EU are traditionally divided. I will indicate below when the proposals of the Convention could make a difference. The rules described beneath refer to the most important group of functions in the EU, the ﬁrst pillar (Economics and the Common Market); the rules for the other two pillars are only slightly different, and are moreover in the process to converge to the ﬁrst pillar rules following the Convention’s amendments. According to the Treaty of Nice (2002), then, to form an ECA: 1. At least eight members of the EU (a third of the members, according to the Convention draft) must declare their willingness to form an ECA in one of the matters covered by the ﬁrst pillar of the Treaty; 2. The European Commission evaluates this proposal and assesses its compatibility with the Common Market, the fundamental freedoms of the Treaty, and the aquis communitaire. Following this assessment, the Commission then decides either to kill the sub-union, or to approve it. In the latter case, the Commission drafts a legislative proposal concerning the ECA, which is then presented to the Council. 3. The Council decides, via Qualiﬁed Majority Voting, whether to accept the draft of the Commission or to reject it.

62

Massimo Bordignon

4. Once approved, an ECA is binding only for those countries that accept to be part of it. The decision rules inside the sub-union are the same as for the Union at large on the same subject. For instance, if the ECA concerns direct taxation, decisions inside the sub-union are taken according to unanimity rule. In other cases, decisions are taken according to qualiﬁed majority rules, with each country maintaining the same number of votes it has in the Council at large. 5. The members of the Council, not belonging to the sub-union, may participate in the discussions concerning the sub-union policy, but only the actual members of the ECA have the right to vote. On the contrary, all members of the European Parliament (including, therefore, representatives of countries outside the sub-union) vote on the proposals coming from the sub-union, according to the same rules for Parliament rulings prescribed by the Union at large for the subject covered by the ECA. 6. All EU members presently outside the ECA may ask to be included, provided that they ‘‘accept to comply with the decisions taken’’. The European Commission is actually given a role in deciding whether non-ECA countries may be allowed to join (the Commission, for example, may decide to set criteria to be fulﬁlled before such a country is allowed in the sub-union), but the wordings of the Treaty makes it clear that, in general, all EU members who wish to join should be allowed inside the sub-union (the Convention draft further weakens the powers of the Commission in this context, making accession to the ECA almost automatic for all EU countries who want to join). Three things should be noted about this procedure. First, in the current arrangements, a paramount role is given to the European Commission. It is the Commission who decides if it should accept the proposal advanced by the countries in the ﬁrst place, it is the Commission who drafts the proposal to be presented to the Council, and it is the Commission who ﬁnally sets up the criteria for admitting other countries. Second, the countries outside the ECA do not have a direct say in the policy of the ECA. Their interests, if any, in the matter covered by the ECA are represented by the Commission (made up, so far, by representatives of all EU countries, but it is in the process of being reduced in number following the recent constitutional reform: see below), and by the European Parliament. However, given the limited role still played by the latter in drafting European legislature, it is clear that it is mainly the Commission that plays the role of a guarantor for the interests of all EU members, both inside and outside the sub-union. Third, if they feel that they are somehow damaged by the ECA, the non-ECA members can always join the sub-union. But note that if they join a sub-union that has already been formed, they are asked to ‘‘comply with the decisions taken’’. Furthermore, their ability to affect the decisions inside the sub-union depends on the governance rules of the ECA itself, rules that in turn depend on the matter covered by the ECA, a point that we will discuss later in the chapter. Summing up, the political compromise found by the EU legislators with the current ECAs’ rules can perhaps be better synthesized as a ‘‘No Veto – No

Institutional Aspects of EU Organization

63

Exclusion’’ agreement. Countries not interested, or opposing an integration policy on some issue, cannot any longer forbid (a consistent subset of ) the other EU members to go on and harmonize their policies on that issue; on the other hand, no EU country can any longer be excluded from an integrated policy reached by other EU countries. Are these rules the solution to the EU’s problems? 2.3. ECAs: the debate According to some observers, the answer is a clear-cut yes. For instance, at the end of their long and detailed report on the Treaty of Nice, Baldwin et al. (2001) conclude by saying that ‘‘After the Treaty of Nicey ECAs could become the main engine of future integration’’. Other observers are less optimistic. Indeed, ECAs have generated a ﬁerce debate among political scientists, lawyers and politicians. Some observes believe that ECAs represent the only realistic solution for a loose and heterogeneous federation such as the EU. A few others are instead more critical, and fear that ECAs will lead to the formation of a ‘‘twospeed’’ Europe. Others believe that ECAs fall way short of what the EU really needs, that is, effective political decision making to be reached through the extension of majority voting to a larger set of matters (maybe even all of them). However, this debate is largely impressionistic in nature; no formal or detailed analysis of the trade-offs induced by the speciﬁc rules for ECAs formation is offered in this literature. Economists, who are more equipped to offer this kind of analysis, have instead been largely silent on the matter, preferring instead to focus on other European problems. There are some exceptions, however. In an earlier attempt to discuss this problem, Dewatripont et al. (MEI, 1995) brieﬂy discuss the trade-offs induced by ECAs, which they term ‘‘ﬂexible integration’’. On the positive side, they emphasize the advantages of the sub-unions in terms of experimentation and learning about the pros and cons of political integration on a particular issue. In particular, they stress the ‘‘snow balling effect’’ of ECAs, an idea which also surfaces in Baldwin et al. (2001). Sub-unions that prove to be successful for their members would promptly be followed by the other members of the EU, thus leading to (only) beneﬁcial integration. On the ﬂip side, they worry about the possible negative externalities that sub-unions could generate on the EU members outside the ECAs. In particular, they fear about the consequences for the Common Market (the ‘‘Hard Core’’ of European Integration, in their words). Accordingly, they advocate a strong role for the European Commission (as the guardian of the Common Market) as a gate-setter for ECAs formation4. Alesina and Angeloni and Etro, (2003), by using a dynamic (median voter’s) model of

4

Dewatripont et al.’s analysis could be interpreted as an application of the standard theory of customs union to ECAs: to determine whether a sub-union is efﬁcient or not, the benchmark is whether ‘‘trade creation’’ or ‘‘trade diversion’’ type-effects dominate.

64

Massimo Bordignon

union formation, detect instead a tendency in the EU versus excessive centralization (see also Alesina and Grilli, 1993). According to their analysis, the EU tends to remain smaller and with more functions allocated at the central level than should be optimal. They consider different possible remedies, among which they also brieﬂy discuss the use of ECAs.5 This is basically all we have on ECAs. Do they fully capture the trade-offs involved in sub-unions formation in the EU? Without disagreeing with some of the insights offered by the above analyses, in a joint work with Sandro Brusco (Bordignon and Brusco, 2003) we make a different point, one that we believe to be more relevant for the issue at hand. The point is simply that the real world is dynamic and stochastic, and that the case for sub-unions should be assessed in a context, which considers these two important real world characteristics. ‘‘Stochastic’’ here simply means that political contingencies change in a way which cannot be precisely predicted at the present; as an effect, countries that decide not to join a sub-union today, may decide to do it tomorrow – perhaps, as suggested by Dewatripont et al. (MEI, 1995), because the sub-union has proved to be successful in the past. ‘‘Dynamic’’ here simply means that what happens today is going to affect what happens tomorrow. Institutions are like ‘‘state’’ variables, with largely irreversible effects: changes today are going to affect the possible changes tomorrow.6 To state an obvious example: the UK decided to opt out from the EMU in 1998; pending a national referendum, it may now decide to enter into the monetary union agreement. But if the UK joined the EMU today, it would not be the same thing as if it had decided to join the union immediately. Institutions, including Central Banks, have their own hysteresis, one which cannot be reverted easily; in its objectives and internal workings, the European Central Bank nowadays certainly looks much more German (or French or Italian); this would not have been the case had the UK joined the EMU immediately. These are rather obvious observations; but they have far reaching implications. First, contrary to what is suggested by Dewatripont et al. (MEI, 1995), the observations imply that ECAs may damage excluded countries even in the absence of any negative externality, either in the present or in the future. If forming a subunion today on some issue changes the standard of integration tomorrow on the same issue, countries who do not want to join the sub-union today, but who may have accepted integration on that issue tomorrow, may ﬁnd themselves worse off than they would have been if the sub-union had been prohibited to start with. Second, this also suggests that governance rules for sub-unions are far from trivial:

5

See also Widgre´n (2001) for an analysis of voting behaviour inside the ECA, in case majority rule determines the policy in the ECA, and Perotti (2001) for a discussion of a case where policy harmonisation inside the EU would actually be harmful. 6 Our point here, as it is based on lack of commitment at the political level, is obviously strictly linked to the ‘‘incomplete contracts’’ literature; see, for instance, Bolton and Whinston (1993).

Institutional Aspects of EU Organization

65

even in the absence of any negative externality, for example, it may make sense to let countries outside the sub-union retain some decision-making powers on the sub-union policy, as this may help in protecting them from future exploitation. If these arguments are correct, the introduction of ECAs would then give rise to a fundamental trade-off. On the plus side, ECAs would allow countries that wish to form a sub-union today to reap efﬁciency beneﬁts that would otherwise be lost (with the correlated possibilities of experimentation and learning to the advantage of all other countries). On the ﬂip side, ECAs introduce a possible risk of exploitation for the excluded countries, if the latter join the subunion in the future. One can then ask, on positive grounds, if this trade-off can actually help in explaining some of the existing arrangements for ECAs in the EU; and, on normative grounds, one may also be led to question the optimality of these same arrangements. 2.4. The analysis To further clarify the issues involved, it may be useful to refer to the results of a simple model, which is discussed in more detail in Bordignon and Brusco (2003). As a working example for our analysis, we focus on the issue of corporate income tax harmonization at the European level. Of course, our argument on ECA is far more general, but this example illustrates very nicely the trade-off we are discussing here. Differences in legal and accounting rules for corporate taxation across the European countries are well known enough to represent one of the main obstacles for efﬁcient allocation of capital in Europe (e.g. Bond et al., 2000; Ruding Committee, 1992). Corporate tax bases and taxable proﬁts are determined in different ways in the individual EU countries. As a result, computing the net of tax return from an investment abroad might be a nightmare for any EU investor, and this may induce ﬁrms, particularly small ﬁrms, to give up otherwise productive investments in the other EU countries. Years of discussions and even several European Commission proposals for across-the-board harmonization have not gone anywhere. The difference in current practices across European countries is simply too large for all of them to agree to pay the costs of the adoption of a common standard for taxing corporations. Furthermore, the overall beneﬁts – and their distribution across countries – of a harmonization policy are very difﬁcult to assess at the moment. For these reasons, and also because direct taxation legislation can only be approved via unanimity ruling in the Council, all attempts of harmonization have so far failed. However, for historical reasons, differences in accounting standards are lower for subsets of the EU countries than they are for the Union as a whole. It is then quite possible that under the new Treaty of Nice rules, the adoption of a common standard for corporate taxation could become one of the ﬁrst examples of an ECA in the EU. But if this ECA is successful, and companies across Europe start to use this common accounting standard for their investment, any further country that wants to join the ECA would be forced to harmonize at this

66

Massimo Bordignon

standard. Hence, there is a serious risk of future exploitation for those countries that initially decide against joining this ECA. To see these arguments more accurately, consider the following formalization of the above example. Suppose there are only three countries, and for historical reasons two of these countries have accounting and taxing standards for corporations that are closer than those of the third. To make this idea more precise; assume that these historical standards can be represented as points on a line, with, say, country 1 having a standard equal to zero, country 3 having a standard equal to one, and country 2 having a standard in the interval between zero and 1/2, so that there is an meaningful sense in which countries 1 and 2’s standards are ‘‘closer’’ than that of country 3. Although these historical standards can be changed, it would come at a cost, as legislation must be passed, lawyers, accountants and tax ofﬁcials must be trained anew, errors in the transition periods must be corrected, and so on. Again for simplicity, assume that these harmonization costs can be represented by a quadratic cost function, under which one country has to pay a larger and increasing cost to harmonize, the further is the harmonized standard from its historical one. Harmonization of the standards between two or three countries may be beneﬁcial because it may increase international capital mobility. But assume, again, following our previous argument, that there is some uncertainty about the actual beneﬁts of a harmonization policy. For simplicity, suppose that there are only two periods, and that uncertainty disappears as time enfolds, so that in the second period there is perfect knowledge of the advantages from harmonization, 7 while in the ﬁrst period these advantages can only be assessed probabilistically. Let P be the (ﬁrst period), ex ante probability that harmonization may be beneﬁcial for the federation. Our three-country federation then has a difﬁcult decision to make in the ﬁrst period. It may decide to follow a decentralization policy in the ﬁrst period (each country choosing a different standard), and wait until period 2 before deciding if it is worthwhile to go on with further harmonization. Although this would save the countries the costs of a harmonization policy in the ﬁrst period, it would also put them at risk of losing the beneﬁts for harmonization in the same period (the increased capital mobility). Or, it may decide to immediately pay the costs of full centralization (all countries agreeing on the same standard) in the ﬁrst period, at the risk of wasting resources if harmonization is not beneﬁcial. Finally, a third possibility is to build an Enhanced Cooperation Agreement between only countries 1 and 2 in the ﬁrst period (choosing the same standard for countries 1 and 2), so

7

This implies that we do not consider here the advantages of learning which can come out from the ECA; whether the countries form an ECA in the ﬁrst period or not, uncertainty dissolves in the second period. But the model could easily be extended to consider this feature too, by making the degree of uncertainty in the second period dependant on the fact if an enhanced cooperation or centralization policy has taken place in the ﬁrst period. Notice that while this may make the case for ECA more robust on efﬁciency grounds, it would not change the basic trade-off we focus on here.

Institutional Aspects of EU Organization

67

that at least these two countries can enjoy the beneﬁts of increased capital mobility in the ﬁrst period (in the case harmonization proves to be beneﬁcial), leaving to the second period any decision concerning the third country. Intuitively, this last option may dominate the other two; as historical standards of countries 1 and 2 are closer to start with, it is certainly less costly to implement a common standard between these two countries than it is to implement one for the Union as a whole. The catch, however, is what to do with the third country. If harmonization turns out to be beneﬁcial (large returns for increased capital mobility as a result of corporate taxation harmonization) it may be advantageous for all countries to have the third country join the harmonization policy in the second period. But if a common standard has already been chosen between the two closer countries in the ﬁrst period, the second period optimal harmonization standard for the Union as a whole may be further from the historical standard of the third country than it would have been had the ECA not been formed in the ﬁrst period. As a result, the third country may turn out to be exploited in the second period by the ﬁrst period ECA; that is, it may have to pay a larger cost to join the harmonization policy than it would do if the ECA had been prohibited to start with. Intuitively, which of the three policies, decentralization, centralization or ECA, is more desirable on efﬁciency grounds depends on several factors. The exogenous parameters of the problem (how large is P and how close are the historical standards of countries 1 and 2) on the one hand, but also on the governance rules inside the sub-union, and in particular on whether the ‘‘losers’’ from the different policies can be compensated by the ‘‘winners’’. To highlight these features, in Bordignon and Brusco (2003), we consider different possible scenarios. We start with the benchmark case in which a benevolent dictator (the European Commission?), equipped with lump sum transfers and subjected to unanimity voting by all countries takes all decisions; we then introduce a number of more realistic political constraints into the analysis, up to discussing the ‘‘no veto-no exclusion’’ policy of the recent Treaty of Nice. Clearly, if a benevolent dictator is in charge of all decisions and can use lump sum transfers to compensate losers, all he has to do is to choose the harmonization policy which maximizes the sum of expected utilities of the three countries; at no efﬁciency costs, the lump sum transfers would then make sure that each country is better off with respect to the alternatives. Referring to Bordignon and Brusco (2003) for any analytical detail, Figure 1 summarizes the results of the analysis for this benchmark case. In the picture, I measure on the vertical axis the expected utility of the federation (the summation of the three countries’ expected utilities) and on the horizontal axis, the ex ante probability for the harmonization policy to be successful. The different lines8 in the picture

8

The curves representing expected utilities are drawn as linear as an illustration only; in reality, except for the centralization case, they are not all linear functions of P. But they intersect only once, so that they intersect only once, and that the picture correctly captures the main intuition of the results.

68

Massimo Bordignon

Figure 1. Enhanced cooperation, centralization and decentralization

represent the expected utility of the federation for each of the three different policies, dependent on the different values of P. As drawn, all curves are increasing function of P,9 but the expected utility of the federation in the three different cases is very different depending on the value of P. For instance, when P ¼ zero (there is zero probability that the harmonization policy could be beneﬁcial), expected utility under centralization is certainly negative (all countries

9

Expected utility under decentralization increases with P, because with a quadratic cost function, each country prefers to move its standard a bit in the ﬁrst period, even if no harmonization occurs in that period in anticipation of the possible move towards full harmonization in the second period.

Institutional Aspects of EU Organization

69

pay the cost of harmonization to a common standard with no beneﬁts), while, upon normalization, it is zero under decentralization (no country moves from this historical standard, and so no harmonization cost is paid). On the contrary, when P ¼ 1 (harmonization is certainly beneﬁcial), centralization certainly dominates decentralization because in the former, countries are assured to enjoy the beneﬁts from harmonization even in the ﬁrst period. Clearly, if one considers only centralization and decentralization as possible policies, there has to be an intermediate value for P, P*, such that for PoP*, decentralization dominates centralization in the ﬁrst period, while for P>P*, centralization dominates decentralization in the same period. Let us now add enhanced cooperation to the picture. ECA is clearly an intermediate case; lower costs are paid in the ﬁrst period (only two countries harmonize their standards), but lower beneﬁts can also be expected (capital mobility increases only across these two countries in the ﬁrst period). It is especially clear that when P ¼ zero, enhanced cooperation is certainly dominated by decentralization (some costs are however paid in the ﬁrst period, with no resulting beneﬁts), while when P ¼ 1, enhanced cooperation is certainly dominated by centralization (if harmonization is certainly beneﬁcial for all, there are no reasons to limit it only to two countries).10 Whether exists an interval of values of P such that ECA can dominate the other two policies depends on the other parameters of the problem. To illustrate this, in Fig. 1, I have drawn two curves for the expected utility of the federation under the ECA policy, ECA1 and ECA2. Clearly, if the case is as drawn in ECA2, enhanced cooperation is always dominated by either centralization or decentralization, so that the efﬁcient policy for the federation would move from decentralization to centralization as P increases without ever considering forming a sub-union. On the contrary, if the case is as drawn in ECA1, then there is an intermediate set of values for P, (P0 , P00 ), where enhanced cooperation dominates both centralization and decentralization. Quite intuitively, it can be shown that the main factor determining which of the two cases is more likely to occur is the distance between the historical standards of countries 1 and 2. If this distance is very small, then the interval (P0 , P00 ) is certainly not empty, while it may be empty in the opposite case (that is, with country 2’s historical standard close to 1/2). Putting it differently, if the heterogeneity of the Union is much larger than the heterogeneity inside the sub-union (the standards of countries 1 and 2 are much closer than that of country 3), then enhanced cooperation agreement may indeed be a good idea, as an intermediate step towards further integration. Lending support to the institutional debate on ECAs in the EU, this indeed suggests that the efﬁciency case for sub-unions is more robust the more the Union is heterogeneous.

10

Strictly speaking, these inequalities depend on the assumptions made on the technology parameters; see Bordignon and Brusco (2003) for details.

70

Massimo Bordignon

However, these results were obtained in the benchmark case where all decisions are taken by a benevolent dictator with full lump sum transfers at its disposal, hardly a realistic description of the functioning of the EU institutions. Suppose now that lump sum transfers are no longer available. Indeed, there are no truly compensatory transfers across countries at work in the European Union; when one country is hurt by some decision, it is usually compensated by distorting other pieces of legislation, so in general inducing further welfare losses (e.g. Tabellini, 2003). For the time being, let us still assume that countries can commit ex ante to harmonize at the efﬁcient standard in the second period, meaning that the countries forming a sub-union can assure the third country that it will not be exploited in the second period if it joins the ECA. Then, it can be shown that lack of transfers makes the case for ECA unambiguously better on efficiency grounds. The intuition is simple: as decentralization does not require transfers, the expected utility of the federation in this case is unaffected by distorting transfers; centralization and enhanced cooperation may instead both require transfers, but because of the smaller variance of the standards inside the sub-unions, these transfers are certainly smaller for the ECA policy than for centralization, resulting in smaller welfare losses. In terms of Figure 1, the line representing centralization would fall more than the line representing ECA, resulting in a larger interval for (P0 , P00 ).11 That is, not only can ECAs be a good idea if the Union is made by very heterogeneous countries, it can be an even better idea if the Union ﬁnds it difﬁcult or very costly to compensate ‘‘losers’’, a situation which certainly characterizes the present state of the EU. These are all good news for the ECAs’ arrangements. But again, all these results were obtained by assuming that the third country could be guaranteed against the fact that the sub-union could generate future welfare losses. Assume now, on the contrary, that countries can no longer commit on a particular harmonization policy. Then what happens in the ﬁrst period may affect the standard of integration in the second period. Indeed, real world federations, including the EU, do not work by committing on a particular contingent policy, but by committing on a particular decision making procedure. Suppose then, for simplicity, that in the second period, decisions are still taken by a benevolent dictator, who, in the case that the harmonization policy proves to be beneﬁcial, would naturally choose the harmonization standard so as to minimize the total costs for the federation. Suppose also that this decision maker can now pay lump-sum compensating transfers, so as to focus here only on the non-commitment issue. Then, it can be shown that the case for decentralization and ECA unambiguously worsens under non-commitment. Centralization is not affected by non-commitment; if all countries agree immediately to a common standard in

11

What we show in Bordignon and Brusco (2003) is that with quadratic cost functions, the ECA never requires transfers, and that the ECA line remains unchanged. This may not be true for other cost functions, although the general intuition discussed in the main text would still remain valid.

Institutional Aspects of EU Organization

71

the ﬁrst period, then that standard will also be obviously optimal in the second period if harmonization proves to be beneﬁcial. On the contrary, under both decentralization and ECA, countries have an incentive to choose their standards strategically in the ﬁrst period in order to affect the harmonization choice of the decision maker in the second period. In a Nash Equilibrium, the effect is then certainly a reduction in the expected utility of the federation under these two policies. In particular, it can be shown that the two countries forming a sub-union are able to exploit their first mover advantage, forcing the decision maker in the second period to propose harmonization at a standard that is closer to their historical standards and further from that of the third country. In terms of Figure 1, the result is then a downfall in both the lines representing decentralization and ECA, while centralization remains unaffected. Summing up, then, on efﬁciency grounds, our theoretical analysis suggests that larger heterogeneity and lack of compensating transfers for losers makes the efﬁciency case for ECAs more robust; and lack of commitment makes it weaker. As the EU has neither compensated transfers nor commitment powers, the case for ECA remains largely undecided on theoretical grounds. But as the analysis developed so far has remained fairly abstract; it may be instructive to see what happens if we introduce the arrangements for ECAs as agreed upon in Nice into our framework. As we have already remarked, the Treaty of Nice has eliminated the veto power that the Amsterdam Treaty gave to each country; at the same time, no country can be excluded by an ECAonce this has been formed. Furthermore, ECAs do not work following the decisions of a hypothetical benevolent dictator; in an ECA, the countries themselves decide the policy of the sub-union either by unanimity or by qualiﬁed majority rule, depending on the subject covered by the ECA. For instance, in the case we are discussing here, corporate income taxation, choices inside the ECA would be taken according to unanimity ruling. Suppose then that starting from our theoretical framework, where the individual rationality constraint of each country had to be satisﬁed by the decision maker, we suddenly introduced the Nice rules, thus allowing countries 1 and 2 to go on with an ECA even against the wishes of the third country. What would then be the equilibrium effect on the Union harmonization policies? Our analytical results unambiguously show that the effect is that of making immediate centralization, rather than Enhanced Cooperation, a more likely outcome. More precisely, the set of parameters under which centralization in the ﬁrst period occurs under the ‘‘no veto – no exclusion’’ rule of the Treaty of Nice, is larger than it would be if the individual rationality constraint of the third country had to be respected. Indeed, under the new ECAs’ rules in the Treaty of Nice, the third country is basically confronted with a ‘‘take it or leave it offer’’; it may join the sub-union immediately and then have a say in the sub-union harmonization policy; or it may enter (with some probability) later, and then, as the sub-union policy is determined by unanimity ruling, it has no option but to accept to harmonize at the standard originally chosen by the two other countries. Putting it differently, the Nice rules have given to the willing countries a

72

Massimo Bordignon

credible threat to go on with the harmonization policy, and forcing the more reluctant countries to follow immediately to avoid future exploitation, even if it may actually have been more efﬁcient for the federation at large, to wait a little longer. 2.5. Policy implications These results offer us some insight to discuss the present ECAs’ arrangements in the EU. On the one hand, our results certainly support the idea that ECAs may play an important efﬁciency role for a heterogeneous federation such as the EU. This role, in turn, is further reinforced by the lack of effective transfer mechanisms. Quite intuitively, if the Union is very heterogeneous and the countries who have more to lose by an integration policy on some issue cannot be compensated, or can only be compensated at very large welfare costs, it may be more convenient on efﬁciency grounds to give up complete harmonization altogether, and concentrate instead on subsets of countries which have more to gain from that particular integration policy and for whom compensatory transfers are less relevant. On the other hand, as the EU is unable to commit at a particular integration policy, this policy may negatively affect the excluded countries in the future. Furthermore, the Treaty of Nice, by revoking the veto power still guaranteed by the Amsterdam Treaty to each EU member, has radically changed the rules of the game, making it possible for a (substantial) group of EU countries to carry on with integration against the wishes of the remaining countries. What kind of insights do these arguments offer on the present ECAs’ arrangements, both on positive and normative grounds? On positive grounds, they may ﬁrst help us understand why some countries in the past have opposed the formation of sub-unions by other EU countries, even without (obvious) negative externalities being at play. Perhaps, these opposing countries were not concerned about present or future externalities; they were concerned about the fact that a sub-union today would have made it more difﬁcult for them to harmonize on that issue in the future on terms they would have found acceptable. Second, these insights may also explain why the current arrangements for ECAs in the EU still attempt to offer some decision role on the sub-union policy to the excluded countries (through the EU Commission and the EU Parliament), although it can be doubted whether this role is completely satisfactory, this is a point we will return to later. Third, they may also help us understanding other puzzling aspects of the present EU institutions. For example, several observers have questioned the presence of the UK in the Ecoﬁn, as this Council is in charge of deciding the sanctions for the EMU countries that fail the Maastricht rules, but the UK is currently not a member of the EMU. And indeed, if one simply thinks in terms of present negative externality, this presence makes very little sense. What would be the negative externality for the UK of, say, France running a 4% current deﬁcit over the GDP

Institutional Aspects of EU Organization

73

instead of a 3% deﬁcit as prescribed by the Maastricht rules? The UK’s role, however, acquires a very different meaning if one thinks dynamically, as suggested by our own analysis. Maybe the UK is in the Ecoﬁn, and it should be there, because what the UK is really doing there is to defend its stake for the future. Should the UK decide to join the EMU in the future, it has an obvious interest to check whether this monetary union satisﬁes a number of desirable characteristics, including ﬁnancial stability, and this can be better guaranteed by helping the current EMU countries to adhere to the Maastricht rules. Four, our analysis also helps putting in evidence other unexpected features of the current ECAs’ arrangements. For example, as our formal analysis proved, the main effect of the introduction of the Treaty of Nice rules for ECAs in the EU may not be that of increasing sub-union formation, but that of increasing straightforward centralization. That is, the present ECAs’ arrangements may simply be a device to force reluctant countries to accept more integration. In fact, many observers have argued that the poor decision making rules agreed upon in the Treaty of Nice have been the result of a deliberate attempt made by a number of countries to stop the European integration process altogether by making it more difﬁcult for the Council to arrive at decisions. If this is the case, then the European Commission, who was the great sponsor of revoking the veto power for ECAs in the Treaty of Nice, may have actually managed to outsmart these countries. What these countries have failed to realize, is that the current ECAs’ rules offers a powerful weapon for the European Commission to force integration on reluctant countries by threatening them of future exploitation if they do not accept integration immediately. This takes us directly to the normative side of the analysis. There is no doubt that ECAs can be beneﬁcial if more centralization is efﬁcient on a particular policy domain, but the inability to make compensating transfers does not allow this efﬁcient outcome to arise. Moreover, it is also interesting to point out that ECAs can be beneﬁcial even though their formation is not part of the equilibrium, as this is sufﬁcient to generate the enactment of EU-wide policies. But the interests of the EU countries outside the ECA must also be taken in consideration, As argued above, the current ECAs’ rules consider a number of devices to defend these interests. However, these arrangements hardly seem to be enough. The European Parliament is largely powerless, and it is also quite likely that a coalition of countries who have managed to have a sub-union approved via qualiﬁed majority rule in the Council would also ﬁnd the numbers to have the ECA policy approved by the Parliament, where the decision rule is simple majority. The naı¨ ve idea that just leaving the excluded countries free to join the subunion subsequently is enough to protect them from exploitation is wrong. As we remarked above, countries who join the sub-unions subsequent of their formation are required to comply with the decisions already taken; and although the decision rules inside the sub-union may also contemplate qualiﬁed majority

74

Massimo Bordignon

ruling and not only unanimity as in our formal example above, it is difﬁcult to believe that new entrants could easily revert decisions already taken.12 To repeat a point that has already been made, policies and institutions have their own hysteresis; and in all cases, you do not go back easily from decisions already implemented. This supports our previous claim that it is the European Commission who has been given by the Treaty of Nice rules, the role of a guarantor of the interests of all EU members, both inside and outside the sub-union. By deciding if the sub-union is acceptable, by drafting the proposal for the Council and by deciding ex-post who can join the sub-union, the European Commission is the key player in the current ECAs’ arrangements. This is not necessarily a bad thing; in fact, the Commission has proved in the past that it can play the role of a super-party (or better super-country) institution. But even leaving aside the fact that this may not happen in the future, the Commission, as an institution, has an in-built mission towards increased centralization among the EU countries. There is then a serious risk that the Commission may use the new ECA’s tools to pursue excessive and unnecessary centralization. At the very least, it is important to be aware of this risk as this may suggest countermeasures or may help avoid mistakes. For example, from this point of view, it is clearly not a very good idea to reduce the number of countries inside the EU Commission, as advocated by the Convention and agreed in the Intergovernmental Conference of June 2004, as this weakens the impartial nature of the main institution which now protects excluded countries. Finally, it should also be noted that the efﬁciency role for ECAs is strengthened by the lack of effective compensatory transfers across EU countries. If these transfers were in place, there would certainly be less need of ECAs, as countries could more easily agree on adopting a common policy if they could compensate the losers. Surely, the introduction of effective compensatory transfers across the EU countries would raise difﬁcult political and economic issues, ranging from asymmetric information problems to the deadweight losses of transfers. However, a lot could be done to improve upon the actual transfer mechanisms, which in reality compensate countries by ﬁctionally performing other roles, as is the case for the structural regional funds or for the CAP system for agriculture. Thus, this paper certainly speaks in favor of substituting this highly imperfect transfer system with more explicit cross-countries compensatory mechanisms.

12

In reality, the presence of qualiﬁed majority ruling for ECAs’ arrangements would create more complex strategic behaviour on the part of the countries joining the sub-union, as there would be the possibility of majority reversals in the sub-union in the future, if other countries also joined the subunion (along the lines, for instance, of the work by Fernandez and Rodrick, 1991). The more extreme countries forming the sub-union at the outset would probably search for compensation, asking for integration closer to their standard so as to make majority reversal more difﬁcult. These issues are brieﬂy touched in Bordignon and Brusco (2003).

Institutional Aspects of EU Organization

75

3. The allocation of powers inside the EU and the ‘‘open coordination method’’ 3.1. The problem One of the most fundamental questions in the theory of ﬁscal federalism concerns the correct allocation of functions to the different levels of government. In the context of the EU, this means asking which functions should be advocated by the Union itself, and which functions should remain at the level of the member countries. The Convention has made an effort to streamline the attribution of functions to the different levels of government in the Union, differentiating these competencies in three broad sectors– exclusive competencies of the Union, concurrent competencies, and exclusive competencies of the countries. This categorization has however left the preceding distribution of powers across the two levels of government largely unchanged. Many observers claim that this distribution is inefﬁcient and should be changed. Economists may have some suggestions to offer. According to Oates’s (1972) celebrated decentralization theorem, for example, functions with more (less) spillover effects and less (more) heterogeneity of preferences across jurisdictions should be centralized (decentralized). In its simplicity, this is a recipe, which can prove to be useful. For example, Alesina and Angeloni and Schuknecht, (2001) use this framework to assess the optimality of power assignments inside the EU (see also Stehn (2002) and Inman and Rubinﬁeld (1998) for exercises along the same lines). But, as we have already remarked, the central government in the case of the EU does not have the same powers or the same characteristics of a central government in a federal state, thus making it problematic to apply Oates’s argument to the EU. Furthermore, another important limitation of Oates’s analysis is that he assumes welfare maximizing governments. It is not entirely clear how far his insights could go in more realistic political environments. Unfortunately, these environments are exactly those where the debate toward more or less centralization at the European level is actually taking place. In particular, while the situation about the demand side (monetary and ﬁscal policies) is more or less settled, there is an on going debate on the role that the EU institutions should play on the supply side, particularly in ﬁelds such as labor markets institutions, competition and regulation policy, education, pensions and environment (see, for instance, the Sapir Report, 2003). These are policy dimensions where the current assignment of functions give the member countries most of the decision power (except, to some extent, for competition policy and environment). However, the EU has taken some steps towards a loose form of coordination among these national policies. Under the Lisbon agreement, the so called ‘‘open method of co-ordination’’ calls for peer review on these policies by the member countries, with the aim to try to promote the best practices across the EU countries. Here, the EU institutions cannot force the member countries to follow any particular policy; at most they can only suggest the adoption of a particular policy. Many believe that this is not enough and that a larger role

76

Massimo Bordignon

should be given to the Union (e.g. Sapir Report, 2003); while others feel that this can only be counterproductive, and reduce beneﬁcial competition across the EU countries (e.g. Tabellini and Wyplosz, 2004). In all cases, the ‘‘open method of co-ordination’’ is a strange combination when looked at from the perspective of traditional ﬁscal federalism theory. This theory usually advocates a clear-cut assignment of functions to the different levels of governments, while the open method of co-ordination is an example where the competencies of the Union and the member countries somewhat overlap. To cast light on this debate, Oates’s decentralization theorem is of very little use. National policies on these supply side dimensions clearly induce some spillover effects across countries, but they do not seem to be so relevant as to call for immediate harmonization (as is the case, say, for the common market). On the other hand, due to historical and cultural reasons, there is possibly some heterogeneity in national preferences across these ﬁelds, but they do not seem to be quite relevant to call for complete decentralization (except perhaps, in the case of education). Most importantly, these are also those dimensions where the assumption that national governments choose their policies to maximize national welfare seems to be quite wide off the mark. In fact, most observers would agree that the policies chosen in these ﬁelds are largely inﬂuenced by the pressure exerted by powerful organized national interest groups on national governments; trade-unions in the labor markets, pensioners in the pension system, ﬁrms in the regulation and competition policy, professors in the education system, and so on (e.g. Tabellini and Wyplosz, 2004). The important policy question over which theoretical analysis should attempt to cast some light would then appear to be whether these pressures are likely to become more or less powerful once these functions are centralized at the EU level. Unfortunately, the theoretical literature is of very little help on this particular issue. The recent attempts to extend Oates’s analysis to a political economic framework (Besley and Coate, 2003; Lockwood, 2002; Seabright, 1996), while largely conﬁrming Oates’s intuitions, do not consider the effect of pressure groups on politics. On the other hand, while there is a very large economic literature on lobbying (e.g. Grossman and Helpman, 2001), very few studies have concentrated on the speciﬁc issue of the relationship between interest groups and decentralization. Furthermore, when studies have been completed, they have only focussed on the higher heterogeneity of preferences under centralization as the main discriminating factor for lobbies’ formation and inﬂuence (e.g. Redoano, 2002). But, as argued above, difference of preferences across European countries hardly seems to be the crucial issue in the present context. More policy-oriented economists seem to have clearer cut opinions. For example, in a very inﬂuential policy paper written for the World Bank, Prud’homme (1994) strongly warned against ‘‘the dangers of decentralization’’, the main danger being the (presumed) stronger inﬂuence of local interest groups on local governments. Prud’homme’s main argument has nothing to do with preferences heterogeneity. It relies instead on a greater natural ‘‘disposition’’ by local

Institutional Aspects of EU Organization

77

governments to ‘‘accept’’ pressures from local interests, presumably because of the fact that supporting a local interest may generate additional beneﬁts for local politicians than supporting an external one, like political consensus. This is the same idea, which continuously surfaces, in the political debate on decentralization inside countries and in the proposals that concern harmonization in the EU context. Is this idea correct? If the answer is yes, then maybe there are politically motivated reasons, additional to those considered by Oates’s theorem for, say, supporting centralization at the EU level in the above mentioned ﬁelds. However, if the answer is no, then maybe there are additional, politically motivated reasons to maintain these functions at the level of member states. 3.2. The analysis To discuss this issue, in Bordignon et al. (2003), we build a simple model of lobbying behavior. In this model there are two identical regions, and in each region, there are two private goods and one regional public good. One private good is the nume´raire (and the only factor of production) and it is sold in a competitive market; the other private good is produced by a resident ﬁrm (wholly owned by residents of the same region) and, if allowed by governments, it is sold both at home and in the other region. When both these ﬁrms operate in the two regional markets, the ﬁrms compete a la Cournot in each regional market; otherwise, there is a monopoly in each market. Each regional public good is complementary in consumption to the private good produced by the two ﬁrms, meaning that the two ﬁrms can gain by an expansion in the production of the regional public good. Finally, the regional public good is ﬁnanced by resident taxation on local consumers (including ﬁrm’s proﬁts), so as to avoid tax competition effects. For simplicity, and because they do not appear to be particularly relevant in the EU context, in the model, we abstract entirely from differences in preferences, mobility of individuals, and intergovernmental transfers. In the paper, we consider two polar cases of decision making, a full centralization case where a central government decides the number of ﬁrms allowed to operate in each market and public good supply in each region; and a full decentralization case where both decisions are taken by the local governments. The crucial difference between centralization and decentralization is that in the former, the central government internalizes as components of social welfare the proﬁts that the national ﬁrms make in both markets; while in the latter , the local government is only interested in the proﬁts made everywhere by its own resident ﬁrm. This is the simplest way to capture into the model Prud’homme’s main idea that local politicians may care more for local interests than for foreign ones. In this model, regional ﬁrms have an incentive to lobby the governments, either for expanding local public good production, or for gaining access to the regional market, as both policies would increase their proﬁts. In particular, we consider two polar cases of lobbying. The ﬁrst, which we call lobbying in the

78

Massimo Bordignon

market, is the case when the two national ﬁrms are already operating in both markets. Note that in this case, the two ﬁrms have a common interest in lobbying the governments to increase regional public good supply in both regions. The second, which we call lobbying for the market, is a situation where the ﬁrms are not yet operating in the market and have a conflicting interest to be the only one to operate in each market, as this could earn them monopoly proﬁts. We model lobbying using the common agency approach developed by Bernheim and Whinston (1986) and popularized by Dixit et al. (1997). According to this approach, under lobbying behavior, governments maximize a weighted social welfare function giving weight (u) to social welfare (the sum of consumers and producers surplus, as perceived by the different types of government) and weight (1-u) to lobbies’ contributions, where 0ouo1 measures the ‘‘honesty’’ of the government. This is clearly a ‘‘reduced form’’ of some (un-modeled) more complex political environment. Politicians are interested in social welfare either because they are benevolent or just because they want to be re-elected; but they are also interested in ﬁrms’ money because this may increase their chances of reelection (contribution campaigns) or may increase their private consumption. Thus, in the centralization case, we have two principals (the two ﬁrms) who lobby a single agent (the central government) either for gaining access to the market or for increased regional public good supply; while in the decentralization case, we have two principals (the two ﬁrms) who lobby two agents (the regional governments) for the same policies. This is a more complicated case of lobbying than the simple common agency approach, a case of a ‘‘game played through agents’’ also more formally studied by Prat and Rustichini (2003). The crucial question raised in the paper concerns the effect of lobbying in the different cases. In particular, when is lobbying more damaging for social welfare, under centralization or under decentralization; and for which of the two policies, access to the market or public good provision? 3.3. Results The analysis shows that the answer crucially depends on the characteristics of the policy being lobbied. When lobbying is in the market, decentralization turns out to be unambiguously better for social welfare than centralization. In this case, under centralization, the two ﬁrms always lobby for both regional public goods, and public goods are distorted upward and the resulting equilibrium is Pareto efﬁcient; under decentralization, there are equilibria where both ﬁrms lobby both local governments as under centralization, but in this case lobbies need to pay more (earn less proﬁts) to induce the same distortion. Furthermore, under decentralization, depending on the market structure, there are also equilibria where lobbies only lobby one local government at a time (often the foreign one), public good supply is less distorted and some of these equilibria are Pareto inefﬁcient, which in our case is good news as it means that lobbying is less effective.

Institutional Aspects of EU Organization

79

On the contrary, when lobbying is for the market, centralization turns out to be unambiguously better for social welfare than decentralization. In this case, under decentralization, the local ﬁrm always outbids the foreign one; in equilibrium, there is then always a local monopoly in each market (each resident ﬁrm always serves his own regional market only), even if a duopoly would be more efﬁcient. Under centralization, on the contrary, the central government is more often prepared to allow both ﬁrms in the market (that is, there is no empty set of parameters where this happens in equilibrium). These different results depend on the fact that under decentralization, as compared to centralization, the local government is only interested in the proﬁts of the local ﬁrm, and this gives the latter an easy way to outbid the foreign ﬁrm. Interestingly, in the lobbying for the market case, we also show that a different regime of split competencies (where the central government decides about who enters, while the local government decides about public good supply) often dominates both centralization and decentralization. This is because this last regime reduces the rents that a dishonest policy-maker can exert from ﬁrms when offering them monopoly powers in the markets, and this forces the policy-maker to make the efﬁcient choices more often. The main intuition behind these results is simple. When the interests of national lobbies are aligned, as is in the case of lobbying in the market, decentralization makes it more difﬁcult for the national lobbies to coordinate their actions; and they have to pay more to convince the local government to internalize even the other ﬁrms’ proﬁts. Hence, decentralization is unambiguously better than centralization, and lobbying is less important and less distortive. And, vice versa, when the interests of national lobbies are in conflict, as is in the case of lobbying for the market, local governments are more easily captured by local interests, as correctly suggested by Prud’homme. Hence, centralization is unambiguously better than decentralization. Clearly, this analysis cries for extensions. In particular, for simplicity, we eliminated any form of external effects (spillover effects in public good productions, mobility of ﬁrms across regions induced by ﬁscal policies, etc.). But clearly these may be important even where lobbying is concerned. For example, while in the traditional theory the presence of spillovers, whether positive or negative, is always an argument for more centralization, one could conjecture that their effects would be different once lobbying behavior is taken into account. In the presence of negative spillovers, lobbies are more likely to have conﬂicting interests, while when there are positive spillovers, lobbies’ interests are more likely to be aligned. Hence, our analysis would suggest moving more towards centralization when externalities are negative and towards decentralization when they are positive.13

13

I owe this point to an anonymous referee.

80

Massimo Bordignon

3.4. Policy implications for the EU This analysis then suggests that the answer to the question if it is better to centralize a particular policy when lobbying behavior is likely to be important crucially depends on how the interests of national lobbies are positioned with respect to that particular policy. If the national lobbies have common interests on that policy, then it is better to decentralize, as decentralization induces a sort of competition across lobbies when there is none; on the other hand, if national lobbies have conﬂicting interests, then it is better to centralize as local governments are more easily captured by local interests. Going back to the EU debate, one notes that in ﬁelds such as consumer and environment protection, foreign and domestic producers would have the same interest to lobby for low consumers’ protection if this policy were decided at the EU level. They would do the same if the policy remained at the local level, of course, but then each country would have no interest to internalize the effects of low consumers’ protection on the proﬁts of foreign ﬁrms, leading to lower distortions. Ceteris paribus, then, our argument would suggest not centralizing these functions. But, in ﬁelds such as production subsidies to national producers, protection of market share of incumbents and ‘‘national champions’’, national lobbies have conﬂicting interests and centralization at the EU level would force the policy maker to take into account the interests hurt by the protection policy. Hence, ceteris paribus, our argument would suggest centralizing these functions. Finally, in the latter case, our argument also suggests that a policy of ‘‘split competencies’’ may often be superior on efﬁciency terms to either centralization or decentralization, because it induces a sort of competition between the different levels of government, which makes them more resilient to lobbying. While this result clearly needs further enquiries, it is worth stressing. We derive concurrent competencies as the optimal institutional solution to a political constraint (the presence of lobbying). This is in contrast to the received ﬁscal federalism theory, which sees overlapping competencies as inefﬁcient institutional arrangements. As concurrent competencies are instead often observed in existing federations, besides the EU, one then wonders if lobbying could provide a possible rationale for these arrangements. In the EU case, our result clearly lends support to some form of concurrent competencies between the Union and the member states, as we observe, for instance, in the ‘‘open method of co-ordination’’ case. 4. Concluding remarks The European Union has some very special characteristics. It is less than a federal state and it is more than an economic union; it follows that in assessing its institutions the knowledge accumulated on other federations or economic unions is often of limited use. In particular, the economic theory of ﬁscal federalism, developed for different institutional contexts, needs to be seriously amended when applied to the EU. In order to offer relevant policy advice,

Institutional Aspects of EU Organization

81

economists ﬁrst need to make a serious effort to understand the functioning of the EU’s peculiar institutions. The previous analysis has offered some examples on how one could proceed. The traditional economic recipe about the optimal assignment of functions to different levels of government makes little sense when the central government has the peculiar characteristics and decision making rules of the EU institutions. Furthermore, in the presence of relevant lobbying behavior, one must also understand how centralization and decentralization would change the political incentives of the different levels of government. The analysis of sub-union formation and governance inside the EU raises subtle theoretical and policy issues, which can only be appreciated by putting them into the context of the functioning of the other institutions of the Union. Dynamic considerations and the strategic behavior of the actors involved reveal unexpected consequences of the existing rules for ECAs, leading one to question their optimality. However, this is only the tip of the iceberg. There are several other ﬁelds that require careful enquiry. The precise relationship between the different decisionmaking bodies of the Union, the Commission, the Parliament and the Council, for example, rests largely unexplored on both theoretical and empirical grounds (for a preliminary analysis, see Noury and Hix and Roland, 2003). The issue of which taxing powers should be given to the Union, if any, is still little discussed in the literature; the problem of the implementation of the economic policy of the Union, lacking executive powers by the center is still an unsolved and interesting element for the analysis. Economic theorizing, and in particular, the economic literature on the relationship between different levels of government could be important for better understanding of these features. But for this contribution to be successful, an effort must be made to incorporate the analysis into the actual functioning of the EU institutions, rather than pretending to directly apply the received theory to a completely different context.

Acknowledgement This is an amended and extended version of a paper I gave at the CESifo-Delphi Conference, Designing the New E, in Delphi, in June 2004. I wish to thank an anonymous referee for her/his very useful suggestions in revising the paper, the two organizers, Helge Berger and Thomas Moutos for their patience, support and encouragement in writing this paper, and all the participants at the conference, in particular, Giuseppe Bertola, Gerard Roland and Jurgen Von Hagen for their very useful comments on the early version of this paper. Furthermore, this paper is also strongly indebted to my joint works with several co-authors, in particular, Sandro Brusco, Luca Colombo and Umberto Galmarini. While their contribution to the ideas of this paper should be emphasised, they are not responsible for the errors, if any.

82

Massimo Bordignon

References Alesina, A. and V. Grilli (1993), ‘‘On the feasibility of a one-speed or multispeed European Monetary Union’’, Economic and Politics, Vol. 5, pp. 145–165. Alesina A., I. Angeloni and L. Schuknecht (2001), ‘‘What does the European Union do?’’, NBER Working Paper, No. 8647. Alesina A., I. Angeloni and F. Etro (2003), ‘‘The political economy of ﬁscal unions’’, American Economic Review, Mimeo, Harvard University, forthcoming. Baldwin, R.E., E. Berglo¨f, F. Giavazzi and M. Widgre´n (2001), ‘‘Nice Try: Should the Treaty of Nice be Ratiﬁed?’’, CEPR, Monitoring European Integration 11, London: CEPR. Berglof, E., B. Eichengreen, G. Roland, G. Tabellini and C. Wyplosz (2003), ‘‘Built to Last: A Political Architecture for Europe’’, CEPR, Monitoring European Union 12, London: CEPR. Bernheim, B.D. and M.D. Whinston (1986), ‘‘Menu actions, resource allocation, and economic inﬂuence’’, Quarterly Journal of Economics, Vol. 101, pp. 1–31. Besley, T. and S. Coate (2003), ‘‘Centralized versus decentralized provision of local public goods: a political economy approach’’, Journal of Public Economics, Vol. 87, pp. 2611–2637. Bond, S., L. Chennel, M.P. Devereux, M. Gammie and E. Troup (2000), Corporate Tax Harmonization in Europe: A Guide to the Debate, London: Institute of Economic Studies. Bolton, P. and M. Whinston (1993), ‘‘Incomplete contracts, vertical integration, and supply assurance’’, Review of Economic Studies, Vol. 60, pp. 121–148. Bordignon M. and S. Brusco (2003), ‘‘On enhanced cooperation’’, CESifo Working Paper, No. 996 (revised 2004), Journal of Public Economics, forthcoming. Bordignon M., L. Colombo L., and U. Galmarini (2003), ‘‘Fiscal federalism and endogenous lobbies’ formation’’, CESifo Working Paper, No. 1017 (revised 2004). Dewatripont, M., F. Giavazzi, I. Harden, T. Persson, G. Roland, A. Sapir, G. Tabellini and J. von Hagen (1995), ‘‘Flexible Integration: Towards a More Effective and Democratic Europe’’, Monitoring European Integration 6. (MEI), London: CEPR. Dixit, A., G.M. Grossman and E. Helpman (1997), ‘‘Common agency and coordination: general theory and application to government policy making’’, Journal of Political Economy, Vol. 105, pp. 752–769. Fernandez, R. and D. Rodrick (1991), ‘‘Resistance to reform: status quo bias in the presence of individual-speciﬁc uncertainty’’, American Economic Review, Vol. 81, pp. 1146–1155. Grossman, G.M. and E. Helpman (2001), Special Interest Politics, Cambridge. MA: MIT Press.

Institutional Aspects of EU Organization

83

Inman R., and D. Rubinﬁeld (1998), ‘‘Subsidiarity and the European Union’’, NBER Working Paper, No. 6556. Keen, M. (1998), ‘‘Vertical externalities in the theory of ﬁscal federalism’’, IMF Staff Papers, Vol.45, No.3. Lockwood, B. (2002), ‘‘Distributive politics and the costs of centralization’’, Review of Economic Studies, Vol. 69(2), pp. 313–337. Noury A., S. Hix and G. Roland (2003), ‘‘How to choose the European executive: a counterfactual analysis’’, Working Paper, University of California, Berkeley. Oates, W. (1972), Fiscal Federalism, New York: Harcourt-Brace. Perotti, R. (2001), ‘‘Is a uniform social policy better? Fiscal federalism and factor mobility’’, American Economic Review, Vol. 91(3), pp. 596–610. Prat, A. and A. Rustichini (2003), ‘‘Games played through agents’’, Econometrica, Vol. 71(9), pp. 989–1026. Prud’homme, R. (1994), ‘‘On the dangers of decentralization’’, World Bank Policy Research Working Paper, No.1252. Redoano, M. (2002), ‘‘Does centralization affect the number and size of lobbies?’’ Mimeo, University of Warwick. Ruding Committee (1992), ‘‘Report of the committee of independent experts on company taxation’’, European Commission, Brussels. Sapir Report (2003), ‘‘An agenda for a growing Europe’’, European Commission, Brussels. Seabright, S. (1996), ‘‘Accountability and decentralization in government incomplete contracts model’’, European Economic Review, Vol. 40, pp. 61–91. Stehn J. (2002), ‘‘Towards a European constitution: ﬁscal federalism and the allocation of economic competences’’, Working Paper, No.1125, Kiel Institute for World Economics. Tabellini, G. (2003), ‘‘Principles of policy-making in the European Union: an economic perspective’’, Mimeo, Bocconi University, Milan. Tabellini, G. and C. Wyplosz (2004), ‘‘Supply-side policy coordination in the European Union’’, Mimeo, Bocconi University, Milan. Treaty of Nice (2002), http://europa.eu.int/comm/nice_treaty/index_en.htm Widgre´n, M. (2001), ‘‘Optimal majority rules and enhanced cooperation’’, Working Paper, No. 3042, London: CEPR.

This page intentionally left blank

Comment Stefan Voigt

The European Union (EU) is one of its kind. This is the starting point of Bordignon’s thoughtful and well-written paper. He is, of course, not the ﬁrst to point out that the EU is neither a traditional international organization nor a typical federation. But he is serious in warning us against assuming that insights valid for federations could simply assumed to hold true for this unique animal. He, however, suggests an approach to tackle this problem that consists of ﬁrst analyzing the existing institutions, second deriving theoretical questions there from, which could then lead both to explanations otherwise hard to get and, eventually, to unexpected policy implications. This approach of dealing with a unique institutional phenomenon (the EU) is applied with regard to two areas, (1) the introduction of formal rules to form sub-unions inside the EU (also known as Enhanced Cooperation Agreements, ECA for short) and (2) the question of how competences should be optimally allocated in the EU. Bordignon has come up with answers to both areas, which deviate from conventional wisdom. With regard to the ﬁrst area, he argues that creating ECAs could lead to further harmonization and centralization. After claiming that the theory of ﬁscal federalism would be of little use with regard to the second topic, he emphasizes the crucial role of lobbying groups: centralization or decentralization of a task can depend on whether the interests of lobbying groups from different states are aligned or competing. The paper is based on a number of more formal papers. Both the methodological approach and the two examples are, at least in general, convincing. Yet, I will try to uncover some problems and deﬁcits in both the methodological approach as well as in the two examples.

CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79011-3

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

86

Stefan Voigt

1. The methodological approach In times of unprecedented developments, one can often hear the call for a speciﬁc theory to deal with such a situation. The events of 1989 and 1990 in Central and Eastern Europe, e.g., soon led to a call for a speciﬁc theory of transition. It would almost seem to suggest that a theory of supranational institutions should be called for with regard to the EU now. This is, however, not what Bordignon calls for. His argument, as I understand it, is less demanding and more convincing. Nomological hypotheses are of the kind ‘‘always and everywhere if x, then y’’. In other words: they pretend to be valid regardless of time and place. The degree to which the antecedence x is speciﬁed is an important aspect to determine the empirical content of the hypothesis. The higher the number of antecedence conditions, the lower the empirical content of the hypothesis is said to be. The antecedence conditions of some nomological hypotheses are often not spelled out explicitly. It is simply stated that rule x will lead to result y, but it is not stated that result y can only be expected if rule x is applied within a speciﬁc context. This could, e.g., be the context of federal institutions. If Bordignon warns us against hastily drawing on insights gained within the theory of federalism, he also reminds us that often, initial conditions are only insufﬁciently speciﬁed. If we force ourselves to spell out the antecedence conditions to the maximum possible extent, two objectionable consequences may pop up: we might notice that we only know these conditions imprecisely. Federalism may also be a case in point: the nomological hypothesis that rule x would lead to result y given that federal institutions are in place might be insufﬁciently speciﬁed if we are unable to precisely spell out what federal institutions are. After all, there are quite a few federal states with seemingly similar institutions that still develop in different directions. The second consequence is a direct result of the ﬁrst: if the antecedence conditions cannot be completely spelt out, we would be forced to question the value of our nomological hypothesis altogether. This leads us directly to a second problem: Suppose the set of our nomological hypotheses is severely limited, but we are still interested in setting up, and possibly improving, a supranational union. Such a situation is not unique. The nation state as a speciﬁc institutional setting emerged during the 17th and 18th century without pre-existing blueprints. The U.S. Constitution of 1787 was unprecedented in a number of dimensions. Hence, it could be interesting to take a look at the approach used by the fathers of the U.S. Constitution and ask whether a similar approach would make sense today. Cain and Jones (1989, p. 12) describe Madison’s approach as ‘‘experimental, empirical, and circumstantial, not deductive and theoretical’’. They point to the three reasons that made Madison refuse the axiomatic-deductive approach: (1) choice always takes place under uncertainty, (2) language is always imprecise, and (3) institutional designers are themselves not perfect (ibid., 13). According to Cain and Jones, in

Comment

87

evaluating theories, Madison appreciated ‘‘(p)racticality and real-life problemsolving, rather than symmetry, elegance, and deductive simplicity’’ (ibid., 14). To achieve this, Cain and Jones ﬁnd three rules in Madison’s thinking: ‘‘(1) Accept the fact that knowledge about political organizations and human behavior is imprecise, incomplete and often circumstantial. (2) Rely on induction from past experiences to develop working hypotheses about institutional proposals. (3) Observe how institutions are actually operating, and aim to improve them incrementally’’ (ibid., 15).

Madison’s approach is thus squarely opposed to the deductive reasoning used in economics. The policy implications that he arrived at, are, however very modern . Vincent Ostrom writes about Madison: ‘‘Since perfect solutions cannot be expected, the practical constitution maker must consider the disadvantages in relation to the advantages inherent in alternatives rather than consider the existence of disadvantages to be evidence that a solution is not perfect, and, thus, to be rejected’’ (Ostrom, 1971, p. 23). Hence, Madison had opted for comparative institutional analysis. We can reconstruct Madison’s approach whereas Bordignon’s stays a bit opaque. Methodological issues are seldom dealt with explicitly. Bordignon deserves credit for begging to differ. Yet, his remarks on the methodology are extremely limited, and a number of important issues remain unclear. Some of them have been mentioned in this discussion. 2. Enhanced Cooperation Agreements The higher the degree of heterogeneity among member states of the EU, the less likely it seems that they will be able to agree on any common policy. In order not to prevent a more intensive cooperation among a subgroup of EU members, it is possible to create the so-called Enhanced Cooperation Agreements (ECAs). Bordignon argues that even in the absence of any negative externality, ECAs can damage the interests of non-participating countries. He spells out the conditions under which ECAs appear to be welfare enhancing (under a large degree of heterogeneity among the members and the absence of compensating transfers for losers) and those under which they appear to be questionable (if the sub-union cannot credibly commit to a particular harmonization policy ex ante). The main insight is that the institutional rules of ECAs have allowed the countries to prefer further harmonization and given them a credible threat to continue further harmonization, which alone could damage the interests of the non-participating countries. This, in turn, could imply that the most important effect of the ECA rules would be to increase straightforward centralization rather than to induce enhanced cooperation among a subgroup of countries. The point depends on the subgroup agreeing on a common policy that is even farther away from the ideal point of the non-participating countries, as

88

Stefan Voigt

compared to a common policy that would arise were all countries to agree. If the common policy of the subgroup creates a standard or a focal point, there is the danger that the non-participating countries will have to accept this standard some time later. In order to prevent this from happening, these countries would be ready to enter into compromises that they would have not agreed to in case ECAs were not possible.1 Bordignon’s point is theoretically convincing, yet difﬁcult to refute empirically: the argument is, after all, that the effect of the possibility of ECAs is not that many ECAs will be factually agreed upon, but is rather that centralization will further increase. One can indeed show that not many ECAs have been concluded so far (to be precise, not even a single one). But can one also show that centralization has received a new push because of the introduction of the institutions dealing with ECAs? In order to (threaten to) found an ECA, one third of all member states need to agree with the proposal. In theory, this can imply that in all policy areas in which a third of all member countries want to centralize substantially more than the rest, a minority could force its will upon the majority. This idea seems to be subject to a number of criticisms, however: it only holds true if the minority countries can systematically deﬁne the standard that will later on be used by everybody. But if the majority of member state governments believe that this is unrealistic, no centralizing drive would result. Suppose countries are made to participate in centralization moves simply to avoid even worse outcomes. However, this does not seem to be a sustainable result in the long run: Eventually, a degree of centralization might result in a country being worse off than not being a member of the EU at all. It may then prefer to completely exit from the Union. Bordignon shows the fact that the ECA members are unable to a speciﬁc policy ex ante makes the welfare effects of ECAs open to debate. Here, it would have been interesting to think about the possible means to increase this commitment. 3. On the optimal allocation of powers After having dealt with positive issues in the ﬁrst example (what are the potential effects of a speciﬁc institution?), Bordignon turns to a normative question: how should powers be allocated in the EU? More speciﬁcally: what supply side

1

The same argument can be made with regard to EU membership: if a club of countries (the EU) agrees on rules that can possibly have negative effects on a current outsider, this could create incentives for the outsider to apply for membership. Enlarged EU membership could also constitute a brake for the centralizing tendencies due to the possibility of (threatening to found) ECAs: the higher the number of members, the more difﬁcult it will be to convince a third of all member states of the potential advantages of an ECA.

Comment

89

policies should be allocated on the EU level? He claims that the insights of ﬁscal federalism a` la Oates would be ‘‘of little use’’ and extends this critique to some recent attempts to integrate political economy considerations into the basic Oates framework. These extensions would be of little help, as they do not explicitly recognize pressure groups to be relevant political actors. Lobbying groups, and their interests, are the key for Bordignon to answer his question. In short, if the interests of national lobbies are in conﬂict with each other, he recommends that this policy area be allocated at the European level, but if they are aligned beyond nation state borders, decentralization would be the better choice. The intuition is straightforward: if national lobbies are in conﬂict with each other, local governments might be more easily captured with detrimental welfare effects by their own interest group. It would, hence, appear that allocating the competence to the EU level would also lead to increased welfare. A number of points can be made with regard to this argument. The reason offered for the non-applicability of Oates’ decentralization theorem does not convince me. Bordignon writes that ‘‘the central government in the case of the EU does not have the same powers or the same characteristics of a central government in a federal state, making it problematic to apply y Oates’s argument to the EU.’’ We are dealing with a normative question here, namely the optimal allocation of policy areas to governance levels, but Bordignon implicitly assumes the institutions and their competences as exogenously given. This is not convincing as a starting point. Yet, at the same time, the critique of Oates does not appear strong enough: if, as Bordignon argues, pressure groups are relevant actors in the political game, and this also holds true for more conventional federations, then the decentralization theorem is seriously incomplete even under more conventional federal settings. To decide issues concerning the allocation of competences on the basis of the degree to which preferences of various lobbying groups are aligned seems debatable, as it is supposedly based on the assumption that both the organizational structure and the preferences of interest groups are very stable.2 They might, however, be endogenous to the policy levels chosen. But suppose the structure and preferences of interest groups are sufﬁciently stable. In this case, I would still conjecture that the alignment and conﬂict of interests does not occur along the lines of policy areas, but rather along the lines of policy instruments. This could imply multi level governance structures. Bordignon does mention concurrent competencies at the end of the chapter. It would have been interesting to learn a bit more about them.

2

Otherwise, one might get into strange circles: in period one, interests of lobby groups are in conﬂict, hence centralization is chosen. Interest groups now have more incentives to unite and stress common interests in Brussels. This would, however, imply that the policy should be decentralized again.

90

Stefan Voigt

4. Summing up This is a very well written paper that convincingly argues that the study of institutional details is important to get the story right. To a degree, both examples question conventional wisdom, which makes reading the paper stimulating. The comments on the methodological approach are an attempt to add to the very short section that deals with these issues in the chapter. And, the comments on the two examples tried to discuss some of the possible implications. References Cain, B. and W. Jones (1989), ‘‘Madison’s theory of representation’’, pp. 11–30 in: B. Grofman and D. Wittman, editors, The Federalist Papers and the New Institutionalism, New York: Agathon. Ostrom, V. (1971), The Theory of a Compound Republic, Blacksburg: Center for Study of Public Choice.

CHAPTER 4

Welfare Policy Integration Inconsistencies Giuseppe Bertola The policy of Europe, by obstructing the free circulation of labor and stock both from employment to employment, and from place to place, occasions in some cases a very inconvenient inequality in the whole of the advantages and disadvantages of their different employments. – Adam Smith, The Wealth of Nations, Chapter X: ‘‘Of wages and proﬁts in the different employments of labor and stock’’.

1. Introduction The current policy of Europe, at least according to the letter of its Treaties, is well aware of the problems discussed by Adam Smith some 250 years ago. The Treaty Establishing a Constitution for Europe (signed on 29 October 2004 by the member countries’ Heads of State or Government, subject to ratiﬁcation at the time of writing) lists mobility and social progress among the Union’s objectives in Article I-3: 2. The Union shall offer its citizens y an internal market where competition is free and undistorted. 3. The Union shall work for y a highly competitive social market economy, aiming at full employment and social progress, y It shall combat social exclusion and discrimination, and shall promote social justice and protection, y

These and other objectives, however, are to be pursued within the institutional framework of ‘‘subsidiarity’’, as deﬁned in Article I-11: 3. y in areas which do not fall within its exclusive competence, the Union shall act only if and insofar as the objectives of the proposed action cannot be sufﬁciently achieved by the Member States, either at central level or at regional and local level, but can rather, by reason of the scale or effects of the proposed action, be better achieved at Union level.

Social policy is not currently among the competences of the European Union (EU). Unfortunately, social protection and undistorted competition need not be achieved in the resulting policy framework any more effectively now than in CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79004-6

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

92

Giuseppe Bertola

Adam Smith’s times when, as discussed in other quotes below from the same chapter of the Wealth of Nations, economic integration was occurring within rather than across nation-states. This paper reviews the purposes of social and labor market policy in modern, interrelated economic systems, where the tension between social equity and productive efﬁciency may or may not be adequately addressed by market and government interactions. It focuses in particular on issues arising from present arrangements regarding the assignment of power at different levels of governance, in this and other ﬁelds, within the European system of economic policies. Section 2 reviews brieﬂy the purposes and tools of social and labor market policy and its predicaments in environment, such as the Europe of Adam Smith and that of present times, where the scope of political decision-making does not coincide with the scope of economic interactions. Section 3 summarizes the conceptual framework of Sapir et al.’s (2004) approach to policy evaluation: conﬂicts of interest, not appropriately mediated by political processes, can easily result in incoherent and ineffective policy structures. Possible solutions are illustrated with examples from European experiences in the ﬁelds of competition and macroeconomic policies. The history and prospects of European labor and social policies are reviewed in Section 4. Broad agreement on the importance of the issues contrasts with very little agreement as regards how desirable aspects of policymaking in this area may be reconciled with the new challenges and opportunities of an ever larger, and perhaps ever closer, union of Europe. Section 5 focuses on a comparison of the Richter (2002)/Sinn-Ochel (2003) and Bertola et al. (2001) approaches to the relevant issues. It returns to Adam Smith’s analysis of his own time’s attempts to regulate or delay integration of diverse populations, and outlines how reform efforts could instead be focused on reappraisal and modernization of social responsibilities at all levels of governance in the EU’s economic system. Section 6 concludes arguing that a comprehensive policy framework can reconcile economic integration and politically important social objectives. The policy of Europe, at a crossroads between development and stagnation, should resist temptations to revert to a dubious Golden Age and develop a forwardlooking, constructive, and coherent approach to reconciling its citizen’s needs and its economy’s evolving constraints. 2. Welfare policy in an integrating world Governments should address problems that markets cannot solve efﬁciently (Sinn, 2003). As regards distribution of labor income across differently fortunate individuals, problems whose market solution is socially and politically unsatisfactory abound. Following Bertola et al. (2001), it is helpful in this respect to recognize that social policy has a variety of goals. Policy intervention aims on the one hand at transferring resources to ‘‘excluded’’ individuals, who would

Welfare Policy Integration Inconsistencies

93

otherwise experience extreme poverty; on the other, at insuring, against future misfortune, individuals who currently enjoy adequate income levels. The ﬁrst objective is primarily pursued by unconditional or targeted beneﬁts, granted as a right of citizenship and ﬁnanced by general tax revenues. Some individuals are both ex ante and ex post unable to earn in the market such an income as would be necessary for them to be included in society, and society chooses to help them (and prevent them from offering unpalatable poverty to the public eye or engaging in crime). The second objective is primarily pursued by employment-based social-security systems, whose experience-rated beneﬁts are ﬁnanced by mandatory contributions. Markets, even in a modern and sophisticated economy, are not well equipped to handle labor income risks. Hence, ex ante similar individuals are exposed to substantial ex post income risk in a laissez faire situation, and may well agree to collectively implement redistribution schemes.

2.1. Choices and changes Different and evolving socio-economic systems have historically emphasized the two objectives differently, and have used widely different policy instruments to address them. The notion of ‘‘citizenship’’ as a body of jus soli obligations and rights that can be granted or acquired (rather than inherited as a birthright on a jus sanguinis basis) was introduced in ancient times, and has always been controversial. Under Pericles, the city-state of Athens restricted citizenship to persons whose mother and father were both citizens, but soon after his death citizenship was granted to many thousands of immigrant residents; in Roman times citizenship was important and importantly restricted until AD 212, when it was ﬁnally granted to all free residents of the empire. Throughout history, much of the reciprocal support that people offer each other has been based on nearness and blood relation, rather than on formal legislated rights and obligations. Before the industrial revolution, extended families and villages shared common resources without formal market or government organization – and still do, as village- and family-level interactions continue to play a crucial role in less developed nations. Family, friendship, and non-proﬁt relationships still shape an important portion of individuals’ access to economic welfare within industrialized societies where a complex and heterogeneous set of formal government interferences with labor income has developed alongside an equally complex, and evolving, set of market interactions. Progressive enlargement of the scope of economic interactions has been an extremely important source of economic progress. While the village might have been enclosed by well-deﬁned and largely impermeable boundaries, members of industrialized societies interact from a distance on markets that have increasingly expanded throughout history. Since any collective policy intervention needs to be mandatory and appropriately enforced, the span of government has

94

Giuseppe Bertola

also expanded beyond the boundaries of prehistoric villages, to enclose broader geographic and political entities. To the extent that economic integration has distributive effects, enforcement and effectiveness of social and labor market policies are obviously undermined by new margins of choice by economic agents. In the modern globalized world, product market integration makes it possible for producers to avoid domestic taxation by relocating abroad, and migration can be motivated both by tax avoidance and by welfare beneﬁts when they are available in the destination country. But these problems are not new, as is clear from Adam Smith’s discussion of the relevant issues. Smith was observing the transition from a parishbased to a national organization of economic interactions, and found that among the sources of comparative advantages limited exploitation the provision of welfare played an important role. Parishes were charged with welfare provision by the Poor Laws: y it was enacted by the 43rd of Elizabeth, c.3. that every parish should be bound to provide for its own poor y. And that overseers of the poor y with the churchwardens should raise, by a parish rate, competent sums for this purpose. ... Who were considered as the poor of each parish became, therefore, a question of some importance

and a ‘‘residence-based’’ entitlement criterion was therefore established: it was at last determined by the 13th and 14th of Charles II when it was enacted that forty days undisturbed residence should gain any person a settlement in any parish; but that within that time it should be lawful for two justices of the peace, upon complaint made by the churchwardens or overseers of the poor, to remove any new inhabitant to the parish where he was last legally settled.

As noted by Hanson et al. (2002), this situation readily generated incentives to dump poor individuals over the boundaries of parishes: Some frauds, it is said, were committed in consequence of this statute; parish ofﬁcers sometimes bribing their own poor to go clandestinely to another parish and by keeping themselves concealed for forty days to gain a settlement there, to the discharge of that to which they properly belonged.

Adam Smith proceeds to outline and discuss very interesting details regarding various permit-based reﬁnements of the Elizabethan residence-based system of welfare provision: those ‘‘origin’’-based rules are discussed below, and the historical experience proves relevant to current discussions of the shortcomings and merits of those and other entitlement rules. Over the following century, the problems he was identifying were aggravated by urbanization. Industrial workers were no longer as able to rely on common properties and family networks, and their welfare needs were only partially addressed by institutions such as the workhouse for the poor, familiar from Dickens’s novels. The bureaucracy-based welfare state, as we know it, was then introduced and developed in the context of the European militaristic nation-state, a fact that has interesting implications as formation and continued integration of the EU may make that concept obsolete. In England and other parts of Britain, during and

Welfare Policy Integration Inconsistencies

95

after World War II, Lord Beveridge engineered a universal system of welfare provision, addressing primarily the ﬁrst (poverty-prevention oriented) general task of redistribution policies. Similar concern with the welfare of increasingly urbanized workers was at the root of other policy systems, most notably, the employment-related old-age and sickness beneﬁts introduced in Bismarck’s 19th century Germany. Individuals’ welfare needs, no longer catered to efﬁciently by villages and families in industrial economies, were thus addressed by national schemes. Of course, Lord Beveridge did not mean to relieve the poverty of immigrants. Chancellor Bismarck was keenly aware of welfare provision’s competitiveness implications: at a conference with trading partners, Germany tried (and failed) to achieve agreement on welfare provision minima. But national welfare systems were clearly meant to protect individual persons against speciﬁc risks, such as that of accidents and bad weather in the construction industry: they could not cope with permanent competition by foreigners. The twentieth century that saw introduction and development of elaborate government redistribution systems and formal citizenship rights at the nation-state level also, not coincidentally, saw the introduction of passports and of increasingly formal regulation of immigration and citizenship – whether on the basis of blood, as in Germany (until recently), Italy, and Japan, or on the basis of birthplace and/or parents’ residence, as in France and in the United States. These and other historical heritages shape the variety of welfare state systems within the EU (Esping-Andersen, 1990; Bertola et al., 2001). Nordic countries (Sweden, Finland, Denmark) and the Netherlands have a tradition of full employment and universal welfare provision, feature relatively generous unemployment insurance beneﬁts, and a very important role for active labor market policies (including job creation in the public sector), while social assistance plays a residual role. Continental countries (Austria, Belgium, France, and Germany) derive from the Bismarckian tradition: wage determination is mostly centralized, and generous unemployment insurance beneﬁts and stringent employment protection legislation leave a residual role for social assistance as a general basic safety net, while pensions and health services are provided on an occupational basis. The Anglo-Saxon countries (UK and Ireland) are closer to the Beveridgian tradition: social assistance schemes as a safety net for a relatively unregulated and unequal labor market, with relatively low unemployment insurance beneﬁts, little employment protection, and decentralized wage-setting. Southern European countries (Greece, Italy, Portugal, and Spain) have more recent and less precisely deﬁned welfare states, where extended family arrangements still tend to play a nontrivial role. As the scope of economic interaction has further enlarged across the borders of nations, the heterogeneous status quo welfare systems need not be able to deliver their intended objectives. Obviously, solidarity concerns depend on a society’s social and racial homogeneity, and it is far from surprising to ﬁnd (Alesina and Glaeser, 2004) that redistribution and poverty-reducing policies are

96

Giuseppe Bertola

far more widespread in Europe’s relatively stable populations than in America’s immigration-rich and racially diverse ones, and that they are especially important in Scandinavian countries characterized (at least historically) by extremely uniform ethnic and social backgrounds. Less obviously, and very importantly, the nature and character of traditional welfare systems has to be revised in light of new economic integration developments. Just as village-based solidarity systems had to give way to state schemes in the aftermath of the Industrial Revolution, a system of economic interactions that generates welfare across the boundaries of nation-states calls for new market and non-markets risk-sharing instruments. The nation has long been the natural decision and implementation unit for social policy. But this need not be a permanent arrangement, and the concept of nation is itself a matter of different and evolving deﬁnition, as exempliﬁed by the different link in the French and the recently reformed German rules between nationality on the one hand, and birthplace, culture, and ancestry on the other. 3. Economic policy in the EU Economies and societies are always shaped by the interaction of market forces and collectively decided policies, neither of which guarantee wholly desirable outcomes. Following Sapir et al. (2004), it is important to acknowledge that all policies have both potentially positive and potentially negative effects; that these effects are differently desirable across different groups of agents; that political interactions among such groups and administrative procedures need not, in general, ensure better outcomes than laissez faire would bring; and that the relationship between the scope of policies and of the governance process that implements policy instruments is crucial. The organization of economic and political interaction across national lines that emerged in the 1800s performed poorly in the early part of the 1900s, which saw two World Wars and a Great Depression deepened by trade and migration barriers. The origins of what is now the EU can be traced to the unfortunate tendency of European nation-states to ﬁght each other’s coalitions for economic and political supremacy. After World War II, the original six members chieﬂy wished to organize their relationship to each other in a way that would prevent future wars. Accession by Spain, Portugal, and Greece had important implications for those countries’ commitment to democracy, and EU accession also has obvious political and implications for the formerly Communist Central and eastern European countries. Thus, a variety of different, if similarly strong, political reasons for European accession led the member countries to ‘‘ever closer integration.’’ Economic policy instruments always played a major role in that framework, pursuing the objectives of growth (economic efﬁciency), stability (prevention of economic crises), and (social) cohesion. The policy instruments available to pursue these objectives are the same in the EU system as in all other economic systems: market liberalization and

Welfare Policy Integration Inconsistencies

97

regulatory policies, tax and spending programs, and the stance of monetary and ﬁscal macroeconomic policies. In practice the degree to which policy objectives have been achieved is not uniform over time and across policy ﬁelds. In theory, as discussed in more detail by Sapir et al. (2004), it is useful to organize a discussion of unsatisfactory policy outcomes in terms of imperfect consistency between the scope of policies and that of decision-making processes across policy effects and policy decisions, either at a point in time, or over time. Dismantling barriers to trade and factor mobility, and organizing a common economic policy framework, not only served the purpose of preventing confrontation and fostering co-operation among European nations. Like the common competition and industrial policies that foster competition and efﬁciency, market integration was and is valuable on purely economic grounds – very much as it was 250 years ago,when Adam Smith noted that The statute of apprenticeship obstructs the free circulation of labor from one employment to another, even in the same place. The exclusive privileges of corporations obstruct it from one place to another, even in the same employment. [y] Whatever obstructs the free circulation of labor from one employment to another, obstructs that of [capital] likewise; the quantity of [capital] which can be employed in any kind of business depending very much on that of the labor which can be employed in it,

and advocated removal of all such barriers. Removing barriers to voluntary exchanges improves the efﬁciency of the integrated economy’s pattern of production, and generally yields economic gains (see, e.g., Bean et al., 1998, and references therein). Opportunities for trade and factor mobility may reﬂect different factor endowments, and the pattern of trade and factor mobility within the EU has been increasingly driven by this endowment-based comparative advantage (rather than by economies of scale at the intra-industry level) as membership has expanded beyond the original core of highly industrialized countries. To some extent, goods-market integration can substitute for factor mobility when either would be triggered by different factor endowments and comparative advantage. When absolute productivity differentials exist, reﬂecting not only natural endowments but also the different level of development of social institutions, migration is instead the natural outcome of integration. Policy instruments designed and implemented to reach a particular objective can have undesirable implications for other objectives, however. Economic integration may foster efﬁciency and growth, but an efﬁcient allocation of economic activity need not in general be equitable and politically acceptable ex post. Market pressure can be perceived to be unfair when markets are imperfect, and not all individuals need gain relative to the pre-integration situation. If the owners of relatively abundant factors in each country gain from economic integration more than the owners of relatively scarce factors, and if per capita income levels reﬂect different endowments of factors (i.e., high-earning factors are more abundant in the relatively rich country), then integration tends to reduce income differentials in the poor country. For example, highly skilled labor may be scarce (and earn high wages) in a relatively poor country, but

98

Giuseppe Bertola

scarcity and incomes should decrease when the relatively abundant supply of skill factors in rich countries becomes available. Conversely, integration should tend to reduce the incomes of relatively poor individuals in rich countries, whose factors become more abundant and less valuable in the integrated economy. Hence, while aggregate inequality may well decline over the whole area, economic integration can have an adverse impact on inequality (or cohesion) within previously separate economies. The European continent has in the last century witnessed many wars, revolutions, hyperinﬂation, and excessive ﬂuctuations resulting in social and political crises. Hence, it is far from surprising that national and supranational policymakers should be concerned not only with economic growth but also with cohesion, in that divergent economic circumstance could foster social conﬂict; and with stability, in that excessive ﬂuctuations of economic, monetary, and ﬁscal circumstances could precipitate political crises. Stability and efﬁciency may be fostered by ﬁscal discipline at the same time as they prevent addressing the social welfare needs of certain groups of citizens, and there may be circumstances where the inability to run government deﬁcits hampers growth by making it more difﬁcult to reform market regulations and social welfare schemes. Since policies can reinforce or offset each other’s desirable effects, policy can lack coherence when it fails to achieve its objectives efﬁciently. Market liberalization may foster faster growth but may also fail to do so if other policies (such as property right enforcement) prevent liberalization from resulting in desirable outcomes. Fostering competition is akin to delivery of a public good. It should be, and is, chieﬂy entrusted to policymakers at the European level when economic interactions span all of the Union’s territory. In other policy areas, however, lack of coherence can result from different policy objectives across policymakers acting at different levels of government, or within different constituencies. The nature of policy failures resulting from improper ‘‘systems competition’’ (Sinn, 2003) is similar to the mechanism underlying undesirable outcomes of laissez faire interactions among individual economic agents. Just like imperfect factor and good markets can fail to appropriately balance the objectives of economic agents with conﬂicting objectives, so political interactions between collective decision-makers whose objectives differ can result in undesirable policy conﬁgurations. Conﬂicts of interests between policymaking entities may lead to attempts to undo the effects of policies implemented at higher or lower levels. The resulting situation can easily be worse than laissez faire in much the same way non-competitive markets can damage economic efﬁciency when individual economic agents are in a position to exploit their market power in pursuit of their own economic welfare. Lack of coherence can also be brought about by more subtle failures of policy coordination. Even policymakers who share an ultimately common view of what would constitute desirable outcomes may fail to take appropriate action. When implementing policies that have effects beyond their immediate constituency, each may rely on others to implement costly actions in pursuit of a

Welfare Policy Integration Inconsistencies

99

common good, and inaction may result even when all share similar views on appropriate actions. Again, the nature of the relevant policy failure is similar to what may be observed when interactions at the individual level fail to address public-good aspects appropriately: just like individuals cannot be expected to spontaneously pay taxes in the absence of collective enforcement, so policymakers cannot be expected to implement the tax, subsidy, and regulation policies that would be optimal from the viewpoint of a large integrated area when their constituency is smaller than the scope of those policies. In both cases, the failure to see policy tradeoffs in their entirety can imply that policy implementation fails to address them appropriately (Sinn, 2003). When this happens, poor coordination results in outcomes that are unsatisfactory from the constituents’ and policymakers’ point of view. Just as market interactions can fail to support efﬁcient outcomes when some markets fail to exist or function properly, so imperfect coordination of (for example) state aid to industry can fail to foster efﬁciency. Political choice and implementation processes may consistently fail to take into account their own ultimate consequences over time. It is again possible to identify a counterpart in laissez faire markets for the resulting policy failure: rather than because of static externalities, markets may ﬁnd it difﬁcult to achieve an efﬁcient allocation because of dynamic market failures of the type that has been identiﬁed as very important by recent theoretical and empirical perspectives on growth and development. Coordination of individual innovation efforts and orderly exchanges in the factor and product markets can be fostered by an appropriately sustainable policymaking environment, able to deliver appropriate incentives to entrepreneurship, low and predictable interest rates, and public institutions conducive to enforcement of property rights and the rule of law. Limiting local constituencies’ ability to run deﬁcits and issue debt fosters ﬁnancial stability: not only by preventing policymakers from disregarding their own actions’ consequences for future citizens and policymakers but also, in a context of actual or potential mobility of individuals and factors of production, by preventing debt burdens from eroding each constituency’s tax base, as local residents may walk away from ‘‘public’’ debt when taxation would be needed to repay it. Accordingly, in the USd (where Alexander Hamilton engineered a Federal bailout of state-level debts in the aftermath of the Revolution) many States have balanced-budget rules. In Europe’s Economic and Monetary Union (EMU), a Stability and Growth Path in principle imposes commonly agreed limits to national ﬁscal policies. Macroeconomic policies can not only stabilize an economy in the face of imperfectly coordinated savings and investment decisions and imperfectly ﬂexible price and wage arrangements, but also generate and propagate aggregate shocks if used in pursuit of objectives different from macroeconomic stability, and precipitate crises if implementation is unsustainable. Fiscal policy is a particularly delicate instrument: it can have favorable stabilization effects through changes (rather than high levels) of ﬁscal deﬁcits.

100

Giuseppe Bertola

Policymaking processes may be unable to change the stance of budgets quickly enough to stabilize business cycles, and may instead tend to fall prey to policymakers’ short-horizon electoral incentives (see, e.g., Gali and Perotti, 2003, for references and evidence). Coherence over time is a very desirable but elusive feature of economic policies everywhere, and it may be particularly difﬁcult for the EU’s young and evolving institutional framework to adopt an appropriately long horizon in its policy implementation and institutional reform processes. An obvious example of incoherence over time in the EU policy framework was that between monetary, exchange rate, and capital control policies (Padoa Schioppa, 1994). Fixing exchange rates in the absence of appropriate supporting policies or institutions, i.e., under independent monetary policies and free capital mobility, led to frequent realignments in the Exchange Rate Mechanism (ERM), and ultimately to unsustainable instability. Adoption of the euro, and agreement to delegate to a European institution the common monetary policy, removed an important source of policy inconsistency across the integrated economic area. 4. Social policy and the EU The economic and social objectives mentioned at the beginning of this paper are pursued by the EU in a context of a rapidly changing political and economic environment (Sapir et al., 2004). In the recent past, change has been driven by technological change, globalization, German reuniﬁcation, as well as by the process of European integration itself. The advantages of wider economic opportunities are obvious to economists, but disadvantages of integration are equally apparent when policies and institutions are not adapted so as to best exploit the new opportunities. As economies integrate and markets (especially ﬁnancial markets) develop both within and across national borders, reforms of collective instruments need to address common problems within a common policy framework, and require coordination of policies that, when implemented locally, have important spillovers on other elements of the integrated economic system. These problems are notuncommon: the same challenges and opportunities arose in previous integration experiences, such as that of villages into kingdoms and empires, or of regional States into national entities. European institutions have developed that address several important policy aspects at the level of the whole integrated economic area. Single market and competition policy target a common good by preventing uncoordinated policy interventions. Adoption of a common currency has also eliminated temptation for member countries of EMU to pursue opportunities to damage each other through uncoordinated expansionary policies and competitive devaluations. To the extent that the performance of European economies is not uniformly satisfactory, however, it is necessary to seek the sources of political and economic unhappiness in inappropriate allocation of decisions and/or in a political process that does not correctly address the relevant trade-offs.

Welfare Policy Integration Inconsistencies

101

Coherence across jurisdictions and decision-makers is not surprisingly difﬁcult to achieve in the EU system of economic policies. The EU is a multi-tier government system, where each of the three policy-instrument-sets mentioned in Section 1 have both Community and national dimensions. The single market – encompassing goods, services and capital – is a Community program, but labor market regulation as well as many ﬂanking initiatives rest with the national authorities and are subject to different degrees of coordination. The income redistribution function is spread across the EU level and the national level, the former dealing mainly with inter-regional and inter-countries cohesion, the latter with inter-personal cohesion. And tools of macroeconomic stabilization are also spread across the two levels: monetary policy is centralized for countries belonging to the euro area, but ﬁscal discipline is decentralized to the national level, subject to Community rules. 4.1. Principles and policies Evidence of social-policy issues’ political importance is given by their prominent status in the draft EU Constitution, cited in the Introduction; earlier Treaties also gave at least lip service to social cohesion goals, which were ofﬁcially incorporated in the 1990s.1 And evidence of their controversial nature is, of course, given by that document’s difﬁculties at the approval and, now, ratiﬁcation stage. The fundamental rights include the right not to be unjustly dismissed; the right to working conditions that respect the worker’s ‘‘health, safety and dignity’’; the right to receive state beneﬁts for unemployment, sickness and old age (without reference to cost); and the right to strike. The broad objectives of EU treaties and the targets set by the Lisbon European Council appear quite wishful. EU-level institutions have successfully dismantling barriers to free and efﬁcient mobility of goods, services, and factors of production, thus indeed creating an area without economic frontiers. In the area of social policy, such a process of ‘‘negative integration’’ tends to enforce deregulation whenever existing policies conﬂict with desirable economic efﬁciency. Even though the scope of economic interactions spans across the national boundaries of the economically integrated EU, the social and labor market policy action advocated by the Treaties is almost completely subsidiary, is left to intergovernmental negotiation and subject to explicit unanimity requirements. Thus, ofﬁcial EU documents hopefully envision desirable socialpolicy convergence and coordination as the automatic result of European

1 See Bean et al. (1998) for a detailed discussion of early and largely ineffectual social concerns in European-level legislation. The 1960 Treaty of Rome mentioned among its goals improved working conditions, ‘‘so as to make possible their harmonization,’’ especially with regard to equal pay for equal work for men and women and paid holiday schemes. Some relevant directives were issued in the 1970s, and the Social Chapter, after initial disagreement by the UK, was incorporated in the Treaty at Amsterdam in 1997.

102

Giuseppe Bertola

countries sharing a common social model faced by common challenges at each national level, rather than of a process of ‘‘positive integration’’ through explicit collective agreement. In practice, an extremely small portion of EU-level policy activity is explicitly devoted to cohesion objectives. Such lack of European-level action may appear surprising in light of the fact that the issues are broadly similar to those more or less successfully targeted by the EU – from the prohibition of state aid and enforcement of single market rules, to adoption of a single currency, to agreement on ﬁscal policy constraints. Reform and possible harmonization of tax and subsidy instruments are even more clearly necessary on theoretical grounds, as such instruments pursue goals whose importance depends on the overall economic environment and their effects depend on elasticities that are sure to change when economic interactions span different and wider environments (Bertola and Boeri, 2002). By opening up opportunities to react to taxation through mobility as well as through reduction of labor supply, integration with poor countries threatens rich countries’ welfare systems. For example, implementation of single market public procurement rules implied that much East German construction activity was performed by British, Portuguese, and Italian ﬁrms posting workers to German construction sites, at the same time as many German construction workers were able to draw unemployment beneﬁts of Bismarckian generosity (Bean et al., 1998). Clearly, the German system of employment-based social insurance was not designed to cope with the new types of labor market risk generated by economic integration. Equally clearly, integration grants both rich and poor countries new trade and specialization opportunities: rebuilding East German infrastructure would have been much more expensive had local labor only been employed. Thus, there is an obvious rationale for ﬁnancial instruments that direct towards poor countries some of the resources that would otherwise be paid in rich countries (in the form of, for example, unemployment beneﬁts for local workers displaced by immigrants). Some such instruments are already in place. Not surprisingly, cohesion-oriented instruments were introduced in synchrony with the design and implementation of Single Market policies (for a more detailed discussion and further references, see Sapir et al., 2004, Chapter 4.3). In the 1980s, enlargement of the EU to relatively poor and peripheral countries prompted the implementation of a complex set of supranational cohesion-oriented policies, aimed at fostering income equality across Europe’s geographical units. The existing instruments do not simply transfer resources to low-income regions: disbursement of funds is conditional on industrial structure, peripherality, and other regional characteristics as well as on income levels, and EU-level policies aimed at fostering cohesion transfer funds with important strings attached. The Treaty itself refers to a reduction in ‘‘disparities between the levels of development’’, not in disposable income levels. Indeed, agreement on unconditional transfers to poor regions would hardly be politically feasible: transparent transfers of funds across the

Welfare Policy Integration Inconsistencies

103

borders of member states’ redistribution schemes would create obvious winners and losers and would be difﬁcult to agree upon at the Community level. But the instruments are not obviously coherent with other objectives. For example, regional and state-aid policies are often designed to work against the structural adjustments, based on comparative advantage, called for by economic integration processes (Midelfart-Knarvik and Overman, 2002) – thus reducing efﬁciency more than would be necessary to achieve a given level of cohesion. And political tensions do surface despite the current scheme’s attempts to link disbursements to objective criteria rather than to income levels. Almost all member states, no matter how rich, claim at least some underdeveloped regions, and region-oriented spending is mired in negotiations involving other ﬁnancial ﬂows, such as those implemented through the Common Agricultural Policy. A paramount concern of national representatives in the relevant negotiations is that of ‘‘getting their money back.’’ Such concerns are politically understandable, as transfers are difﬁcult to accept for the electorates of net-contributor member states. By largely neglecting income per capita considerations, however, political interactions not only fail to embody the notion of a common European concern with poverty, but also leave unaddressed the need to avoid undesirable policy spillovers. Policy can be far from coherent and lack efﬁciency when decision-making is decentralized, policymaking authorities are accountable to constituencies that are small compared with the scope of economic interaction across a large integrated area, and each policymaker focuses on a small portion of aggregate trade-offs or relies on others to provide common goods. In the resulting system of economic and political interactions, it is difﬁcult to address market failures by appropriate tax-and-subsidy or regulation policies in the social and labor market area, as well as in ﬁnancial and other markets. The difﬁculties of European policy interventions on labor income determination and redistribution derive from two interrelated problems (Bertola et al., 2001). First, there are many reasons why welfare systems designed decades ago need to be redesigned in light of new demographic and technological trends, and of changes in the structure of market and non-market economic interactions. But while the relevant changes are broadly similar in all industrialized economies, their impact and implications are quantitatively and qualitatively different across EU member countries’ diverse conﬁgurations of their systems of social protection, summarized by the distinction introduced above, between Scandinavian, AngloSaxon, Continental, and southern European welfare states. Budget problems are the most pressing cause of distress for the public-employment-based Scandinavian Welfare State. The UK (like many non-European Anglo-Saxon countries) faces increasing social exclusion in the form of permanent ‘‘working poor’’ status. The Continental countries are troubled by low employment rates. Last, but not least, the southern European countries also ﬁnd it increasingly difﬁcult to target poverty as their family- and pension-oriented social policies are challenged by demographic and labor-market trends.

104

Giuseppe Bertola

Second, and as a result, the EU policymaking framework in the relevant area is much less well developed than (for example) in the area of monetary and ﬁscal policy. This can of course be a source of policy inconsistency. Just like uncoordinated macroeconomic policies, ﬁxed exchange rate, and free trade with capital mobility before EMU, so free mobility of goods and/or factors, local decision-making powers in the labor-market and social protection area, and social inclusion coexist uneasily (Bertola, 2004). And, as in that case, the relevant issues are most clearly seen in terms of an ‘‘inconsistent trio’’ of policy characteristics and goals:pursuing two of the three to the limit necessarily implies forsaking the third (see Padoa-Schioppa (1994) for such reasoning in the monetary and ﬁscal policy area). 4.2. Incoherence Consider ﬁrst the implications of completely unrestrained economic competition among constituencies with completely independent social policymaking authority. This situation cannot foster social inclusion: economic competition leads local constituencies to forgo social objectives, and social policies are theoretically predicted to enter a ‘‘race to the bottom’’ downward spiral. The pursuit of equity always needs to be traded off lower economic efﬁciency. From the point of view of local policymakers the trade-off is clearly worse when more generous subsidies and higher tax rates lead to relocation of production rather than lower labor supply, and economic integration makes it easier for individuals to opt out of supposedly mandatory redistributive schemes. Forgoing the ‘‘protection’’ afforded by barriers to trade and labor mobility reduces the effectiveness of social policies and increases their cost, and theory predicts that, for each decision-maker, preservation of economic competitiveness prevails on pursuit of social inclusion, uncoordinated policy choices by local constituencies should trigger race-to-the-bottom tensions. Thus, the EU can have complete economic integration and subsidiary social policies, but only by accepting much less generous social protection. The far from satisfactory performance of many member countries’ social and labor market policies may lead some to favor such an outcome, but the resulting caricature of the US system (local social policies without any Federal competence in the area) is not politically feasible. Economic integration and subsidiary policy are not likely to be the chosen members of the inconsistent trio at the cost of social exclusion. In most European countries, social policies are politically stable and resistant to reform. Such stability is often read as evidence that ‘‘race to the bottom’’ fears are unjustiﬁed. But it can also be read as evidence that the extent of economic integration (especially as regards labor mobility) is not yet as full as would be necessary in order to reap its full specialization and ﬂexibility beneﬁts. To some extent, the pre-EMU situation of strongly limited international economic competition and personal mobility and strong national redistributive and regulatory policy (which in terms of the trio forsakes competition,

Welfare Policy Integration Inconsistencies

105

but preserves local decision-making powers and local social inclusion) is closer to the political consensus of European countries, and much of the resistance encountered by dismantlement of international economic barriers appears dangerously related to such social and political feelings. The third possible extreme conﬁguration of the inconsistent trio, namely unrestrained competition and effective social policy intervention at the cost of local decision-making power in the area, also presents obvious problems. It would require EU-level competence in social and labor market policy but, given the high degree of heterogeneity in the status quo levels of economic development, would very much be in danger of replicating at the continental level the current conﬁguration of large and heterogeneous member countries such as Italy, Germany, or Spain, where homogeneous national institutions tend to reduce the intensity of interregional and inter-occupational competition at the cost of generating pockets of persistently high unemployment in the relatively less developed areas (Sinn and Ochel, 2003; Bertola, 2004, and references therein). And, as in those countries, it would require substantial ﬁscal transfers from the richer regions, whose political sustainability is very much dubious in the absence of pan-European solidarity feelings – but can hardly be ruled out completely, to the extent that national political decision processes do often appear to prefer protection to economic efﬁciency at substantial economic and ﬁscal costs. 5. Enlargement, delays, and reforms An incoherent approach to policy issues in an integrating economic system faces two dangers: ﬁrst, that of slow and limited economic integration with less than optimal exploitation of the resulting economic gains. Second, that of imperfect policy implementation along other very relevant dimensions – such as welfare provision, as the national systems of redistribution are challenged by new sources of labor market risk. The issues outlined in the previous section are at least as old as Adam Smith’s writings. Historically, social policy has been implemented by nation-states with strong central governments, and has encountered difﬁculties when diverse levels of development and imperfectly integrated labor markets happened to coexist within such States. Conﬂicts between integration and redistribution remain important within the larger members of the EU, in particular Italy, Germany, and Spain. As regards international integration, such problems certainly become more visible and important as the EU expands to encompass increasingly diverse countries. Absolute productivity differentials are very wide across western and central/ eastern European economies, reﬂecting not only natural endowments but also the different level of development of social institutions, particularly as regards security arrangements, political stability, and the enforcement of property rights. Hence, migration is a natural outcome. The new members acceding to the EU in

106

Giuseppe Bertola

the current enlargement, however, are economically small compared with the existing integrated area. Hence, not only trade ﬂows but also migration trends from the central and eastern portions of the EU are likely to be relatively small (Boeri, Bru¨cker and others, 2001) and probably of less consequence, at the aggregate level, than existing migration pressure from northern Africa and other non-EU neighboring regions. Concentration of immigration in certain industry, regions, and population segments can have important consequences in interaction with social policies, however. While trade among the core EU members may be largely based on economies of scale and intra-industry in character, trade, and especially, migration ﬂows with the Central and Eastern European Countries (CEEC) should at least in part reﬂect differential endowments of different factors. Hence, the distributional effects of CEEC accession (if any) within older EU members can be expected to be negative, i.e., income inequality can be expected to increase in western Europe. Further issues are raised by legislation that, like the recent Directive on Free Movement (2004/38/EC), grants rights of residence and potential access to welfare assistance to all EU citizens. In many previous trade and labor market integration experiences, the effects of actual or potential migration in relatively rich labor markets were largely avoided by explicit or implicit subsidization of unemployment in relatively poor labor markets. But it is hard to envision that EU transfers could, or indeed, would be used for this purpose to cope with its Eastern enlargement. First, the size of the EU supranational budget is tiny relative to the size of the transfers effected, for example, in the German uniﬁcation episode. Second, and more importantly, the income differentials between the EU-15 and the CEECs are much wider than those featured by other European integration episodes. Anything resembling the policies enacted in East Germany could not be effected even if Poland were to be integrated fully within a federal ﬁscal budget, let alone in the current institutional structure of the EU (where ﬁscal policy is essentially subsidiary). Migration to older EU members of workers from the relatively poor CEECs may therefore put important stress on social policy arrangements. As shown by the theoretical model of Wellish and Wildasin (1996), external immigration affects the strategic interaction between decentralized redistributive policies, such as those prevalent in the EU institutional arrangements. Depending on whether immigrants are net ﬁscal contributors or recipients, local welfare policies may attempt to attract or repel them; hence the very existence of potential immigration magniﬁes the adverse welfare effects through coordination failures. While the relatively limited income differentials and substantial cultural differences among western European EU members do not excite much migration, the incentives of CEEC immigrants to ‘‘beneﬁt shop’’ once they have been uprooted by the much larger income differential between their country and any of the richest Western labor market may potentially disrupt the current state of affairs, and multiply the difﬁculties of local social-welfare systems within the EU.

Welfare Policy Integration Inconsistencies

107

As further discussed below, a ius loci basis for ﬁscal treatment of economic interactions is economically naı¨ ve. Trade unavoidably spans the boundaries of different jurisdictions, and persons do not need to move to make a difference in the age of the Internet. But even though foreigners need not be present to affect the sustainability of welfare state arrangements, fears of immigration pressure are widespread. In the run-up to Eastern enlargement, the increasingly apparent inconsistency of the current EU social and labor market policy framework has spurred interest in previously neglected issues. And the reaction of heterogeneous countries to the possibility of worker inﬂows from the new members conforms nicely to the same countries’ heterogeneous characteristics in terms not only of geographic position, but also of status quo welfare-state conﬁguration. The Nordic and Anglo-Saxon EU countries (with the exception of Ireland, at the time of writing) have chosen to restrict only immigrant workers to their citizenbased welfare beneﬁts. Austria and Germany have chosen to restrict labor market access for the longest allowed period, and this is of course unsurprising in light not only of these countries’ geographic location but also of their labormarket based approach to welfare issues. 5.1. Integration, delayed? When social policy instruments are decentralized in an integrated economic system, unfettered access to local social policy subsidies (on a residence or employment basis) fosters mobility but reduces local incentives to provide welfare beneﬁts. Conversely, exclusion from welfare beneﬁts of individuals hailing from other constituencies, on an origin principle basis, improves the sustainability of local welfare schemes, but at the cost of reducing mobility incentives and providing doubtful support to migrants.2 Since there are advantages and disadvantages to both extreme conﬁgurations, Richter (2002) sensibly suggests that an intermediate solution may be desirable, perhaps in the form of a time period (longer than zero, and shorter than inﬁnity) to elapse before a migrant’s welfare-system membership shifts from the original to the destination constituency. This suggestion is debatable in

2

The origin principle is still adopted within Switzerland (Art. 48 of the Swiss Constitution stipulates: Needy persons shall be assisted by the Canton in which they are living. The cost of this assistance shall be borne by their canton of domicile, and the Confederation can order that recourse be had to a previous canton of domicile or the canton of origin.). That country did not develop into a nation as regards political organization (and wars of conquest), and remains rather medieval as regards welfare provisions, which play a residual and not very effective role in its society. The Swiss economy’s personal mobility requirements have been fulﬁlled by (temporary) migration from other countries, and its system of welfare provision, like other Swiss peculiarities, is better viewed as an exception in the context of worldwide integration trends, not as a model for broader economic areas: only one Switzerland can exist in the heart of a warring and then integrating Europe, and it would be moot to argue that all European countries try and imitate its ﬁnancial, trade, and personal mobility policies.

108

Giuseppe Bertola

terms of the dimension (time) along which an intermediate solution should be sought– in terms of the trigger (migration) for the switch, and in terms of ease of implementation. Choosing time as the dimension along which extremes should be bridged can of course be justiﬁed if advantages and drawbacks of extreme solutions materialize with different delays. Sinn and Ochel (2003) base such an argument on the notion that migration may be driven by misguided welfare-shopping incentives in the short run, while in the longer run capital ﬂows would support income levels in initially poor countries and stem migration. Theoretically, however, capital or labor mobility can play the same role as trade in goods in a wide variety of circumstances, and both should be responsive to market and policy circumstances on a forward-looking basis when adjustment is costly. Empirically, a piece of evidence brought in support of delayed integration is the dynamics experienced in the German uniﬁcation episode (Sinn and Ochel, 2003, footnote 7). In that case, nearly all migration occurred in the ﬁrst two years after uniﬁcation, while capital ﬂows were smaller and persistent. That single observation, however, is very special along both the cross-country and time-series dimensions. Migration (not only for economic reasons) was certainly a more natural option for reunited East Germans than for citizens of truly different countries, who face much higher language and cultural barriers. And migration did not cease because of the mere passage of time but because of public policies meant to stop it. Public transfers from western Germany were at times as high as 80% of East Germany’s GDP, and the convergence of wage rates (if not of employment opportunities) was accelerated by extension of national wage agreements rather than by productivity growth. Large subsidies to poor integrating regions were in this case the other side of the coin of the adoption of a single regulatory framework. As shown by the experience of the Italian Mezzogiorno over a long period, and so far conﬁrmed by the experience of eastern Germany after uniﬁcation, this type of integration results in slow growth, longterm unemployment and inactivity in the poor regions (See Boeri and others, 2002, and references therein). And while intermediate solutions are indeed likely to be better than extremes, the correct mix should depend on the relative prominence of pros and cons in a variety of respects. Focusing on migration-related aspects also has some elements of extremism. Migration is more visible and politically sensitive than other symptoms and means of economic integration. However, it is neither the most important nor the most natural convergence vehicle, and it is not well predicted (in light of different cost of living and system productivities) by the initial difference between Eastern and Western labor costs. Immigration is intense from non-EU areas, and accession of CEECs to the EU may increase migration pressures. Product market integration, via factor-price equalization, can have the same effects unless labor mobility (rather than product mobility) is necessary, as in the personal services sector. The case of construction workers posted to East Germany mentioned above remains a very clear instance of tension

Welfare Policy Integration Inconsistencies

109

between national welfare systems and economic integration. Building has to be performed on-site, hence (temporary) migration of construction workers made the tension more visible than in the case of, for example, relocation of automotive plants. But posting of construction workers – where a major factor of change was the single market provision of EU-wide bidding for public projects – illustrates that immigrants need not draw beneﬁts themselves to trigger important consequences: when they compete with low-skill workers and the latter’s wages fall below the non-work beneﬁts, substitution of the indigenous poor by foreigners is effectively subsidized by the taxpayers of more generous constituencies. An excessive focus on immigrants’ visible welfare beneﬁts also risks disregarding important elements of the broader integration-and-social-policy picture. Some migration may of course be driven by welfare-seeking motives.3 But welfare beneﬁts can hardly be the only or even the main motive for migrants, and any impact on welfare provision is certainly not the only consequence for the economies of host countries – economies that can clearly proﬁt from immigrants’ willingness and ability to perform tasks different from those of natives, and at lower cost (as in the German Public Works case) to local consumers and taxpayers. Economic development relies on specialization and gains from trade; hence diversity is a productive asset for an economy. Like trade, migration is beneﬁcial when it occurs across the boundaries of diverse regions, and can be expected to beneﬁt typical (but not all) individuals in most cases. Ethnic diversity is a potential obstacle to adoption of beneﬁcial policies when it is a source of conﬂict but – other things being equal – does appear to foster economic productivity (see Alesina and La Ferrara, 2003, for references and evidence). Consider next ease of implementation and labor mobility in a system whereby migrants remain entitled to beneﬁts (and liable to taxes) in their original, independently managed constituency. As pointed out by Poutvaara (2002), decisions and negotiations regarding the length of delay are no less difﬁcult than those regarding rates of contribution and other aspects of harmonization or, perhaps, immigration quotas meant to balance the advantages of economic integration and the disadvantages of welfare-system externalities. And if an agreement could be reached on such rules, administration of a system of constrained access to welfare policy instruments would not be easy. It is interesting

3

In the US in 1969, the Supreme Court’s Shapiro decision ruled that state-level welfare beneﬁts could not be conditioned on previous residence. This exposed states offering generous beneﬁts to the danger of attracting welfare recipients. Not surprisingly, some welfare-motivated migration by single mothers and non-US citizens can be detected: Meyer (2000) offers a careful empirical analysis and a review of the extensive literature on beneﬁt-induced migration in the US. Of course, each state of the US ﬁnds it difﬁcult to implement generous welfare support systems, and this ‘‘welfare magnet’’ effect may explain why that country’s social policies are mostly co-funded and implemented at the Federal level.

Giuseppe Bertola

110

to revisit the 18th-century situation analyzed by Adam Smith and referred to above, where [yThe obstruction given to the free circulation of labour] by the poor laws [y] consist[ed] in the difﬁculty which a poor man ﬁnds in obtaining a settlement, or even in being allowed to exercise his industry in any parish but that to which he belongs.

Entitlement issues were then addressed by a complex system of written notiﬁcation rules and certiﬁcation requirements: In order to restore in some measure that free circulation of labour which those different statutes had almost entirely taken away, the invention of certiﬁcates was fallen upon. By the 8th and 9th of William III it was enacted, that if any person should bring a certiﬁcate from the parish where he was last legally settled, subscribed by the churchwardens and overseers of the poor, and allowed by two justices of the peace, that every other parish should be obliged to receive him; that he should not be removeable merely upon account of his being likely to become chargeable, but only upon his becoming actually chargeable, and that then the parish which granted the certiﬁcate should be obliged to pay the expence both of his maintenance and of his removal.

This system was remarkably similar to a possible conﬁguration of ‘‘delayed integration,’’ and to the several German–Turkey treaties whereby (temporary) workers remained attached to the origin country’s social security systems. In 18th century England, a formal origin-principle system failed to foster labor mobility: in equilibrium, certiﬁcates ought always to be required by the parish where any poor man comes to reside, andy ought very seldom to be granted by that which he proposes to leave

and, as a result, The very unequal price of labor which we frequently ﬁnd in England in places at no great distance from one another, is probably owing to the obstruction which the law of settlements gives to a poor man who would carry his industry from one parish to another without a certiﬁcate.

Lack of agreement regarding welfare provision as a common good in support of economic integration condemned 18th century England, and may condemn the EU, to forgo important mutual gains from factor reallocation. 5.2. Redesigning and modernizing welfare delivery Is the EU doomed to repeat English history, or can historical experience help it devise a coherent policy framework and reconcile integration and welfare provision? As trade already puts important pressure on existing policy arrangements, and as both migration and capital ﬂows are driven by forward-looking decisions, a coherent and sustainable policy framework is needed to foster migration that has beneﬁcial effects for both the source country and the destination

Welfare Policy Integration Inconsistencies

111

country, and discourage migration driven by exploitation of policies that have different purposes. Such a framework is that proposed by Bertola et al. (2001). Rather than accepting unquestioningly the assumption that all welfare provision must be local (or ‘‘subsidiary’’) in an integrated economy, appropriate reform of welfare systems should recognize that harmonization and coordination can support a desirable process of economic and social integration that is advantageous to all parties involved. To imagine how this could be achieved, it is useful to start from the distinction, introduced in Section 2, between social policies motivated by market failures in the face of labor market risks, and social policies motivated by solidarity and prevention of social unrest. This distinction is not always clear in actual policy frameworks, but is crucial in theory to devise appropriate resolution of issues in the relevant area: the potential tensions between economic integration and the provision of adequate social protection are different for the two sets of policy goals, which therefore call for differently conﬁgured coordination and harmonization processes at the EU level. Policies meant to implement quasi-market redistribution of ex post income across workers are of course very much needed within each country, especially in the presence of underdeveloped and inefﬁcient market provision of insurance and savings vehicles: Bertola and Koeniger (2004) and their references show that differences across countries in the efﬁciency of legal enforcement, and the development of ﬁnancial markets, do appear theoretically and empirically relevant to the desirability and implementation of redistribution and labor market regulation. As better ﬁnancial markets develop in Europe, it may well be possible to reduce government intervention in this ﬁeld. And to the extent that collective organization of insurance against life risks remains necessary, no race to the bottom need develop in an integrated economic area as regards policies meant to protect labor income against such risks. Social insurance is not subject to pressure from ﬁscal competition when there is a clear link between contributions and beneﬁts (Atkinson, 1998). Hence, collective schemes can coexist if they offer actuarially fair beneﬁts. Insurance against labor market risk may generally be more efﬁciently provided by collective schemes rather than private contracts: governments can enforce continued participation by ex post lucky individuals, and can in principle monitor and implement such schemes more efﬁciently than private providers. Just like competition among private insurers, however, so competition among collective systems needs to be appropriately regulated and monitored, so as to ensure that individual participants are adequately informed. EU-level institutions can potentially play an important role in certifying and clarifying to citizens the appropriateness and sustainability of the relevant schemes. Supranational authorities can play a useful role in testing and enforcing budget rules that ensure coherence over time of public ﬁnances; they can also play a

112

Giuseppe Bertola

similar role as regards pension, health, and unemployment insurance schemes. Hence, Quasi-market arrangements meant to provide insurance, such as unemploy-

ment beneﬁts and pension schemes, should not redistribute resources ex ante and should not be funded at the EU level, because economic incentives for factor mobility and allocation are best preserved when beneﬁts and contributions balance each other within appropriately deﬁned local markets. Participation in such schemes needs to be mandatory and comprehensive. Minimum contribution rates and standard conﬁgurations should be agreed upon centrally to eliminate opportunities for individuals and local constituencies to ‘‘opt out’’ (Bertola et al., 2001). In practice, social insurance schemes always entail ex ante as well as ex post redistribution. Targeting risk outcomes is difﬁcult if information is incomplete and asymmetric: just like other insurance, social insurance faces moral hazard problems and can hardly avoid actuarial unfairness. Moreover, the political determination of entitlements tends to build redistributive components into schemes that should be meant to insure participants; and to the extent that the distributional impact of pensions and unemployment beneﬁts reﬂects political power rather than need, it too often fails to improve the lot of truly poor individuals. Conceptually, however, insurance objectives should be pursued separately from solidarity-based social cohesion goals to the fullest extent possible. And to prevent inconsistencies in an integrated economic area, the instruments meant to target the latter goals do require coordination and harmonization, rather than only monitoring and certiﬁcation. Redistribution is motivated by goodwill and solidarity, as well as by a more selﬁsh desire to prevent crime and unrest and to maintain social peace. Neighbors are more immediately disturbing when poor, but as noted above, economic integration need not involve labor mobility to disrupt national welfare systems. As evidenced by the German integration experience, rich citizens often help poor ones in their own interest, to preserve the welfare and labor market structure they prefer while (hopefully) exploiting economic and social gains from integration with previously isolated regions. To ensure the long-run sustainability of social protection and free personal mobility within the EU, the social rights of European citizens should be deﬁned in the form of minimum welfare standards at the EU level (Atkinson, 1998; Bean et al., 1998; Bertola et al., 2001). Agreement should therefore be possible on an appropriately devised scheme whereby Solidarity-based transfers, guaranteeing a minimum welfare level, should be

coordinated at the central level to prevent competition among subsidiary levels of government from resulting in either unacceptably low levels of welfare provision, or in more or less implicit limits, to economic integration. Minimum-welfare transfers and services should be co-ﬁnanced by a speciﬁc

Welfare Policy Integration Inconsistencies

113

budget line item at the EU level. Central co-ﬁnancing of social assistance programs would also provide means for enforcement of EU-wide guidelines (Bertola et al., 2001). The wide income differentials among EU members (especially after enlargements) do imply that minimum-welfare transfers and services would need to be co-ﬁnanced at the EU level. If targeted narrowly on poverty, however, the relevant EU instruments could be much less important than those of the US Federal government, which does regulate and ﬁnance or co-ﬁnance the bulk of that country’s social policies (Bertola et al., 2001). Disbursement in each country of beneﬁts based on low-income status is large when the country’s population is poor on average and, for given country-speciﬁc average income, when incomes are widely dispersed within the country. Some relevant data are reported in Morrisson and Murtin (2003), and can be combined with income and population data so as to assess the orders of magnitude of a possible EU-wide welfare scheme (see the Appendix). Beneﬁts should be indexed to the cost of living, to prevent poverty and welfare-shopping incentives. Suppose in all countries a fund amounting to US$10 per day (corrected for purchasing power parity) was disbursed to citizens earning less than US$20 per day in 1998, the overall size of the relevant budget would be in the order of 1.3% of EU GDP, and cross-border redistributive effects would range between a maximum net contribution of 1.3% of GDP (from Luxembourg) and a maximum 2.8% of GDP net receipt (by Portugal). Both the overall size and redistributive component of a theoretical scheme targeting economic income appear broadly comparable with those of the current schemes based on residence in certain regions or activity in agriculture, which currently absorb a little more than a percentage point of GDP and generate net imbalances in EU budget contributions ranging from -0.45% (Germany) to 3.66% (Greece). Since roughly half of total EU funds is spent in the Common Agricultural Policy context, that obsolete scheme’s funding would appear to be adequate to eradicate the most extreme European poverty, without national intervention, if it were rerouted to that end. These are only very rough computations, of course, and do not take into account the implications (much more dramatic for the new, if not for the existing, members) of the recent enlargement to Central and Eastern economies with much lower average income levels. More detailed income distribution data could be used to compute more sophisticated estimates, and a cost-of-living adjusted US$10 beneﬁt distributed among individuals earning less than US$20 a day (in 1998) may or may not be an appropriate benchmark for EU-level poverty-prevention schemes. Minimal standards in theory could correspond to what the least generous countries would be willing to offer if they were forced to take account of spillovers across the boundaries of local constituencies, and could be determined in terms of the market value (at local prices) of a basket of goods and services, thereby preventing any ‘‘race to the bottom’’ tendency from undermining political support for EU integration. In practice, many poor

114

Giuseppe Bertola

individuals are currently cared for by member States: it would be hard to compare the labor-supply effects of a hypothetical European scheme to those of existing national schemes, which are very different in terms of overall ﬁnancial generosity (which largely reﬂects the countries’ different per capita income levels), and, especially, as regards their qualitative structure. In practice, the design and implementation of such a program would need to address many difﬁcult issues, chieﬂy that of appropriately balancing the undesirable welfare-shopping and labor-supply incentives of any poverty-prevention policy in a heterogeneous environment. In the long term, however, Europe must resolve these issues, because the removal of barriers to personal mobility over increasingly large areas, and the preservation of cohesion in the resulting new environment, will be and have always been key ingredients of its economic and social progress. Many of the same issues, in fact, already arise within nations, especially large and heterogeneous ones such as Germany and Italy. Within Europe, most nations are too small or too large to cope with current challenges: small countries can inﬂict externalities upon other countries, and large countries tend to impose uniform beneﬁts on their regions. Within and across countries, social policy standards do need to vary according to local circumstances and traditions, as they do across US States. And they also need to be ‘‘harmonized’’ for mobility to be possible and not distorted by welfare-seeking incentives. Minimum assistance levels should be speciﬁed on a relative basis, as a proportion of local (not necessarily national) average earnings, so as to make them politically acceptable and economically affordable in light of different development levels. They should also be adjusted to reﬂect local price levels: both criteria would imply lower beneﬁts in the less-developed regions of the EU, and any co-ﬁnancing at the EU level should be ﬁne-tuned so as to address cross-border spillovers while at the same time minimizing moralhazard effects, in the form of a tendency for individual constituencies – in the style of English parishes of old – to neglect their own needy citizens’ welfare and dump them on other systems of welfare provisions. A welfare ﬂoor preventing absolute poverty at the EU level can and will need to be designed and enforced, as economic interactions gradually achieve ‘‘ever closer Union’’ objectives, and an appropriately harmonized policy of this type should also reduce or eliminate incentives for welfare-motivated mobility by disadvantaged groups. A pan-European safety net would also contribute to easing social tensions if established as a European citizenship right, and provide a strong rationale for EU-wide immigration policies addressing the obvious coordination problems (Wellish and Wildasin, 1996) arising when local constituencies grant EU-wide citizenship entitlements. 6. Summary and prospects Market economies are unavoidably imperfect, and relying on competition among imperfect policymaking systems cannot yield better outcomes than a

Welfare Policy Integration Inconsistencies

115

well-informed policymaking framework addressing the relevant trade-offs coherently. To preserve both redistribution and economic integration, elements of social policy must be centralized or coordinated. In an integrated European economy, decision-making power in the social ﬁeld must not be exclusively national, for this would either make social policy ineffective or, more realistically, let calls for protection from foreign social dumping strengthen barriers to trade and factor mobility. Disappointment with the performance of European labor markets has been high and growing for more than 20 years. The US’s better employment performance and lower degree of interference with labor markets may of course reﬂect less concern for inequality, but even that economy’s well-developed ways of addressing income-distribution issues through ﬁnancial and product market efﬁciency fail to deliver equality to its citizens; in comparison to the EU the US has developed a much more integrated system of redistribution and welfare provision at the Federal level. Is Europe ready to reform its labor and other markets, maybe so as to make them more similar to their American counterparts? Elements of the reforms advocated in Section 5 are dimly apparent at the national level, where pension and labor reforms are increasingly oriented towards actuarial fairness and economic incentives; and at the EU level, where the Constitution (like previous Treaties) does envision coordination of minimum welfare standards – albeit still with very blunt tools, as the draft that may yet fail to be ratiﬁed explicitly excludes the relevant issues from majority voting and in many ways aims at minimizing interference with national schemes. Reforms of pension and unemployment insurance schemes in the direction of actuarial fairness are all the more desirable in an integrated economic area, and need not call for explicit coordination or co-ﬁnancing at the EU level. At the same time as such schemes are reformed to address demographic-trend problems, they should be tailored to the new integrated economic systems, and redesigned so as to encourage mobility driven by comparative advantage. They should also be monitored at the supranational level, and the component of social policies that implements redistribution should be explicit, coordinated, and co-ﬁnanced at the same level at which economic interactions take place. The EU’s ever closer integration of factor and product markets implies that uncoordinated national policies simply cannot achieve the redistribution it deems desirable in light of ﬁnancial market imperfections and political cohesion objectives. Directly or indirectly, via migration and capital mobility or via market interactions, tax bases react elastically to taxation, and beneﬁts are effectively paid to members of jurisdictions other than those legislating them. Social policy issues are more difﬁcult but arguably more urgent than those relevant to common defense and foreign policy initiatives. An entity that can decide milk production in each member country should presumably be able to inﬂuence and coordinate issues that may not be any easier to reach agreement on, but are certainly more important.

116

Giuseppe Bertola

Sapir et al. (2004) identify a number of such issues, and point out that only historical heritage explains the supranational (Community) status of the Common Agricultural Policy. Currently, allocation to the EU level of expenditures is not clearly targeted to common goods. This justiﬁes suspicions that all money allocated to the EU is wasted, suspicions that in turn motivate budget negotiation strategies that are narrowly and unproductively focused on ‘‘money back’’ concerns. Arguments voiced against the re-nationalization of agricultural policies suggested by Sapir et al. (2004) stressed possible likely distortions induced by competing policies in that area. But much the same arguments are valid for national subsidiary of tax and welfare policies, and the higher transparency of agricultural subsidies granted at the national level would presumably decrease or eliminate their political attractiveness in times of scarce ﬁscal revenues. Of course, the pitfalls of centralized policies must be avoided. Misguided application of homogeneous rules and tax systems to heterogeneous economic entities can stiﬂe economic development, as in Italy’s Mezzogiorno and Germany’s Eastern Landers. But the problems encountered by social and labor policies in these large and diverse countries are just another element of the case against preservation of national policies through, e.g., delayed-integration entitlement schemes. Welfare provision should be modernized within countries and harmonized across them: harmonization does not imply uniformity, and the alternative of forgoing integration is not practically possible (let alone desirable) in the modern world. Citizens’ social concerns have to be addressed at the same level where economic interactions take place, for choices made at lower levels too easily heed resentment against economic integration as a source of unfair competition and reduction of already inadequate protection. A coherent policy framework should address the shared goals of all EU countries. Nations, like parishes, are in some respects obsolete policymaking units. If mobility of people and goods over an increasingly large integrated area needs to be preserved as a crucial engine of economic development, both increasing attention to regional speciﬁcities and construction of a layer of European social policy are ultimately unavoidable. The latter already exists in the limited and imperfect shape of agricultural, regional, and structural funds in the EU budget, and of a relatively small but intrusive body of EU regulation of working time, worker rights, and other social-policy issues. Both are motivated by concern about the possible adverse effects of economic integration on distribution and on the governments’ willingness and ability to address them. That concern is politically strong in Europe, and it is not surprising to hear it expressed in the same treaties that extol the virtues of economic integration. What is missing, and needed, is a translation of that desire into a clear system of rules and adequate ﬁnancial resources addressing the relevant trade-offs coherently. EU-level ﬁnancial and regulatory interference with national social policy is far from negligible, but its opacity and complexity fail to generate goodwill: quite the opposite, each national government only sees incentives to retrieve funds that

Welfare Policy Integration Inconsistencies

117

are not perceived to address common concerns at the European level. Agreement on how to best address the relevant policy issues at the appropriate integrated level is of course not politically easy to achieve: Sapir et al. (2004) argue that European economic policy governance could and should be improved by a clearer distinction between objectives to be agreed politically, and implementation to be delegated to accountable supranational agencies (such as the European Central Bank) operating at the Community rather than the intergovernmental level. As mentioned in Section 5.2, similar principles can be readily applied to social and labor market policy instruments. In a reality that sees large nations unable to cope with their own regions’ economic heterogeneity, and all European policymakers concerned with possible cross-border spillover effects of local welfare policies, it is both theoretically and empirically difﬁcult to see why national competence on labor market and social policies should be preserved by limiting that personal mobility which, with other channels of economic integration, always was and remains the main engine of economic progress. The political and technical difﬁculties of implementing the appropriate reform of policymaking structures should not lead European citizens and policymakers to accept status quo passively. A clear understanding of the relevant static and dynamic trade-offs can lead policymaking processes to target their objectives more effectively, and offer a credible alternative to perhaps comfortingly stable, but eventually unsustainable, economic stagnation. Acknowledgments I am grateful for helpful comments by an anonymous referee and by the discussant, the editors, and other conference participants. References Alesina, A. and E. Glaeser (2004), Fighting Poverty in the US and Europe: A World of Difference, Oxford: Oxford University Press. Alesina, A. and E. La Ferrara (2003), ‘‘Ethnic diversity and economic performance’’, Working Paper. Atkinson, A.B. (1998), Poverty in Europe, Oxford: Blackwell. Bean, C., S. Bentolila, G. Bertola and J. Dolado (1998), Social Europe: One for All?, London: CEPR. Bertola, G. (2004), ‘‘National labor market institutions and the EU integration process’’, CESifo Economic Studies, Vol. 50, pp. 279–298. Bertola, G. and T. Boeri (2002), ‘‘EMU labor markets two years on: microeconomic tensions and institutional evolution’’, pp. 249–280 in: M. Buti and A. Sapir, editors, EMU and Economic Policy in Europe: The Challenge of the Early Years, Aldershot: Edward Elgar. Bertola, G., J.F. Jimeno, R. Marimon and C. Pissarides (2001), ‘‘Welfare systems and labor markets in Europe: what convergence before and after

118

Giuseppe Bertola

EMU?’’, in: G. Bertola, T. Boeri and G. Nicoletti, editors, Welfare and Employment in a United Europe, Cambridge, MA: MIT Press. Bertola, G. and W. Koeniger (2004), ‘‘Consumption smoothing and the structure of labor and credit markets’’, IZA DP No. 1052, Bonn. Boeri, T., H. Bru¨cker and others (2001), The Impact of Eastern Enlargement on Employment and Labor Markets in the EU Member States, Berlin and Milan: European Integration Consortium. Boeri, T. and others (2002), Who’s Afraid of the Big Enlargement? Economic and Social Implications of the European Union’s Prospective Eastern Expansion, London: CEPR. Esping-Andersen, G. (1990), The Three Worlds of Welfare Capitalism, Cambridge, UK: Polity Press. Galı´ , J. and R. Perotti (2003), ‘‘Fiscal policy and monetary integration in Europe’’, Economic Policy, Vol. 37, pp. 533–572. Hanson, G.H., K.F. Scheve, M.J. Slaughter and A. Spilimbergo (2002), ‘‘Immigration in the US economy’’, pp. 286–289 in: T. Boeri, G. Hanson and B. McCormick, editors, Immigration Policy and the Welfare System, Oxford: Oxford University Press. Meyer, B.D. (2000), ‘‘Do the poor move to receive higher welfare beneﬁts?’’, Northwestern University, Working Paper. Midelfart-Knarvik, K.H. and H.G. Overman (2002), ‘‘Delocation and European integration: is structural spending justiﬁed?’’, Economic Policy, Vol. 35, pp. 323–359. Morrisson, C. and F. Murtin (2003), ‘‘Inequality among Europeans (1970–2000)’’, Report for the Group of Policy Advisors, European Commission, http://www.crest.fr/pageperso/lmi/murtin/UEInequality.pdf. Padoa Schioppa, T. (1994), The Road to Monetary Union in Europe, Oxford: Clarendon Press. Poutvaara, P. (2002), ‘‘Comment on Wolfram F. Richter: social security and taxation of labor subject to subsidiarity and freedom of movement’’, Swedish Economic Policy Review, Vol. 9, pp. 75–78. Richter, W. (2002), ‘‘Social security and taxation of labor subject to subsidiarity and freedom of movement’’, Swedish Economic Policy Review, Vol. 9, pp. 47–74. Sapir, A., P. Aghion, G. Bertola, M. Hellwig, J. Pisani-Ferry, D. Rosati, J. Vin˜als and H. Wallaceothers (2004), An Agenda for a Growing Europe – The Sapir Report, Oxford: Oxford University Press. Sinn, H. (2003), The New Systems Competition, Oxford: Blackwell Publishing. Sinn, H. and W. Ochel (2003), ‘‘Social union, convergence and migration’’, Journal of Common Market Studies, Vol. 41(5), pp. 869–896. Wellish, D. and D.E. Wildasin (1996), ‘‘Decentralized income redistribution and immigration’’, European Economic Review, Vol. 40, pp. 187–217.

Appendix: Order of magnitude of an European Union poverty prevention scheme [B]

[C]

[D]

[E]

[F]

[G]

[H]

[I]

3.6 3.2 0.5 0.4 5 8.9 27.2 11.7 13.4 0.8 2.8 27.2 17.5 1.8 10.2

8.0 10.2 5.3 5.2 58.4 82.0 10.5 3.7 57.6 0.4 15.7 10.1 39.9 8.9 58.6 374.4

211,898 250,321 172,428 129,499 1,451,953 2,144,483 121,958 86,989 1,196,663 18,900 393,471 112,386 588,022 248,287 1,423,237 8,550,494

0.29 0.33 0.03 0.02 2.92 7.30 2.86 0.43 7.72 0.00 0.44 2.74 6.97 0.16 5.98 38.18

1.08 1.05 1.41 1.14 1.16 1.17 0.32 0.90 0.76 0.93 1.09 0.47 0.64 1.09 0.96

1133.27 1247.95 136.58 85.98 12406.89 31264.42 3335.62 1428.40 21333.06 11.65 1752.00 4719.74 16407.59 635.72 20834.67 116733.55

0.53% 0.50% 0.08% 0.07% 0.85% 1.46% 2.74% 1.64% 1.78% 0.06% 0.45% 4.20% 2.79% 0.26% 1.46% 1.36%

–0.83% –0.86% –1.28% –1.29% –0.51% 0.10% 1.38% 0.28% 0.42% –1.30% –0.91% 2.84% 1.43% –1.10% 0.10%

– 0.22% – 0.09% 0.15% 0.22% – 0.05% – 0.42% 3.66% 1.77% 0.11% – 0.28% – 0.39% 1.95% 0.91% – 0.45% – 0.19%

119

[A] % of population earning less than 20US$/day, 1998; Source: Morrisson and Murtin (2003). [B] Population (millions) in 1998; source: World Bank. [C] Gross Domestic Product (millions current US$) in 1998; Source: World Bank. [D] A*B ¼ Millions of citizens earning less than US$20/day. [E] Purchasing power parity conversion ratio, US ¼ 1.00; source: OECD. [F] C*D*E*365: cost (millions US$/year) of paying 10 PPP dollars per day to each of the citizens in A. [G] F/C: total paid in country as in B, % of GDP. [H] G-1.36%: Cross-border net ﬂow if G ﬁnanced in proportion to GDP. [I] Net contributions to EU operational (excl. administration) balance, 2000; source: Commission Services.

Welfare Policy Integration Inconsistencies

Austria Belgium Denmark Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain Sweden United Kingdom EU

[A]

This page intentionally left blank

Comment Laszlo Goerke

At what institutional level should social policy issues be decided upon in the enlarged European Union (EU) of 25 member states? This is the question tackled by Guiseppe Bertola. It is closely linked to the issue of the adequate extent of the welfare tate in a world in which, ﬁrst, not only labour market participants but also people from outside the labour force are becoming increasingly mobile and, second, income levels differ substantially across countries. The underlying problems are by no means new. As Bertola points out, quoting from Adam Smith’s Wealth of Nations, at least the consequences of an inadequate assignment of decision-making rights have already been discussed almost 250 years ago. To set the stage for the subsequent discussion, Bertola refers to three objectives enshrined in the Treaty Establishing the European Union that are relevant for his analysis. In Art. 2 it is stated: ‘‘The Community shall have as its task, y, to promote throughout the Community a harmonious, balanced and sustainable development of economic activities, a high level of employment and social protection, y, and economic and social cohesion y’’. This is to be achieved, inter alia, by establishing ‘‘an internal market characterised by the abolition, as between member states, of obstacles to the free movement of goods, persons, services and capital;’’ (Art. 3 1(c)). Finally, the ‘‘Community shall act within the limits of the powers conferred upon it by this Treaty and of the objectives assigned to it therein. In areas which do not fall within its exclusive competence, the Community shall take action, in accordance with the principle of subsidiarity, only if and

Corresponding author. E-mail address: [email protected] (L. Goerke). CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79012-5

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

122

Laszlo Goerke

insofar as the objectives of the proposed action cannot be sufﬁciently achieved by the member states and can therefore, by reason of the scale or effects of the proposed action, be better achieved by the Community’’ (Art. 5). Bertola interprets this selection of objectives as obligation for policymakers to promote social and economic cohesion and a high level of employment by establishing a world in which the free movement of persons is assured, taking into account the principle of subsidiarity. In Section 2, Bertola distinguishes two aims of the welfare state, namely the redistribution of income among individuals who may live in poverty without according transfers, and the provision of insurance for risks against which individuals cannot insure on a private market. He traces the development of institutions dealing with these objectives through history, and links their emergence to the occurrence of market-based production processes. In Section 3, Bertola outlines in general terms for the EU– with particular reference to principles discussed in the Saphir Report (Saphir et al., 2004, pp. 11ff) – how policies, which are decided upon at one institutional or geographical level but have broader consequences, distort economic incentives. This externality problem is illustrated for social policy issues in the EU in Section 4. Bertola argues convincingly that unhindered mobility, the application of the subsidiarity principle and social cohesion, in the sense of preventing poverty, ‘‘coexist uneasily’’ because pursuing any two of the three objectives necessarily implies that the third has to be forsaken. In Section 5, Bertola contrasts two approaches to overcoming this dilemma: the ﬁrst is the delayed integration methodology proposed, for example, by Richter (2003, 2004), Sinn and Ochel (2003), and Sinn et al. (2002). The second is Bertola’s own proposal, which focuses on the distinction between the insurance function of the welfare state and its redistributive objective. Bertola’s central idea is that, ﬁrst, mandatory, comprehensive and actuarially fair social insurance systems at, for example, the level of the individual state, can co-exist without forsaking any of the three objectives of free mobility, social protection and subsidiarity. This is, implicitly, argued to be the case because actuarially fair social insurance schemes do not affect the decision to migrate but can provide the level of insurance against income risks which individuals desire but the market will not offer. The second element of Bertola’s proposal responds to the hypothesis that the redistributive components of the welfare state provide incentives for migration which are not due to price or productivity differentials and, therefore, not desirable from an efﬁciency point of view. Accordingly, ‘‘To ensure the long-run sustainability of social protection and free personal mobility within the EU, the social rights of European citizens should be deﬁned in the form of minimum welfare standards at the EU level (p. 112)’’. Such minimum standards may be deﬁned in relative terms and may require a centralised European institution that could (partially) fund these transfers. While Bertola obviously points out the advantages of his proposal, a comprehensive comparison with alternative approaches may also be desirable. In

Comment

123

this comment I will attempt such a comparative evaluation. As has been made clear above, Bertola proposes to restrict the subsidiarity principle to the insurance of labour income risk and to allocate the decision on the amount of redistribution to the central European level. The gain for partially giving up the principle of subsidiarity is an equal treatment of all residents within a country. The delayed integration methodology, however, preserves subsidiarity to a much greater extent but may be argued to limit the extent of social protection, since it temporarily excludes immigrants from social protection measures in the country to which they have moved. The two approaches, hence, give different weights to the three objectives of free mobility, social protection and subsidiarity. The subsequent assessment also involves two additional alternatives, which Bertola does not look at in detail, namely the full and immediate inclusion of all immigrants in the welfare systems of their host country and their complete exclusion. Including the two approaches can help to show what is gained and lost by going for intermediate solutions, such as the one suggested by Bertola. Table 1 summarises the impact of different assignments of social policy competences. Rows 1 to 3 evaluate the delayed integration approach and Bertola’s proposal, as well as the two additional options of not restricting access to social protection measures for immigrants at all and of excluding them completely, with respect to the three EU objectives as deﬁned by Bertola. A plus sign indicates that the relevant objective is fulﬁlled, whereas a minus sign is awarded if a proposal does not warrant the respective objective or does so to a signiﬁcantly lesser degree than the other approaches. A sign has been bracketed to indicate that the evaluation is less positive (negative) than in the case of a ‘‘full’’ plus (minus) sign.

Table 1.

Comparative evaluation of alternative proposals

Objective

Approach Complete integration

Impediment to free mobility Social protection Redistribution Insurance Subsidiarity Preclusion of welfare migration Nondiscrimination

1

+

2a 2b 3 4

+ (+) +

5

+

Complete exclusion

( ) (+) + +

Delayed integration

Bertola

(+)

+

(+) (+) + (+)

(+)

( )

+

+

124

Laszlo Goerke

In particular, row 1 assesses whether the four options for assigning social policy competences restrict migration below the level that is desirable for efﬁciency reasons. Full integration into the welfare state implies that immigrants are insured against labour income risks in the same way as residents. Accordingly, migration, which is desirable from an efﬁciency perspective, is not prevented. Since both Bertola’s proposal and the complete integration of immigrants make no distinction between residents and immigrants with respect to insuring labour income risks, they certainly fare best on account of not restricting mobility for efﬁciency reasons. The delayed integration principle excludes immigrants from social insurance systems for a limited duration, such that immigrants will be less willing to accept risky jobs than they would be if an actuarially fair insurance were available. Hence, not all migration that is desirable from an efﬁciency perspective will take place. Since the exclusion is limited, the evaluation is still generally positive. The more limited the delayed inclusion is– for example, because primarily redistributive social protection measures are unavailable to immigrants for a limited duration (see Sinn and Ochel 2003, pp. 892f, for an according proposal) – the fewer and less important are impediments to efﬁcient migration. The complete exclusion of immigrants from social protection measures represents the strongest obstacle to free mobility of labour. Table 1 follows Bertola’s notion of differentiating between the redistribution objective of social protection and the social insurance motif. The evaluation of the various approaches with respect to the aim of providing insurance against labour income risk (row 2b) resembles to some extent that of impediments to free mobility of labour. An actuarially fair insurance fulﬁls the objective of insuring individuals against variations in their labour income due to shocks optimally almost by deﬁnition. All other approaches either include redistributive components or exclude immigrants from the beneﬁts of this component of social protection. Accordingly, their assessment is less positive than that of Bertola’s proposal. Turning to the impact on redistribution (row 2a), the evaluation of the four approaches is based on their consequences for the extent of redistribution. Clearly, Bertola’s proposal would imply a massive reduction in redistribution for a majority of EU residents to a minimum welfare standard. Insofar as the aim of the EU is to provide a high level of social protection for all its citizens, the complete integration approach or a delayed integration scheme warrants this objective to the greatest extent. Looking at the principle of subsidiarity (row 3), it is obvious that Bertola’s proposal requires a coordination of minimum welfare levels and may also involve a central mechanism of ﬁnancing them. Hence, it implies severe restrictions on the ability of individual states to decide about the extent of the welfare state. All other approaches basically leave this right to individual member countries. The ﬁndings contained in rows 1 to 3 of Table 1 are based on the three European Union objectives highlighted by Bertola, namely ‘‘free mobility of goods and/or factors, local decision-making powers in the labor-market and social protection area, and social inclusion’’ (p. 104). Table 1 shows that the

Comment

125

immediate and complete inclusion of immigrants into the welfare state of the host country and the delayed integration principle warrant the three criteria best. Hence, Bertola’s proposal for dealing with the externalities caused by local or national decisions in the presence of mobile individuals is inferior to other approaches, at least when applying his own evaluation criteria strictly. However, my perception of the public debate, EU objectives, and my reading of Bertola’s paper is that the list of objectives contained in rows 1 to 3 in Table 1 needs to be extended. First, differential transfers and beneﬁts can induce workers and also people from outside the labour force to migrate to countries characterised by a more extensive welfare state system. Such mobility, which may be labelled welfare migration, is not desirable from an efﬁciency point of view. Moreover, the ﬁnancial burden of the welfare state for countries which experience welfare migration will rise so that either taxes have to be raised or the extent of the welfare state has to be reduced, that is, a ‘‘race to the bottom’’ may occur. In either case, welfare immigration will be in conﬂict with the EU objective of promoting the development of economic activity and a high level of employment and social protection. Since the extent of social protection is already being evaluated, the additional evaluation measure focuses on the impact on the development of economic activities (row 4). Second, the discussion on the European Convention has clariﬁed that the prohibition to discriminate on the basis of nationality is likely to be extended beyond work-related issues in the future.1 If a differential treatment of potential welfare recipients on the basis of nationality is prohibited or at least restricted, this feature should also be taken into account in the search for an optimal allocation of social policy decision-making rights in the EU (row 5). Incorporating the two additional objectives of non-discriminatory treatment of immigrants and of preventing welfare migration, respectively ensuring economic development or progress, demonstrates that the evaluation of the various approaches may be altered substantially. While the complete exclusion principle ideally prevents welfare migration at the expense of discriminating immigrants, the complete inclusion approach may foster such behaviour in exchange for nondiscriminatory rules. The assessment of Bertola’s proposal is somewhat mixed. It does not rely on discrimination and can reduce welfare migration. Insofar as a reduction in welfare immigration alleviates the ﬁnancial problems of welfare states, Bertola’s proposal is conducive to economic progress. However, less redistribution is not automatically equivalent to enhanced economic development. A delayed integration of immigrants into the redistributive elements of the

1

The draft of the European Convention, for example, states: ‘‘Everyone residing and moving legally within the European Union is entitled to social security beneﬁts in accordance with Union law and national law and practices’’ (Art. II-34). While the Convention does not represent binding law yet, it is obvious that the European Union will not be able to implement policies which would grossly violate the Convention’s spirit.

126

Laszlo Goerke

welfare state of their host country clearly clashes with a principle of non-discrimination. As far as welfare migration can be restricted, it will prevent further ﬁnancial burdens on welfare states and, hence, be conducive to economic development and high employment. Moreover, this outcome is achieved without necessarily reducing the amount of redistribution to natives, which may be argued to be a superior outcome with respect to the objective of social progress than would result when applying Bertola’s proposal. Without probing further into the linkage between the extent of redistributive transfers and economic development, no clear-cut distinction between the delayed integration approach and Bertola’s proposal with respect to their impact on welfare migration and economic progress seems to be feasible. The above discussion has clariﬁed that a comprehensive evaluation of different proposals for an assignment of social policy decision-making rights requires the incorporation of at least two additional criteria beyond the ones explicitly employed by Bertola. One of them, the requirement of non-discrimination, strongly favours his proposal. The assessment with respect to the other, namely on ensuring economic and social progress by limiting or preventing welfare immigration, is ambiguous since the reduction in welfare migration is obtained at the expense of reducing redistribution. However, Bertola’s proposal is superior to the viable alternative of delayed integration. The analysis of four proposals for redesigning EU welfare policy has highlighted the elements that are necessary for a successful approach. First, welfare immigration can be limited if a distinction is made between an insurance against labour income risk and redistributive activities. Thus, unrestricted mobility of labour or people does not require a complete centralisation of the welfare state to prevent welfare migration. Second, to foster migration that enhances efﬁciency, immigrants have to be granted the same insurance against labour income risk as residents. Third, discrimination on the basis solely of nationality or migration status is likely to become difﬁcult in the future. However, the analysis has also shown that none of the proposed approaches is dominated by any other. It may, hence, be worthwhile to consider piecemeal reforms of the welfare state which reduce the incentives for welfare migration, without, ﬁrst, preventing the efﬁciency-enhancing ﬂow of labour and, second, entirely giving up the principles of subsidiarity and non-discrimination. Such partial reforms cannot warrant all objectives simultaneously, but if they fulﬁl the criteria and requirements to some extent, they may be politically more viable than the somewhat dramatic changes that, for example, Bertola’s proposals involves. One component of such a piecemeal approach would be a substitution of welfare savings accounts for traditional social security systems.2 The central idea

2

See, inter alia, Feldstein and Altman (1998), Fo¨lster (1999), Fo¨lster et al. (2003), Stiglitz and Yun (2005) and Brown et al. (2006) for according proposals and (generally favourable) evaluations.

Comment

127

of such savings accounts is that individuals (are forced to) save a ﬁxed amount of their income that can be used to smooth income over time if individuals become unemployed or ill, or suffer income losses for other reasons. A potentially remaining balance in the account, for example at the end of one’s working life, can be transferred to a pension payment. In general, such proposals for welfare savings accounts also include regulations concerning individuals who have either not been able to save a sufﬁcient amount before they experience a pronounced decline in income, or have run down their savings accounts because, for example, an unemployment spell or an illness has been persistent. These accounts, therefore, combine insurance elements and redistributive components. If only the redistributive component is tied to a previous labour market experience in the country that has to make such payments, the incentives for welfare migration are reduced, while discrimination of immigrants is limited. Moreover, each country can decide on the exact speciﬁcation of such savings accounts, thus conforming to the principle of subsidiarity. Finally, savings accounts can reduce the disincentive effects of the welfare state. Thus, such accounts can be employed to reduce the problems that the welfare state may face owing to any increased labour mobility in a larger EU. Related to the advantages of welfare savings accounts are those of funded pension schemes. Funded, instead of pay-as-you-go, systems can reduce the incentives for migration. This will be the case if funded schemes contain fewer redistributive elements, or if such components only become operative upon retirement. Implicitly, therefore, funded pension schemes incorporate elements of a delayed integration principle and allow for a stronger emphasis on actuarially fair insurance systems. The more important funded pension systems become, the less the need to modify the principles governing the pension element of the welfare state at a European level. Another potential component of a piecemeal reform of the welfare state may be to link not only social insurance payments but also redistributive transfers to labour market activity. If individuals who are able to supply labour are only entitled to beneﬁts if they are actually working, the incentives for welfare immigration will be reduced. All these elements of an alternative, piecemeal response to greater labourmarket integration have been proposed as mechanisms for enhancing domestic labour market efﬁciency. According proposals may be politically more viable, conform to a greater extent with EU legislation, take into consideration differential preferences of countries, and yield similar gains in terms of economic and social development. Redesigning welfare policy at the level of the EU may not be necessary after all. References Brown, A., M. Orszag and D. Snower (2006), ‘‘Unemployment accounts and employment incentives’’, IZA Discussion Paper No. 2105.

128

Laszlo Goerke

Feldstein, M. and D. Altman (1998), ‘‘Unemployment insurance savings accounts’’, NBER Working Paper 6860. Fo¨lster, S. (1999), ‘‘Social insurance based on personal savings accounts: a possible reform strategy for overburdened welfare States?’’, pp. 93–115 in: M. Buti, P. Daniel and R. Luis, editors, The Welfare State in Europe: Challenges and Reforms, London: Edward Elgar. Fo¨lster, S., R. Gidehag, M. Orszag and D. Snower (2003), ‘‘Assessing the effect of introducing welfare accounts in Sweden’’, pp. 255–274 in: T.M. Andersen and P. Molander, editors, Alternatives for Welfare Policy – Coping with Internationalisation and Demographic Change, Cambridge: Cambridge University Press. Richter, W. (2003), ‘‘Delayed integration of mobile labor: a principle for coordinating taxation, social security, and social assistance’’, pp. 495–518 in: S. Cnossen and H.W. Sinn, editors, Public Finance and Public Policy in the New Century, Cambridge, MA: Cambridge University Press. Richter, W. (2004), ‘‘Delaying integration of immigrant labor for the purpose of taxation’’, Journal of Urban Economics, Vol. 55, pp. 597–613. Saphir, A. (2004), An Agenda for a Growing Europe – The Saphir Report, Oxford: Oxford University Press. Sinn, H.W., C. Holzner, W. Meister, W. Ochel and M. Werding (2002), ‘‘Aktivierende Sozialhilfe – Ein Weg zu mehr Bescha¨ftigung und Wachstum’’, Ifo Schnelldienst, Vol. 9/2002, pp. 3–52. Sinn, H.W. and W. Ochel (2003), ‘‘Social union, convergence and migration’’, Journal of Common Market Studies, Vol. 41(5), pp. 869–896. Stiglitz, J. and J. Yun (2005), ‘‘Integration of unemployment insurance with retirement insurance’’, Journal of Public Economics, Vol. 89, pp. 2037–2067.

CHAPTER 5

The EU Budget: How much Scope for Institutional Reform? Henrik Enderlein, Johannes Lindner, Oscar Calvo-Gonzales and Raymond Ritter ‘‘Now that the European Council has decided to give us the mandate to simplify the system, to make it more functional and visibly democratic, are we supposed to hold onto all the procedures which history has laid down in succession, or are there some which can we do away with?’’ Giuliano Amato, Vice-President of the European Convention, on budgetary decision-making in the EU (September 12, 2002)

1. Introduction Calls for reforming the procedures governing the ﬁnances of the EU have enjoyed some prominence in recent discussions on the future institutional framework of the EU. In academic as well as policy-making circles, it has become routine to refer to the ‘‘problem’’ arising from the present EU budgetary framework at both the multi-annual and annual levels. President Giscard d’Estaing, chairman of the European Convention, while commenting on the present budgetary procedure claimed that ‘‘there is indeed a real problem in that area’’ (Plenary Session of the Convention on September 12, 2002). The representative of the United Kingdom in the Convention, Peter Hain, was equally bold on this issue and called upon his colleagues to ‘‘simplify where we can’’. The Sapir Report prepared at the request of Commission President Prodi calls upon Member States to ‘‘refocus the EU budget’’ (Sapir et al., 2003), while the special report of the European Parliament on the budgetary procedure stressed the ‘‘need for reforming, updating and simplifying’’ (European Parliament, 2003). These demands for an institutional overhaul of the EU budgetary procedure contrast with the results achieved in the recent round of discussions on changes to the EU legal framework, particularly in the context of the European Convention and the Intergovernmental Conference (IGC). While the Constitutional CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79005-8

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

130

Henrik Enderlein et al.

Treaty for Europe (which was adopted by the IGC on June 17–18, 2004 and is presently put on hold after failing to win ratiﬁcation in referendums in France and the Netherlands) introduces a few minor changes to the budgetary procedure, many fundamental issues have either not been raised or have rapidly disappeared from its agenda. The goal of this paper is threefold. It intends to provide an assessment of the EU budgetary procedure based on theoretical

considerations and brief comparisons with the budgetary procedure of the United States; to present an overview of the various reform proposals presently or recently under discussion, and assess their potential implications for the balance between efﬁciency and legitimacy in the EU budgetary procedure; and to suggest an explanation of the nature of the reform discussions on the EU budgetary procedure and their (limited) impact on the outcome of the recent constitutional negotiations within the EU. Although due to its limited size, the EU budget does not bear major economic and ﬁscal importance for the EU economies, a close analysis of the rules and procedure that govern budgetary decision-making at the EU level is nevertheless of considerable interest. First, by identifying the factors that determine the shape and the design of the budgetary procedure, this paper indicates why the EU budget has remained limited in size compared to the national budgets. The small size of the EU budget and the requirement that it must always be in balance or in surplus have clear implications for the conduct of macroeconomic policy in the euro area: it makes the coordination of decentralised ﬁscal policies all the more crucial. The dominance of national budgets reﬂects the current state of the integration process, but it is not necessarily a steady state. Understanding why the EU budget remains small in comparison to national budgets, how it has developed over time, and the debate about its future development, is therefore of great interest to people who are involved in policy formulation with a euro areawide perspective. Second, modiﬁcations to the EU budgetary procedure are a particularly intriguing example of institutional reform in the EU. The reform discussions concerning the EU budgetary procedure reveal that individual policy areas are embedded in the overall state of political integration and that this fact can signiﬁcantly constrain the scope of institutional reform. The assessment of the reform discussions and their outcome, therefore, contributes to a better understanding of the impediments in launching area-speciﬁc reforms. The paper is structured as follows. Section 2 outlines the main theoretical considerations. Section 3 presents the main criticisms of the current EU budgetary procedure and reform proposals, and assesses them on the basis of comparisons with the budgetary process in the United States and on the basis of the relevant theoretical literature. Section 4 reviews the actual changes contained in the EU Constitution and Section 5 concludes the paper.

The EU Budget

131

2. Some theoretical considerations The assessment of the EU budgetary procedure has to be placed into two sets of wider theoretical analyses. On the one hand, focusing on the particular political and institutional context of the EU, which is based on the pooling of national sovereignties in a supranational framework, appears necessary in order to understand the implications of that particular context for the budgetary procedure. On the other hand, an overview of the main theoretical analyses of budgetary procedures, in general, is necessary to allow for a description of the main features that should be provided by an appropriate EU budgetary procedure. Three main concepts are used to assess the current EU budgetary procedure: efﬁciency, legitimacy and political integration. These will be considered in more detail later in this section, but it is worth mentioning them brieﬂy at this point. We consider that any budgetary procedure has to strike a balance between the considerations of efﬁciency and legitimacy under the constraint of political integration. The efﬁciency of a budgetary procedure can be described as the timely and ﬂexible allocation of resources in order to ensure the appropriate provision of the main public goods required. The legitimacy of a budgetary procedure is derived from the degree of democratic control by citizens in a way that resources are allocated according to the will of the people and that any kind of ‘‘rentseeking’’ is minimised. It follows that from a purely institutional perspective there is a conﬂict between the basic requirements of efﬁciency and legitimacy, with large units enhancing the former and small units the latter (Scharpf, 1988). However, we believe that beyond this simple institutional relationship, the compatibility of efﬁciency and legitimacy is largely conditioned by a political dimension related to the willingness of citizens to accept the redistributive implications of a common budgetary authority. The legitimacy of a large budgetary unit (e.g. in a large nation state) increases in line with the citizens’ sense of belonging to that unit, and thus their willingness to be part of a pooled system of income redistribution also increases. We use the term ‘‘political integration’’ to refer to this willingness.1 2.1. The EU’s institutional framework between supranational and intergovernmental decision modes The origins of the EU’s present institutional set-up can be assessed in terms of a principal-agent analogy. This starting point beneﬁts from the use of constitutional choice literature (e.g. Buchanan and Tullock, 1962), which gives normative and positive accounts of the origins of constitutional orders and their

1

This deﬁnition of political integration obviously relates to the two large bodies of literature on ﬁscal federalism (for an overview see Oates, 1999), and on the appropriate size of nations (see Alesina and Spolaore, 1997).

132

Henrik Enderlein et al.

legitimacy and efﬁciency. The focal point of the constitutional choice literature is the pooling of individual citizens’ sovereignty and the delegating of functions and powers to elected representatives. Citizens, as principals, allow elected representatives, as their agents, to be in charge of public institutions and take political decisions. While the principal-agent analogy allows the constitutional choice literature to explain the rationale behind the existence of states, it can also be used to analyse institution building at the international level, and, in the speciﬁc European context, at the supranational level (Pollack, 2003). Here, states are regarded as the principals. Under certain circumstances, cooperation among states is beneﬁcial and states may decide to pool their political sovereignty in selected areas. A key example of such an institution based on a certain degree of pooling of sovereignty among states is the United Nations (UN). Other examples include the UN specialist bodies such as the International Monetary Fund and the World Trade Organisation. In the context of pooled sovereignties at the supranational level, there are, however, two types of decision-making procedure: Intergovernmental cooperation. States agree to take decisions in the relevant policy area by unanimity, thus preserving a considerable degree of ‘‘ultimate’’ sovereignty stemming from the power of each individual member state to veto decisions. These decisions are legitimised by the direct link between citizens and their national governments and the fact that these governments cannot be outvoted. At the same time, the unanimity rule renders negotiations difﬁcult, as decision-making is clearly hampered by national vetoes. Supranational governance. States might realise that it is beneﬁcial to take some decisions by majority and to delegate, as principals, certain functions to an (independent) agent helping them to overcome collective action problems. This move towards majority voting and delegation is motivated by an interest in increasing the efﬁciency of international decision-making. The states sacriﬁce their veto power and assign political tasks such as overseeing implementation and mediating between states to a supranational agent. However, supranational governance may, in the view of citizens, be considered less legitimate, as national governments can be outvoted and supranational agents exercise power without a direct mandate from the citizens.2 In order to counteract this lack of direct legitimisation, supranational actors may be directly elected by the citizens, thus circumventing national governments as the sole source of legitimacy. However, this increase in legitimacy may come at the cost of efﬁciency, as the involvement of an additional directly elected agent may increase the complexity of decisionmaking.

2

However, it is also argued that efﬁcient decision-making procedures yielding effective policy outcomes can also create a high degree of legitimacy (see the distinction between input and output legitimacy in Scharpf, 1999).

The EU Budget

133

Applying these theoretical considerations to the EU, we can see a political system that has reached an advanced but still limited state of integration. Important areas of national sovereignty, such as competition and trade policy and the four freedoms that underpin the single market, are pooled at the EU level and a certain degree of political identiﬁcation and acceptance has been achieved. However, the nation state still provides an important reference point for political identity and national institutions continue to determine the daily life of citizens in a vast number of areas. With regard to the institutional set-up, the EU level is governed by a combination of supranational and intergovernmental forms of decision-making. In the intergovernmental sphere, heads of state or government set the broad policy guidelines in the European Council by consensus. Moreover, they adopt Treaty changes, which are subsequently ratiﬁed, according to national domestic procedures. In the Council of Ministers, representatives of national governments make detailed policy decisions on the basis of legislative initiatives from the supranational Commission. Voting rules vary depending on the policy ﬁeld. Where unanimous voting still applies, the Council acts as an intergovernmental body, but in cases of qualiﬁed majority voting, ministers switch to supranational mode. The directly elected European Parliament provides a link between the supranational decision-making process and the citizens. Its involvement as a strong veto-player is largely connected to qualiﬁed majority voting in the Council, where, in terms of legitimacy, it compensates for the loss of Member States’ veto power. Moreover, it fulﬁls control functions vis-a`-vis the Commission, and thus, contributes to an appropriate level of accountability. Hence, the present institutional set-up of the EU, combines elements of intergovernmental cooperation and supranational governance and strikes a balance between legitimacy and efﬁciency. This overall balance lays the foundation for the speciﬁc rules and procedures in the different policy ﬁelds of the EU. The involvement of the Commission, the European Parliament and the Council is a common feature of most policy domains. It caters for similar legitimacy and efﬁciency concerns, regardless of the particular characteristics of the speciﬁc policy ﬁeld. This does not mean that there is no variation in the degree to which political authority is delegated to supranational agents in the different domains. Indeed, in certain policy ﬁelds,such as competition policy, the Commission exercises signiﬁcant decision-making powers, while in others, such as the Common Foreign and Security Policy, decisions are taken by unanimity in the Council. However, policy ﬁelds evolve as part of the overall institutional framework. Moreover, they reﬂect the general scope of pooled sovereignty and the degree of political integration and acceptance. Therefore, calls to reform the decision-making procedure in a particular policy ﬁeld should not be assessed independently of the general state of integration. On the contrary, their contents should be put under scrutiny with a view to the ‘‘meta-level’’ that the state of integration constitutes. This meta-level not

134

Henrik Enderlein et al.

only overarches but also determines decision-making in the different policy ﬁelds of the EU. Therefore, the complexity or alleged inefﬁciency of a decision-making procedure cannot be exclusively attributed to the institutional provisions governing the speciﬁc policy ﬁeld, but need to be linked to the characteristics of the meta-level. Moreover, the degree of embeddedness in the meta-level constrains the scope of institutional reform in a policy ﬁeld. In other words, while farreaching reform proposals could increase the efﬁciency and legitimacy of the procedure, they would actually require a higher degree of political integration than is currently the case. 2.2. Budgetary procedures – a balance between efficiency and legitimacy Shifting the analysis from the general reﬂections on the institutional set-up of the EU to the procedural features of the budgetary ﬁeld, it becomes clear that the complex interplay between the concerns of legitimacy and efﬁciency is even further accentuated in that particular area. Decisions on the utilisation of public ﬁnances need to be subjected to tight control by citizens so that resources are allocated according to the will of the people and any kind of rent-seeking is minimised. At the same time, such decisions need to ensure efﬁcient resource allocation and guarantee the provision of the main public goods required. There is a solid body of academic literature linking these fundamental requirements of a budgetary procedure to the underlying institutional framework. In relation to the effectiveness and efﬁciency of the budgetary procedure, an inﬂuential literature has emerged from a number of cross-country analyses comparing different types of institutional frameworks. Mainly developed by von Hagen (1992), von Hagen and Harden (1994), and Hallerberg and von Hagen (1999), this literature establishes theoretically grounded and empirically detectable links between institutional components of the budgetary procedure and its success in achieving speciﬁc policy goals (in this literature mainly ﬁscal discipline). The starting point for these analyses is the assumption that an optimal allocation of public resources is unlikely to be achieved within a complex interplay of numerous actors pursuing different interests. They show that a strong political authority at the top of the process (e.g. a strong ﬁnance minister), a parliament with limited amending powers, and a strict implementation process ensure a smaller in size budget and more restricted use of debt as a way of ﬁnancing public expenditure. What stems from this literature is the claim that budgetary procedures need to be capable of yielding swift and effective decisions on ﬁscal measures, even if such measures seem unpopular in the short run. The delegation of powers to a strong political authority for efﬁciency reasons, however, needs to be embedded in an effective democratic control mechanism to ensure that voters’ legitimacy concerns are met. As a number of stylised analyses in political economy convincingly demonstrate, an effective separation of powers needs to be enacted for voter utility to be maximised in a system of delegated authority (Persson et al., 1997; Persson and Tabellini, 2003). The

The EU Budget

135

analyses are based on the key distinction between presidential and parliamentary constitutional regimes, and argue that direct accountability is more effective in presidential than in parliamentary democracies. Persson et al. (1997, 2000) give two reasons for this. First, the chain of delegation is shorter under the presidential system as the executive is directly accountable to the voters. In a parliamentary system, where the executive is directly accountable to the legislature and not the voters, the scope for collusion at the voters’ expense is greater. Second, the degree of separation of powers tends to be larger, and the system of checks and balances tends to function more effectively in presidential systems. Persson et al. (2000) generally argue that checks and balances in a presidential system improve accountability and strengthen the incentives for good behaviour among politicians, as voters can exploit diverging interests among the bodies involved in decision-making to prevent abuse of power or to reduce information asymmetries between themselves and the policymakers (see also Persson and Tabellini, 2003).

2.3. The EU budgetary procedure and the constraint of limited political integration Looking at the EU budgetary procedure from the twin perspectives of ‘‘efﬁciency’’ and ‘‘legitimacy’’, it becomes clear that the institutional complexities arising from the articulation of supranational and intergovernmental decision modes are even greater in the budgetary sphere. While it is a straightforward argument that an appropriate budgetary procedure for the EU would seek to strike a balance between citizens’ legitimacy concerns and procedural efﬁciency concerns, it has to be taken into account that such balance has to be determined under the constraint of still limited political integration. Indeed, as a large number of studies and surveys indicate (see overview in Marx and Hooghe, 2003), the delegation of sovereignty from European citizens to the EU is still grounded to a large extent in a strong national identity rather than a European identity. The mechanism of democratic control by European citizens of decisionmaking bodies of the EU still seems to be largely perceived as requiring the intervention of national governments rather than being exercised directly through the European Parliament. In this context, options for reforms of the budgetary procedure need to be considered within the scope granted by the present degree of political integration. The EU budget basically reﬂects the range of those areas in which sovereignty is pooled at the EU level. Such pooling is often legitimised on the basis of efﬁciency. However, taking into account the varying degree of European citizens’ political identiﬁcation with supranational decision-making at the EU level and also the very tangible ﬁnancial component of budgetary decisions, possible legitimacy concerns need to be carefully addressed. The involvement of Member States’ representatives serves to redress such concerns, as budgetary

Henrik Enderlein et al.

136

Figure 1.

The efficiency-legitimacy trade-off

decisions are usually regarded as an issue close to national sovereignty.3 However, such intergovernmental elements may come at the expense of efﬁciency. This paper thus argues that the exogenously given state of political integration is the key to understanding the current debate – and its outcome – on the reform of the EU budgetary procedure. Proponents of far-reaching reforms seem to have underestimated the scope for change allowed by the general state of integration as well as the high degree to which the current procedure is already in tune with the present degree of integration. To illustrate this assessment, Figure 1 (which should be viewed as a metaphorical depiction rather than a formula-based curve) indicates possible solutions to the identiﬁed trade-off between legitimacy and efﬁciency. These solutions are limited by a ‘‘Pareto frontier’’ determined by the state of integration – i.e. outward/ inward shifts of the frontier can only be triggered by increases/decreases in political integration.

3

This is comparatively less the case in the area of monetary policy. Prior to the start of the EMU, monetary matters had already been delegated to independent central banks on the basis of the understanding that politically independent monetary authorities would be conducive to give signiﬁcantly better results to the society. Given the success of central banks’ independence from citizens’ direct electoral choices and government inﬂuence, the transfer of monetary authority from the domestic to the European level was perceived as an obvious natural political process.

The EU Budget

137

Proponents of reform assume that the budgetary procedure is currently situated at the suboptimal point A. This paper, in contrast, argues that point B is probably an accurate description of the current procedure. While the proponents of reform take the view that reforms would signiﬁcantly and simultaneously increase efﬁciency and legitimacy, this paper argues that the improvements can only be marginal and would have to concentrate on enhancing either efﬁciency (leading to B0 ) or legitimacy (leading to B00 ). 3. Reforming the EU budgetary procedure: an assessment This section assesses the main proposals for reforming the EU budgetary procedure, and argues that simultaneous increases in legitimacy and efﬁciency are highly unlikely without any major progress in political integration. In the current political discussions, an array of different reform proposals has been presented by various actors. The European Parliament set out its ideas in a report that was initiated by the Committee on Budgets (European Parliament, 2003). The European Convention focused extensively on the budgetary procedure in a special discussion circle (European Convention, 2003a) during the preparatory phase leading to the issue of its draft Treaty establishing a Constitution in June 2003 (European Convention, 2003b). Finally, an independent, high-level study group, which was established at the initiative of the President of the Commission, also covered the budgetary procedure in its report on the economic governance of the EU (Sapir et al., 2003). Table 1 gives an overview of these different proposals. These reform proposals can be assessed in the context of a comparison of the EU budgetary procedure with the decision-making process of the US federal budget.4 Obviously, comparing the United States and the EU is limited in its legitimacy. The United States is already a fully-ﬂedged federal system where ‘‘intergovernmental’’ cooperation has made way for ‘‘supranational’’ governance and states have pooled much of their sovereignty, as is also reﬂected in the actual contents of the federal budget (see Table 2). In the United States, political decision-making takes place in the context of dominant federal actors. The executive power enshrined in the Presidency largely depends on agreements with the House of Representatives, representing the interests of electoral districts, and the Senate, which represents the interests of individual states. Imposing an analogy with the EU, the role of the European Commission can be compared with that of the US President, the role of the European Parliament with that of the House of Representatives and the role of the Council of Ministers with that

4

The description of the US budgetary process is based on the seminal treatment of the matter by Schick (2000), as well as Ofﬁce of Management and Budget (2003) and Congressional Budget Ofﬁce (2003). For the EU budgetary procedure Laffan (1997), Brehon (1997), Nava (2000) and European Commission (2002) serve as the main references.

Reform Proposals

Content of the budget (a) Re-focussing the expenditure side of the EU budget (b) Modifying the revenue base of the EU budget

Annual budgetary procedure (a) Eliminating the distinction between compulsory and non-compulsory expenditure (b) Giving either the Council or the European Parliament the last word

Overview of reform proposals

European Parliamenta

The Convention’s Discussion Circle on the Budgetary Procedureb

Sapir Groupc

European Conventiond

— Yes

— —

Yes Yes

— —

Yes

Yes (preferably with a ‘‘super majority’’)

Yes

Yes —

Yes No

— —

Yes for the FP, but no for later revenue Yes Yese

No Yes

No No (at least ﬁve years)

— —

No No (at least ﬁve years)

Yes

No (but with provisions for ﬂexibility reserve)

Yes

No

Yes

Yes

—

Yes

Yes (EPf)

Yes (EPf)

—

Yes (EPf)

a European Parliament (2003), ‘‘Report on the reform of the budgetary procedure: possible options in view of the revision of the treaties’’, 20 February, A5-0046/2003. b European Convention (2003a), ‘‘Final report of the discussion circle on the budgetary procedure’’, 14 April, CONV 679/03. c Andre´ Sapir et al. (2003), An Agenda for a Growing Europe. Making the EU Economic System Deliver, July. d European Convention (2003b), ‘‘Draft treaty establishing a constitution for Europe’’, 18 July, CONV 820/1/03 REV 1, CONV 847/03, CONV 848/03. e But possibly allowing for the Council of Ministers to meet at the level of the Heads of State or Government. f Within certain limitations.

Henrik Enderlein et al.

Financial Perspective (a) Adopting the Financial Perspective and/or the revenue side by qualiﬁed majority among Member States (b) Institutionalising the Financial Perspective (c) Adopting the Financial Perspective in the Council of Ministers instead of the European Council (d) Taking the veto power from the European Parliament (e) Shortening the time frame of the Financial Perspective and synchronising it with the European Parliament elections (f) Increasing the ﬂexibility of the expenditure headings

138

Table 1.

The EU Budget

Table 2.

The content of the EU and US budgets in 2003

EU budget

US federal budget

Total expenditure Amount As a percentage of GNI

EUR 99.7 billion 1.0%

Selected expenditure

139

As a percentage of total expenditure in 2003 45%

Agricultural policy and rural development Structural and Cohesion Funds Internal policies External actions

7% 5%

Pre-accession aid

3%

34%

Total expenditure Amount As a percentage of GDP Selected expenditure

Social security

Health (including Medicare) National defence Income security (e.g., unemployment beneﬁts) Education

USD 2,158 billion 19.8% As a percentage of total expenditure in 2003 22%

22% 19% 15%

4%

Sources: European Union Financial Report (2003) and Ofﬁce of Management and Budget.

of the Senate. From this analogy, one clear difference between the broader political systems of the United States and the EU becomes immediately apparent: the degree of representation of the citizens. In the United States, rather than being representatives of state governments, Senators are directly elected by their constituents. Similarly, although the President of the United States and the European Commission could be characterised as the executive power respectively in the two systems, there is a clear difference in terms of how they are appointed. In other words, there is a direct link between the electorate and all three players in the US budgetary process, which is further enhanced by a national identity that attaches US citizens to the federal level. In terms of the criteria outlined above, though this difference matters most in checks and balances, it also plays a role when it comes to the effectiveness of decision-making and transparency. This brief and broad-brush comparison between EU and US procedures is nonetheless a worthwhile exercise as it illustrates areas for possible reform of the EU procedure, while also showing that even the fully integrated political system of the United States still functions on the basis of a budgetary procedure based on a complex interplay between legitimacy and efﬁciency concerns within a multi-level polity.

140

Henrik Enderlein et al.

3.1. Proposals for reform at the general level One of the main criticisms voiced with regard to the general level of the EU budget relates to the expenditure structure of the EU budget, which also implicitly raises the issue of the size of the EU budget. In this context, it should be noted that at present the size of the EU budget is almost negligible in economic terms (around 1% of EU GNI),5 whereas the US budget amounts to roughly one third of US GDP, of which the federal level accounts for approximately twice the expenditure than that of the states. The large difference in size between the EU and US budgets reﬂects the very different compositions of the respective expenditures and sources of revenue. While in the United States the federal budget contributes to the ﬁnancing of state and local government through annual federal grants amounting to around 3.5% of GDP, the relationship in the EU is the opposite. Moreover, large parts of the US budget are concentrated on welfare, military spending, education and servicing federal debt. But, in the EU, spending on welfare, defence and education is done at the national level. The EU budget focuses mainly on a small number of policies with strong redistributive bent, especially agricultural and regional policies, which alone account for more than 80% of expenditure in the EU budget. While agricultural spending provides European farmers with subsidies, regional spending redistributes funds to less wealthy regions. Although often presented as such (e.g. Leonardi, 1999), these policies may only to a limited extend be regarded as public goods beneﬁting the Union as a whole. Together with a revenue side that de facto is based largely on national contributions, these policies result in a system in which Member States are under pressure to demand ‘‘juste retour’’ for the payments into the budget. The present EU budget is expenditure-led within a limit set in an ‘‘Interinstitutional Agreement’’ with a multi-annual horizon (see below). Resources are raised to match what is needed to carry out the EU policies. This creates very different incentives for the European Parliament and the Council as the two arms of the budgetary authority. For Member States, there is an incentive to reduce expenditure to achieve a reduction in their direct contributions, while for the European Parliament, there is an incentive to propose expenditure programmes, as Member States will automatically furnish the matching ﬁnancing up to the overall ceiling. Members of the European Parliament (MEPs) thus enjoy a unique position – they gain credit for expenditure agreed by the European Parliament, but are not associated with the related costs. This is not the case for US politicians, who also set the tax levels necessary to ﬁnance expenditure. In practice, however, differences in the incentives for Member States and MEPs have been reduced, on the one hand, by a growing acceptance among MEPs of an austerity approach towards budgetary decisions and, on the other

5

The size of the EU budget is currently formally capped at 1.24% of combined national GNIs.

The EU Budget

141

hand, by the interests of individual Member States about expenditure in areas in which they gain more from the budget than they contribute. While one of the main rationales for the existence of a public budget is the ﬁnancing of the provision of public goods which would otherwise be underprovided, the fact that current expenditure at the EU level is for the most part redistributive in nature, shows that equity rather than efﬁciency criteria plays a dominant role in guiding the expenditure side of the EU budget. Although it does not question the existence of a redistributive element in the budget, the Sapir report favours a ‘‘change in the current mix of equity and efﬁciency considerations’’ when deciding upon expenditure ﬁnanced by the EU budget as well as an increase the provision of public goods, which do not prompt a calculation of juste retour among contributors. The Sapir report suggests that the EU budget should promote growth through expenditure on R&D, education and training, and infrastructure. Other proposals have included internal and external security, foreign policy, research, immigration, and single market-related issues (see e.g. Tabellini, 2002). The key problem related to changes on the expenditure side is obviously the one of allocating more legitimacy for decision-making at the European level. It has to be borne in mind that the present structure of EU expenditure reﬂects the current institutional set-up of the EU and the political consensus on the policy tasks assigned to the EU and national levels respectively.6 Any changes to be made on the expenditure side of the EU budget would thus necessitate an agreement among policymakers on the modiﬁcation of the present assignment of policy tasks at the EU level. Reforms of the spending programmes depend on the willingness of citizens to allocate additional tasks to the EU and thus may only be realised as a consequence of changes in the degree of political integration (see also Strauss-Kahn et al., 2004). A related ﬁeld of criticism with regard to the present general features of the EU budget is its revenue base. At present, the EU is funded predominantly by direct transfers from the Members States, although the current system is called a system of ‘‘own resources’’. Direct transfers obviously facilitate the computation of net positions and do little to curb Member States’ pursuit of juste retour. The system of own resources has become less and less autonomous since its creation in the 1970s when it was intended to give resources to the Community that would ‘‘belong’’ to it and would not depend on decisions by national governments. This, together with the granting of the ‘‘power of the purse’’ to the European Parliament in the 1970s, was a development of a federal nature, aimed

6

As Padoa-Schioppa (2002, p. 200) puts it: ‘‘There can be no doubt that it would be a good thing, for the Union, to have more room for manoeuvre in the area of budgetary policy. But it is also my belief that this can only come as a natural consequence of political union. No country has ever adopted a large budget just in an effort to obtain more instruments for economic policy. Historically, the size of the budgets grew because the functions attributed to the Union grew.’’

142

Henrik Enderlein et al.

at enhancing the supranational element of the Community. However, as Community expenditure increased, the traditional own resources, namely customs duties, agricultural levies and VAT contributions, proved insufﬁcient and a fourth resource – in form of a set percentage of national GNP – was established as part of the Delors-I package in 1988 which also introduced the multi-annual Financial Perspective (see below). This fourth resource, which now accounts for approximately two-thirds of total EU budget revenues, can be regarded as similar to the pre-1970 period in which the Community was ﬁnanced by contributions from the Member States (Begg and Grimwade, 1998). In line with the notion of an expenditure-led budget and in consideration of the sovereignty concerns of Member States, changes in the revenue structure of the EU have to be taken by unanimity following a procedure that is completely separate from the annual budgetary procedure. In the United States, around one-half of current revenues are raised by personal income tax, a third by social insurance receipts and a further 10% by corporate income tax. Almost all of these sources of income are referred to in the US federal budget as general funds. Decisions over the level and structure of taxes are often directly linked with expenditure decisions. The US budget is allowed to run deﬁcits, but, in line with the current state of integration, the Treaty clearly forecloses this option in the EU and obliges the budgetary authority to adopt balanced budgets.7 In the light of the decreasingly autonomous character of the EU’s ‘‘own resources’’, a number of proposals have been made. For example, the Sapir report proposes the introduction of a new revenue source with a clear EU dimension, such as a tax with a very mobile tax base within the EU. This could be capital income taxes or stock exchanges taxes. While such a tax would arguably convey a strong and positive symbolic message and enhance the transparency of citizens’ contributions to the EU, it is doubtful that such a proposal would be acceptable given the present state of political integration. On the one

7

In a sense, the situation with regard to the EU budget more closely resembles the one that characterised the US federal budget during the nineteenth century. In fact, until 1921, the budgetary process in the US can be characterised as one of legislative dominance (Schick, 2000), in which the Congress effectively constrained the executive not only over total expenditure but also over individual items of expenditure (see description of the EU budgetary procedure below). The size of the US federal budget was very limited and stable at around 2% of GDP until the Civil War in the 1860s, and mainly devoted to ﬁnancing public works projects, defence and the operations of government agencies. In addition, in the absence of a federal income tax, the US budget lacked true own resources to commit itself to make major investments in transportation and ﬁnance. The major source of US federal revenues throughout the nineteenth century remained the tariff, which accounted for around 80 to 90% of federal revenues (Wallis, 2000). Moreover, and despite the lack of a formal constraint on budgetary outcomes, the norm of balanced budgets was broadly maintained. It was not until the 1930s that the US federal budget gained the prominence that it currently enjoys in the allocation of ﬁscal activity between the different levels of government.

The EU Budget

143

hand, Member States seem very reluctant to move the current set-up on the revenue side in a more supranational direction, which would ultimately give the EU resources that cannot be redirected to Member States when they exceed the level of expenditure. On the other hand, the creation of an ‘‘EU tax’’ would certainly require a well-functioning system of checks and balances at the EU level. As Persson et al. (1997) and Persson and Tabellini (2003) argue, rents from power and incomplete information at the expense of voters are accentuated in a common pool situation in which the bodies participating in the budgetary procedure are ‘‘residual claimants’’ over the budget (i.e. they can keep the beneﬁts of spending within the majority, putting part of the costs on the excluded minority). Put in the context of the EU institutional set-up, this argument is certainly far too simple. There are no clear lines of division between majorities and minorities in the EU context. Alliances tend to vary greatly between different groups of actors in both the European Parliament and the Council. The fundamental demand of ‘‘no taxation without representation’’ is still likely to be voiced as an argument against an EU tax. This argument would be technically wrong, as representation could be ensured by the European Parliament, but in the perception of most EU citizens, the time does not seem ripe for introducing an EU tax. Overall, looking at general aspects of the budgets in the United States and the EU, it is evident that the role of the EU budget is commensurate with the much lower degree of political integration in the EU: the size of the budget is small, its content is not geared towards the provision of basic public goods but has a very limited focus and it is redistributive. This combination of a budget which is funded by direct transfers from Member States and spent largely on regional or agricultural programmes, which are easily tracked to regional or national recipients, rather than on European-wide public goods, creates a situation particularly prone to bargaining. 3.2. Proposals for reform at the level of multi-annual planning The Financial Perspective is the EU’s multi-annual budgetary framework (7 years), which lays down the maximum amounts of both total annual expenditure and annual expenditure in speciﬁc policy areas.8 The Financial Perspective was a welcome development when it was ﬁrst introduced in 1988 (Lindner, 2003). In the late 1970s and early 1980s, EU budgetary negotiations was characterised by confrontations between the

8

While the Financial Perspective sets an upper limit on the annual EU budget, there is some limited leeway for revisions in response to unforeseen circumstances. However, any revision has to remain within the margin for unforeseen expenditure as speciﬁed in the Financial Perspective. Moreover, revisions amounting to more than 0.03% of Community GNI require the agreement of both the European Parliament and the Council, whereby the Council has to act unanimously (in the case of revisions below 0.03% of Community GNI, the Council can act by qualiﬁed majority).

144

Henrik Enderlein et al.

European Parliament and the Council which eventually led to the rejection of entire draft budgets by the European Parliament (La¨ufer, 1990). The early 1980s also saw complaints from the United Kingdom and a lack of sufﬁcient resources. The latter was prompted mainly by the increase in Common Agriculture Policy (CAP) spending and was aggravated when the UK rebate was agreed to. In this context, the Financial Perspective was intended to signiﬁcantly restrict the scope for political choice during the annual budgetary procedure. By linking the revenue and expenditure sides, it ensured that expenditure-led budgets would no longer exceed existing resources (Shackleton, 1990). The Financial Perspective combines intergovernmental and supranational elements. After a proposal by the Commission, it is discussed in the European Council, where it requires unanimous agreement among Member States. The involvement of heads of state or government ensures that differences and stalemates between EU ministers are overcome. The Financial Perspective is ﬁnally adopted as an ‘‘Interinstitutional Agreement’’ between the European Parliament, the Council and the Commission. In contrast to indicative ﬁnancial programming, the ceilings of the Financial Perspective are binding on the three parties. While the revenue ceilings are codiﬁed in legal acts adopted unanimously by Member States in the Council, the expenditure ceilings gain their binding character from the political willingness of actors to adhere to the ‘‘Interinstitutional Agreement’’, which by itself does not have the status of an enforceable legal act (Monar, 1994). The semi-voluntary character of the Financial Perspective obliges actors to cooperate, as the multi-annual framework would otherwise break down. Planning stability and a reduction in conﬂict comes at a price: the Financial Perspective clearly limits the ﬂexibility of budgetary actors and introduces a strong status quo bias. Annual expenditure ceilings for regional spending, for example, have the status of spending targets and thus commit annual budgetary decisions over a period of seven years. Moreover, when the Financial Perspective is renegotiated in the European Council, national governments use their veto power in order to maximise budgetary gains. Although the Financial Perspective is not automatically renewed at the end of the seven year period, the use of veto power and the resulting bargaining dynamics lead to a largely incremental revision of the ceilings, thus respecting the key spending interests of Member States, such as the rebate for the United Kingdom, regional spending for the countries beneﬁting from the Cohesion Fund and unaltered CAP spending.9 In

9

Renegotiations in the European Council take place in conjunction with other political issues. Hence, bargaining among Member States combines budgetary issues with non-budgetary issues. In this context, it has been argued that the EU budget fulﬁls an important ‘‘compensation function’’. It compensates those Member States that might incur costs from the integration process and thus, facilitates a consensus for further integration among Member States (Folkers, 1997).

The EU Budget

145

such a setting, major changes and far-reaching reforms are very unlikely to occur (Begg, 1999) unless there is signiﬁcant progress in political integration.10 In the United States, there is no true multi-annual planning, as there are no binding multi-annual constraints. The President’s budget, which is transmitted to Congress each February, pertains exclusively to the following ﬁscal year, while the horizon envisaged in the congressional budget resolution has no statutory implications. The only multi-annual budgetary implications stem from the mandatory spending implied by past acts providing for entitlements such as social security, but then again, the President and Congress can change the law in order to change the spending on entitlements and other mandatory programmes in any given year. Overall, in the EU, a binding multi-annual budget plan adopted by heads of state or government plays a key role in reducing the conﬂict between budgetary actors and ensuring planning stability. This comes at a price: the ﬂexibility of the annual procedure is seriously curtailed and major shifts between the main spending blocs are de facto almost impossible. The intergovernmental level, at which legitimacy concerns clearly dominate over efﬁciency concerns, thus uses multi-annual planning to strictly limit the scope for supranational decision taking in the annual procedure. Although lengthening the budgetary cycle is discussed, decision-making in the United States proceeds almost exclusively within the remit of the annual procedure. A number of proposals have been made to reform the Financial Perspective. One set of proposals for reform relates to the suggestion of the European Parliament, the Convention’s discussion circle on the budgetary procedure and the Sapir report that the Financial Perspective be adopted by qualiﬁed majority voting among Member States. Such reform would address the key bottleneck of the veto power by Member States. It would also probably reduce the tendency for the negotiations on a new Financial Perspective by-and-large to conﬁrm the existing spending structure. It may also help to curtail the inclusion in the Financial Perspective of amounts earmarked for speciﬁc spending projects in individual Member States without a direct link to existing expenditure programmes, as has arguably been the case with the expenditure for ‘‘particular situations’’ detailing speciﬁc amounts of structural funds for certain regions (Begg, 1999). Combined with the introduction of qualiﬁed majority voting on the revenue side, this may enable Member States to revisit the rules that govern own resources. While these arguments may look very appealing from the perspective of increasing the efﬁciency of the EU budgetary procedure, it is rather doubtful,

10

Although the Commission and the European Parliament have a veto power over the ceilings of the Financial Perspective as signatories of the International Agreement, they rarely exercise it. Usually, the European Parliament grants its consent to the ceilings in exchange for informal extensions of its budgetary powers.

146

Henrik Enderlein et al.

given the present state of political integration, that such reforms would indeed yield the desired increases in efﬁciency. Moreover, they probably lack the necessary basis in terms of transfers of legitimacy. It is not at all certain that an abolition of the unanimity requirements would eradicate the ‘‘pork-barrel’’ character of the Financial Perspective. Experience from areas in which qualiﬁed majority voting has been adopted show that bargaining still ﬁgures prominently in the discussions, as Member States generally try not to outvote each other. Even more importantly, as the decisions regarding the Financial Perspective and the revenue side are ultimately redistributive in nature (Member States will probably always try to work out whether they are net contributors or net recipients), it could be difﬁcult to legitimise the contributions to the EU, budget of Member States that are outvoted in the decision on the Financial Perspective (imagine, for example, a decision taken by majority that signiﬁcantly increases the budget contribution of one of the large net contributors).11 The unanimity requirement in the area of the Financial Perspective and in particular on the revenue side presently serves as an important element in addressing the sovereignty concerns of Member States. Although at ﬁrst glance, the introduction of qualiﬁed majority voting seems conducive to further reducing conﬂict by limiting the veto power of Member States, it might actually have the opposite effect. Currently, Member States are assured that their distributive interests will be taken into account in the negotiations for the renewal of the Financial Perspective. Potential conﬂict is channelled into these negotiations. If this were no longer the case, outvoted Member States may altogether question the advantage of having a Financial Perspective and may carry their discontent into the annual procedure. As long as the binding nature of the Financial Perspective is dependent on the political will and acceptance of all actors involved, the introduction of qualiﬁed majority voting might seriously threaten the functioning of the Financial Perspective. A second set of proposals for reform relates to including the Financial Perspective in the Treaty and making its ceiling legally binding, as proposed by the European Parliament and the Convention’s discussion circle. Budgetary actors would be obliged to adopt a multi-annual budget plan and adhere to its ceilings during the annual budgetary procedure. Such an institutionalisation of the Financial Perspective would give recognition to the special relevance of the Financial Perspective for budgetary decision-making and would thus close the gap between the Treaty provisions and current practices. At the same time, depriving the Financial Perspective of its voluntary nature could undermine its conﬂict-reducing nature. At present, negotiations take place

11

Legitimacy concerns could to some extent be reduced by involving the European Parliament in the decision-making procedure on the revenue side.

The EU Budget

147

in the shadow of the qualiﬁed majority voting prescribed by the Treaty for the annual procedure. The Financial Perspective is not renewed automatically and its existence hinges on the ability of Member States, and subsequently of the Commission and the European Parliament to agree on a new multi-annual budget plan (although the Interinstitutional Agreement stipulates that the ceilings of the old Financial Perspective continue to apply until a new one is adopted, the Council and the European Parliament can veto such automatic continuation). If the Financial Perspective were to be institutionalised and such a veto did not exist, the tendency for negotiations over a new Financial Perspective to conﬁrm the existing spending structure might increase further. Budgetary actors could simply prevent change by relying on the automatic prolongation of the status quo. And, the threat to exit the Financial Perspective (and the Interinstitutional Agreement) or to veto its renewal has so far served as a healthy warning to all actors involved, thereby forcing them to cooperate. A third set of proposals on the Financial Perspective relates to its adoption in the Council of Ministers instead of the European Council. This proposal is included in the EU Constitution, with the possibility, however, that the Council of Ministers could meet at the level of heads of state or government. The adoption by the Budget Council would indeed bring the multi-annual budget plan more in line with the annual budgetary procedure and place it on the same level as other EU policy areas. Having said this, the involvement of the European Council ensures that differences between the spending ministers are transcended. The Budget Council may not be powerful enough to free itself from the grip of the spending Councils (the inﬂuential role of the Agriculture Council is often cited in this respect). Moreover, heads of state or government in the European Council are able to broker an agreement on the Financial Perspective by linking ﬁnancial negotiations with non-budgetary issues, as has been the case so far every time the Financial Perspective has been renewed. A fourth set of proposals on the Financial Perspective relates to synchronising the time frame with the term of ofﬁce of the European Parliament, as has been proposed by the Parliament. This would shorten the time frame from seven to ﬁve years and thus (slightly) reduce the high degree of pre-commitment that the current seven-year cycle entails. Moreover, synchronising the Financial Perspective with the European Parliament elections would politicise the multi-annual budget plan. MEPs could put their budgetary ideas before the electorate and thus gain strength vis-a`-vis the Council. European voters would become more aware of the EU budget and may exercise pressure for a more effective use of EU funds. There are thus clear arguments in terms of legitimacy, and checks and balances in favour of this proposal. That said, a shorter time frame is likely to increase the number of budgetary negotiations, thus reducing the degree of efﬁciency of the procedure. Indeed, politicisation would only be feasible if the European Parliament were fully involved in the negotiations on the Financial Perspective. As long as the ceilings of

148

Henrik Enderlein et al.

the Financial Perspective are set by Member States in the European Council and the European Parliament is presented with a take-it-or-leave-it choice at the end of the negotiations, voters will not see the relevance of European elections for the setting-up of the Financial Perspective. The last proposal on the Financial Perspective relates to increasing the ﬂexibility of the expenditure headings, as proposed by the European Parliament and the Sapir report. The ceilings imposed on speciﬁc expenditure headings under the current Financial Perspective can be considered questionable. By setting rigid limits on expenditure for speciﬁc purposes for a rather long period of time, a potentially useful degree of ﬂexibility is lost. It appears that the preference for detailed bargaining at the multi-annual or annual level may be driven by the voting mechanism through which the multi-annual framework and annual budgets are approved. In the current situation, the unanimity requirement makes agreement on the Financial Perspective particularly difﬁcult, and the costs in terms of lost ﬂexibility may not outweigh the beneﬁts in terms of predictability and stability over a relatively long period of time. However, if voting requirements for approval of the Financial Perspective and the annual budgets were to be relaxed, the beneﬁts of greater ﬂexibility than is currently the case could perhaps be enjoyed without running the risk of a protracted stalemate. However, loosening the headings and facilitating regular revisions may turn the Financial Perspective into a mere ‘‘ﬁnancial planning’’ exercise, largely reshifting the decision-making on budgetary matters back to the annual budgetary procedure and thereby reducing the efﬁciency of the procedure. As a result, the high levels of conﬂict that dominated the 1970s and 1980s may creep back into budgetary decision-making, thereby seriously threatening the timely adoption of annual budgets. Moreover, with its ﬂexibility and emergency reserves, the Financial Perspective already provides buffers to allow for unexpected developments. Theoretical approaches support this point. As mentioned above, voter utility increases in a two-stage system based on separate allocations of responsibilities for the size of the budget and its composition (Persson et al., 1997). Increasing the ﬂexibility of the Financial Perspective (which is de facto set by the European Council) may run the risk of creating a ‘‘common pool’’ bargaining environment, thus leading to increased rent-seeking activity by ofﬁce-holders and a decrease in utility for voters. 3.3. Proposals for reform at the level of the annual procedure The annual budgetary procedure for the EU budget, set out in detail in Article 272 of the Treaty, appears to conform to the division of labour between an executive branch that proposes the budget and a legislative branch that adopts it (Figure 2) The Commission prepares a ‘‘preliminary draft budget’’, which is amended by the Council; this version of the budget, which at this stage of the procedure is called the ‘‘draft budget’’, is forwarded to the European Parliament

The EU Budget

Figure 2.

149

The current budgetary procedure (Article 272 of the Treaty)

for proposals for amendments or modiﬁcations. Following two readings by each of the two institutions, the European Parliament adopts the ﬁnal budget. However, the role of the Commission is actually much more limited than this division of labour suggests. Although the Commission is present at all stages of the budgetary process, its main function ends once the ‘‘preliminary draft budget’’ is submitted. The Treaty bestows each of the two arms of the EU budgetary authority, i.e. the European Parliament and the Council, with speciﬁc powers that largely rest on a distinction between compulsory expenditure (that results directly from Treaty application or from acts adopted on the basis of the Treaty) and noncompulsory expenditure. Compulsory expenditure accounts for around 45% of the EU budget and is mainly used for the CAP. While the European Parliament can – within the limits set by the Financial Perspective12 – overrule the Council’s amendments on non-compulsory expenditure, it has to accept the Council’s

12

The Treaty provides for a maximum rate of increase as a limit on non-compulsory expenditure. Under the Interinstitutional Agreement, the ceilings of the Financial Perspective replace the maximum rate of increase.

150

Henrik Enderlein et al.

prerogative in the domain of compulsory expenditure. And, the European Parliament has the exclusive power to reject the overall budget. For years, the distinction between compulsory and non-compulsory expenditure has been a bone of contention between the European Parliament and the Council (Lindner and Rittberger, 2003). It was introduced in 1970 to limit the budgetary powers of the European Parliament and to ensure the exclusive control of Member States in the Council over legislative decisions with ﬁnancial implications. With that distinction, the Council was able to continue its practice of introducing legally binding entitlements. As the deﬁnition of compulsory expenditure in the Treaty left scope for interpretation, the European Parliament and Council fought intense battles over the classiﬁcation of certain expenditure lines. Since the introduction of the Financial Perspective and the extension of the European Parliament’s legislative powers, the distinction has gradually lost relevance. Close cooperation and a series of formal and informal meetings during the course of the annual procedure give the two arms of the budgetary authority the opportunity to confer over both types of expenditure. A compromise is often found at the conciliation meeting that is held shortly before the second reading in the Council, which is then endorsed by both institutions in their respective readings. As mentioned above, in the United States the President, as the executive, plays a much stronger role, and is the main dividing line in the budgetary process runs between the President and the two chambers of Congress – the Senate and the House of Representatives. Not surprisingly, the current annual budgetary process of the US federal budget has been shaped by a number of swings in the relative strengths of the Presidency and the Congress, which have resulted in a rather complex procedure.13 The ﬁrst step in the US federal budgetary cycle involves the submission by the President of the budget for the following ﬁscal year. Once the President’s budget has reached the legislature, Congress passes its own ‘‘budget resolution’’, which provides a framework within which the different congressional committees will work and includes targets for total spending and revenues. Once Congress has passed its budget resolution, it turns its attention to passing the legal instruments that will, subject to the signature of the President (who retains the power of veto), allow the disbursement of funds. The procedure for such approval depends on the type of expenditure in question. Formally, the US federal budget contains two types of spending categories – discretionary and mandatory – that have some resemblance to the distinction

13

Most notably the 1921 Budget and Accounting Act, by which the President gained a formal role in the federal budget prior to Congress action, and the 1964 Congressional Budget and Impounding Control Act, which provided for Congress to adopt an annual budget resolution setting revenues and spending and also established the Congressional Budget Ofﬁce.

The EU Budget

151

between compulsory and non-compulsory expenditure in the EU budgetary procedure. The President and Congress decide discretionary spending for the following ﬁscal year on an annual basis. Discretionary spending, which currently accounts for around a third of all federal outlays, covers items such as agency budgets, defence programmes, education, foreign aid, etc. Mandatory spending, which accounts for two-thirds of all spending, is authorised by permanent laws. It includes entitlements, such as social security, through which individuals receive beneﬁts because they are eligible on the basis of certain criteria. It also includes interest on the national debt, which the Government pays to individuals and institutions that hold Treasury bonds and other government securities. The President and Congress can change the law in order to change the spending on entitlements and other mandatory programmes – but they do not have to. For discretionary programmes, Congress and the President must act each year to provide spending authority. For mandatory programmes, they may act in order to change the spending that current laws require. On an annual basis, Congress prepares 13 bills to appropriate funds for discretionary expenditure and, if it so chooses, modiﬁes or enacts new authorising bills governing mandatory spending and revenues. In each case, for any authorisation or appropriation to be enacted, ratiﬁcation by the President is required. If the President vetoes a bill, it will be returned to Congress where a two-thirds vote by a quorum of members in each chamber is necessary to override the veto of the executive branch. Conﬂict between the President and Congress can be pervasive from the moment the President’s budget reaches the Congress in February until the beginning of the ﬁscal year in October. The very nature and the high proﬁle of the budget resolution makes it prone to conﬂict between the President and Congress. Congress can pass any resolution but ultimately must gain the President’s approval to enact its proposals. As a result, since 1974, the process and timing of the congressional budget resolution has been broadly different depending on the year. The veto power enjoyed by the President is over entire bills. In other words, the President cannot pick and choose from among the provisions of a particular act. Congress can also combine several appropriation bills with other legislation into one single omnibus measure, thus forcing the President to sign it or veto it as a whole. This tactic can put signiﬁcant pressure on the President, particularly if, as it is sometimes the case, spending bills are presented after the new ﬁscal year has already started and a presidential veto would risk shutting down the Government. Partial shutdowns of federal government have, in fact, occurred in the not too distant past (e.g. 1981, 1984, 1986, 1990 and 1995). Conﬂicts between the executive and legislative branches are aggravated by pork barrel spending, in particular earmarking, contained in bills forwarded by Congress to the President. By incorporating earmarks into the 13 annual spending bills that it passes each year, Congress speciﬁes in law how a certain amount of money be spent, rather than giving the executive branch discretion. In doing

152

Henrik Enderlein et al.

so, provided the President signs the respective bill, Congress manages to keep a tight grip on an important element of government expenditure. Overall, both annual budgetary procedures display a high degree of complexity that stems from the involvement of a wide range of institutions and actors representing different constituencies. In tune with the political system of the United States, the budgetary procedure seeks to strike a balance between the powers of the directly elected President, who heads the executive branch, and the role of Congress, the legislative branch, which represents the interests of voters in the districts and states. In the EU, the largely supranational annual budgetary procedure involves delegates of national governments in the Council and directly elected MEPs. The Commission, as a representative of the executive branch, is a bureaucratic actor that plays a limited role in the ﬁnal stages of the procedure. The ceilings of the Financial Perspective provide a largely intergovernmental framework around the annual procedure that limits the conﬂict between the different actors. A number of proposals for reform have been made with regard to the annual budgetary procedure in the EU. One set of proposals relates to eliminating the distinction between compulsory and non-compulsory expenditure (proposals made by the European Parliament and the Convention’s discussion circle). Such reform would address one of the complexities of the annual budgetary procedure and would acknowledge not only that there is no real difference in nature between the two types of expenditure, but also that in practice the European Parliament and the Council have developed a system of close cooperation, which has signiﬁcantly reduced the relevance of the distinction. Moreover, the rationale that originally lay behind the distinction, namely to accommodate the asymmetry in the distribution of budgetary and legislative powers between the Council and the European Parliament, has largely ceased to apply as the European Parliament has become an important co-legislator. However, it should be kept in mind that abolishing the distinction is closely linked to the question of whether one of the two arms of the budgetary authority should dominate the annual procedure. At the moment, the distinction serves as an instrument for dividing the power to adopt the annual budget equally between the European Parliament and the Council. Should non-compulsory expenditure become the single all-encompassing expenditure classiﬁcation, the European Parliament can reject the Council’s amendments and would thus have the last word on the annual budget. A second set of proposals for reform relates to giving either the Council or the European Parliament the last word on the annual budget (proposals made by the European Parliament, the Convention’s discussion circle and the Constitution – all of these favour giving the last word to the European Parliament). The current system with two arms of the budgetary authority having two readings of the annual budget can be streamlined by having only one reading, and giving the last word on the annual budget to one of the two arms. Such a

The EU Budget

153

reform would help increase the efﬁciency of the annual budgetary procedure, especially as the Council’s ﬁrst readings seem to mainly serve presentational purposes. Moreover, the reform may also increase the coherence of annual budgets, as one arm of the budgetary authority would assume responsibility for the budgets as a whole and would be freed from ﬁnding detailed compromises with the other institution. For reasons of democracy,14 the European Parliament would seem to be more suited than the Council to take over this position, especially as the European Parliament already has the right to reject the overall budget. However, it is important to bear in mind that, taking into account legitimacy concerns, the current balance between the Council and the European Parliament helps to ensure the acceptance of annual budgets among MEPs, as well as among Member States, whose contributions form the EU budget. Therefore, proposals to grant the European Parliament (or the Council) the last word over the annual budget may need to be embedded in a structure of close cooperation, building on the existing channels between the European Parliament and the Council. The conciliation committee speciﬁcally should play a prominent role, and ensure that the position of the Council (or the European Parliament) is duly taken into account. The theoretical literature on efﬁciency and legitimacy provides no clear answers to the question about the ﬁnal decision-taking authority in the annual budgetary procedure. The speciﬁc set-up of the EU institutions makes it difﬁcult to attribute the traditional roles of ‘‘executive’’ and ‘‘legislature’’ to the various players involved in the annual budgetary procedure. It is clear, however, that a joint decision mode, bringing together two or more actors in the budgetary procedure, would face the ‘‘common pool problem’’. In this context, a two-stage approach with either the European Parliament or the Council taking the ﬁnal decision on the annual budgetary procedure would appear more desirable, in theory. This point also derives from the efﬁciency literature, which clearly favours a streamlined process with a clear allocation of power resources. 4. The state of play following the Intergovernmental Conference15 Against the background of these different proposals and on the basis of the draft Constitution of the Convention, representatives at the IGC negotiated on the possible institutional reforms in the budgetary ﬁeld. The European Constitution, which was ﬁnally adopted by Heads of State or Government in June 2004 and

14

As MEPs campaign on European platforms and are directly elected to decide on European issues, it may be argued that their democratic legitimacy is stronger than that of national governments that are chosen (in most Member States) by national parliamentarians who are elected largely on the basis of national issues. 15 This section is based on Laffan and Lindner (2005).

154

Henrik Enderlein et al.

formally signed in Rome on 29 October 2004, introduces a budgetary procedure that remains de facto close to the current budgetary decision-making process – even though de jure the two procedures may appear very different. In fact, the main innovations that the ‘‘new’’ budgetary procedure entails are taken from the rules and procedures that are currently laid down in the Interinstitutional Agreement. Thus, it simply institutionalises existing informal arrangements and does not enact institutional change. For the annual procedure, three innovations are introduced (see Figure 3): (i) the abolition of the distinction between compulsory and non-compulsory expenditure, (ii) the introduction of a Conciliation Committee into the Constitution, and (iii) the granting of the right of rejection to the Council. All three correspond closely to current practice. In the ﬁrst case, the distinction between the compulsory and non-compulsory expenditure has lost its relevance over time. The Interinstitutional Agreement already gives the European Parliament some say over compulsory expenditure through the ad hoc conciliation procedure. Second, the equivalent of a Conciliation Committee is already in place. The annual budget is mostly adopted de facto in a conciliation meeting between the Council and the European Parliament shortly before the second reading in Council. Third, given that negotiations at the conciliation meeting cover all areas of the budget, the exclusive budgetary powers over the non-compulsory expenditure for the European Parliament and the compulsory expenditure for the Council have de facto amounted to a right of rejection for both arms of the budgetary authority. Both the Council and the European Parliament use their budgetary powers as bargaining chips to strike deals over the different parts of the budget. Under the new procedure, the Council will be able to prevent an agreement in the Conciliation Committee and thus trigger a new budget proposal by the Commission. The granting of the right of rejection to the Council simply maintains the current balance. With regard to the Financial Perspective, the new provisions largely institutionalise the current procedures for the multi-annual budget plan. From the outset, it was clear that the institutionalisation of the Financial Perspective would probably be a minimum result of the constitutional negotiations. Given the objectives of the constitutional process (i.e. to update and streamline the Treaty), the contrast between the Treaty provisions and current practice was simply too pronounced in this area of budgetary decision-making for the drafters of the Constitution to ignore it. However, any attempts to go beyond institutionalising the current Interinstitutional Agreement were strongly opposed. Although the Convention settled on the introduction of qualiﬁed majority voting for the Financial Perspective after 2013, the Heads of State or Government adopted a provision that allows for the introduction of qualiﬁed majority voting only on the basis of a unanimous decision by the European Council. Similarly, the proposed provisions concerning the revenue side do not alter the unanimity requirement for own resources decisions.

The EU Budget

155

Figure 3. The amended annual budgetary procedure in the Constitutional Treaty (Article III-404)

156

Henrik Enderlein et al.

Overall, the IGC settled on reforming the current budgetary procedure only to the extent that proven rules and procedures from outside the Treaty are brought into the Constitution. Taking into account the close link between the overall state of integration in the EU and procedural issues related to budgetary matters, it is not surprising that constitutional negotiators were risk-averse with regard to changes to the budgetary procedures that would have entailed signiﬁcant consequences for the general set-up of the EU.

5. Conclusion This paper has discussed the question why, despite widespread calls for reforms of the EU budgetary procedure, the European Convention and, even more importantly, the Intergovernmental Conference seem to have generated only minor institutional adjustments in this area. The analysis of the EU budgetary procedure and various reform proposals reveals that the current institutional design corresponds by-and-large to an equilibrium between all the actors involved. Altering that equilibrium would require a shift in the level of integration. The ‘‘embeddedness’’ of the budgetary ﬁeld in the overall state of European integration thus constrains the scope of reform. Given that neither the European Convention nor the IGC focused on questions related to the general state of political integration, it is not surprising that the calls for far-reaching reforms of the EU budgetary procedure have not resulted in the desired institutional change. While it is beyond the scope of this paper to consider the pros and cons of changes in the state of political integration, a number of elements are crucial in demarcating the room for manoeuvre for the EU budgetary procedure. Such key elements, which we call dimensions of political integration, are (i) the assignment of tasks to the different levels of government, (ii) the balance of power between the various EU actors and the corresponding voting modalities, and (iii) the degree to which citizens identify themselves with the EU and with the politicians who are supposed to represent them at that level. These conclusions should not be read as a panglossian view of current EU budgetary procedures, but rather as a cautionary tale about the limited impact that can be expected from a simpliﬁcation of budgetary procedures that does not change the broader EU institutional and political set-up. To fundamentally change current EU budgetary outcomes would require the modiﬁcation of the current set of broader constraints in which any EU budgetary procedure is necessarily embedded. Such constraints are naturally not set in stone: the future will show whether recent and current signiﬁcant advances in European integration – especially in the area of monetary integration – will one day have helped in shifting out the symbolical ‘‘Pareto frontier’’ of political integration, thus also contributing to stronger integration in the EU budget.

The EU Budget

157

Acknowledgement The opinions expressed in this paper are those of the authors and do not necessarily reﬂect those of the European Central Bank. We thank Pierre van der Haegen, Pierre Petit, Theodor Martens, Hedwig Ongena, Demosthenes Ioannou, Wouter Coussens, Helge Berger, Magarita Katsimi, Ge´rard Roland, two anonymous referee, and the participants of two CESifo-Delphi Conferences, ‘‘Designing the New EU’’, in Munich and Delphi, and of the 14th International Conference of Europeanists in Chicago for their helpful comments.

References Alesina, A. and E. Spolaore (1997), ‘‘On the number and size of nations’’, Quarterly Journal of Economics, Vol. 107(4), pp. 1027–1056. Begg, I. (1999), ‘‘Reshaping the EU budget: yet another missed opportunity?’’, South Bank European Paper, No. 5/99, South Bank University. Begg, I. and N. Grimwade (1998), Paying for Europe, Shefﬁeld: Shefﬁeld Academic Press. Brehon, N.-J. (1997), Le budget de l’Europe, Paris: LGDJ. Buchanan, J.M. and G. Tullock (1962), The Calculus of Consent, Ann Arbor: University of Michigan Press. Congressional Budget Ofﬁce (2003), The Budget and Economic Outlook: Fiscal Years 2004-2013, US Government Printing Ofﬁce, January (available at http://www.cbo.gov/ftpdoc.cfm?index=4032&type=1). European Commission (2002), European Union Public Finance, Luxembourg: Ofﬁce for Ofﬁcial Publications of the European Communities. European Convention (2003a), Final report of the discussion circle on the budgetary procedure, CONV 679/03, April 14. European Convention (2003b), Draft Treaty establishing a Constitution for Europe, CONV 820/1/03 REV 1, CONV 847/03, CONV 848/03, July 18. European Parliament (2003), Report on the reform of the budgetary procedure: possible options in view of the revision of the treaties, A5-0046/2003, February 20. Folkers, C. (1997), ‘‘Finanz- und Haushaltspolitik’’, pp. 561–663 in: P. Klemmer, editor, Handbuch der Europa¨ischen Wirtschaftspolitik, Munich: Verlag Franz Vahlen. Hallerberg, M. and J. von Hagen (1999), ‘‘Electoral institutions, cabinet negotiations, and budget deﬁcits within the European Union’’, pp. 209–232 in: J. Poterba and J. von Hagen, editors, Fiscal Institutions and Fiscal Performance, Chicago: University of Chicago Press. Laffan, B. (1997), The Finances of the European Union, Basingstoke: Macmillan Press.

158

Henrik Enderlein et al.

Laffan, B. and J. Lindner (2005), ‘‘The EU budget’’, in: M. Pollack, H. Wallace and W. Wallace, editors, Policy-Making in the European Union5th edition, Oxford: Oxford University Press. La¨ufer, T. (1990), Die Organe der EG – Rechtsetzung und Haushaltsverfahren zwischen Kooperation und Konflikt, Bonn: Europa Union Verlag. Leonardi, R. (1999), The Socio-economic Impact of Projects Financed by the Cohesion Fund: A Modelling Approach3 vols, Luxembourg: Ofﬁce of Ofﬁcial Publications of the European Communities. Lindner, J. (2003), ‘‘Institutional stability and change: two sides of the same coin’’, Journal of European Public Policy, Vol. 10(6), pp. 912–935. Lindner, J. and B. Rittberger (2003), ‘‘The creation, interpretation and contestation of institutions – revisiting historical institutionalism’’, Journal of Common Market Studies, Vol. 41(3), pp. 445–473. Marx, G. and L. Hooghe (2003), ‘‘National identity and European integration. A multi-level analysis of public opinion’’, Unpublished paper (available at http:// www.unc.edu/hooghe/downloads/national_identity_25march2003.pdf). Monar, J. (1994), ‘‘Interinstitutional agreements: the Phenomenon and its new dynamics after Maastricht’’, Common Market Law Review, Vol. 31, pp. 693–719. Nava, M. (2000), La finanza europea, Stocia aualisi e perspettive della politica fiscale europea, Rome: Carocci. Oates, W.E. (1999), ‘‘An essay on ﬁscal federalism’’, Journal of Economic Literature, Vol. XXXVII, pp. 1120–1149. Padoa-Schioppa, T. (2002), ‘‘A European economic constitution’’, The Federalist – A Political Review, Vol. XLIV(3), pp. 190–202. Persson, J., G. Roland and G. Tabellini (1997), ‘‘Separation of powers and political accountability’’. The Quarterly Journal of Economics, pp. 1163–1202. Persson, J., G. Roland and G. Tabellini (2000), ‘‘Comparative politics and public ﬁnance’’, Journal of Political Economy, Vol. 108, pp. 1121–1161. Persson, J. and G. Tabellini (2003), The Economic Effects of Constitutions, Boston: MIT Press. Pollack, M.A. (2003), The Engine of European Integration – Delegation, Agency, and Agenda Setting in the EU, Oxford: Oxford University Press. Sapir, A., P. Aghion, G. Bertola, M. Hellwig, J. Pisani-Ferry, D. Rosati, J. Vinals and H. Wallace (2003), An Agenda for a Growing Europe. Making the EU Economic System Deliver, July (available at http://www.euractiv.com/ ndbtext/innovation/sapirreport.pdf). Also forthcoming with Oxford University Press. Scharpf, F.W. (1988), ‘‘The joint-decision trap: lessons from German federalism and European integration’’, Public Administration, Vol. 66, pp. 239–278. Scharpf, F.W. (1999), Governing Europe. Effective and Democratic?, Oxford: Oxford University Press. Schick, A. (2000), The Federal Budget: Politics, Policy, and Process, Washington, DC: Brookings Institution.

The EU Budget

159

Shackleton, M. (1990), Financing the European Community, London: The Royal Institute of International Affairs, Pinter Publishers. Strauss-Kahn, D. et al. (2004), Construire l’Europe Politique – 50 Propositions pour l’Europe de Demain, April (available at http://europa.eu.int/comm/dgs/ policy_advisers/experts_groups/gsk_docs/rapport_europe_strauss_kahn_fr.pdf). Tabellini, G. (2002), ‘‘Principles of policymaking in the European Union: an economic perspective’’, Paper presented at the Munich Economic Summit, June. von Hagen, J. (1992), ‘‘Budgeting procedures and ﬁscal performance in the EC’’, European Commission, Directorate for Economic and Financial Affairs, Economic Papers No. 96. von Hagen, J. and I.J. Harden (1994), ‘‘National budget processes and ﬁscal performance’’, European Economy Reports and Studies, Vol. 3, pp. 315–418. Wallis, J.J. (2000), ‘‘American government ﬁnance in the long run: 1790 to 1990’’, Journal of Economic Perspectives, Vol. 14(1), pp. 61–82.

This page intentionally left blank

Comment Margarita Katsimi

This very informative paper provides an overview as well as an assessment of the current reform proposals of the EU budgetary process. The authors note that although supranational governance is more dominant in certain policy ﬁelds, the present institutional set-up of the EU strikes a balance between legitimacy and efﬁciency by combining elements of supranational governance and intergovernmental co-operation. An important point of the paper is that reforms in the decision-making procedure should be assessed in the framework of the general state of integration. This is depicted in Figure 1 of Enderlein et al.’s paper.

Fiscal Federalism

Figure 1.

Accountability CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79013-7

r 2007 PUBLISHED BY ELSEVIER B.V.

162

Margarita Katsimi

According to the current budgetary procedure, we are at point B on the ‘metaphoric’ Pareto frontier, which is determined by the state of integration. For an exogenous level of integration, any reform will either increase efﬁciency at the expense of legitimacy (point B0 ), or it will increase legitimacy at the cost of lower efﬁciency (point B00 ). The authors argue that the EU is currently at point A, which implies that the current procedure provides the best possible combination of efﬁciency and legitimacy given the level of political integration. Table 1 summarises the authors’ assessment of the various reforms proposed. My first point is that this assessment should be discussed as a package. In several cases, the main costs of the reform would not apply if they are combined with some of the other reforms proposed. For example, the authors fear that the introduction of the Financial Perspective (FP) in the Treaty may increase the status quo bias. This fear will, however, be weakened if the FP is adopted by qualiﬁed majority voting (QMV) as reforming the voting rule will facilitate changing the ceilings. In turn, let us assume that QMV for the FP is adopted. One of the disadvantages suggested is that this reform will give an incentive to outvoted Member States to carry their discontent into the annual procedure. However, reforms of the annual budgetary procedures that give less power to the Council may discourage such behaviour. The same consideration holds true for the beneﬁts of various reforms: Giving more power to the European Parliament (EP) in the annual budgetary procedure may prove to be more costly if the FP becomes less binding (e.g. by introducing QMV). Though the Members of the European Parliament (MEP) get credit for expenditures agreed by the EP, the related costs are actually borne by the Council; this may lead to a situation of ﬁscal illusion (e.g. the marginal beneﬁt of a public activity is overestimated). In such a situation, MEP will be encouraged to pursue the interests of their political constituency without taking the social beneﬁt into account. Retaining the unanimity rule for the FP in this case may be optimal as ﬁscal performance improves under a procedure-oriented approach (see Von Hagen and Harden, 1995). My second comment relates to the general nature of the reforms proposed. A common characteristic among most of the reform proposals is that they aim to increase the supranational element of the union. For that reason, stressing the main advantages and disadvantages of moving from intergovernmental governance to supranational governance in the ﬁeld of budgetary policy is important in my opinion. A federalist structure implies beneﬁts (e.g. the coinsurance effect, economies of scale in public goods provision and internalizing externalities at a lower cost) and costs (e.g. information costs and single policy costs) that should be mentioned (see e.g. Persson and Tabellini (1995), Inman and Rubinfeld (1996) and Oates (1999) for surveys). Although at the beginning of Enderlein et al.’s paper there is some discussion about the pros and cons of intergovernmental co-operation versus supranational governance, it is not focused on budgetary policy. Third, I think that an important implication of the paper – that is not explicitly mentioned – is that in the EU, there is a trade-off between supranational governance and optimal institutional framework for the budgetary process. This

Comment

Table 1.

Reform proposals

Reforms

General level Modify the expenditure structure Creation of an ‘‘EU tax’’

163

Assessment Pros

Cons

Improves efﬁciency by allowing the provision of some public goods. Conveys a strong and positive symbolic message and enhance transparency of EU contributions.

Requires willingness to allocate further tasks to the EU. Requires willingness to allocate further tasks to the EU.

Allows the emergence of a ‘common pool’ problem. Financial perspective Adoption by qualiﬁed majority

Reduces status quo bias and increases efﬁciency.

Introduction into the Treaty and making the ceiling legally binding.

Closes the gap between the Treaty provisions and current practices.

Adoption in the Council of Ministers

More in line with the annual budgetary procedure.

Synchronizing the time-frame with the EP term of ofﬁce

Enhances legitimacy.

Increasing the ﬂexibility of the expenditure headings

Improves ﬂexibility.

Avoids the cost of reaching an agreement about the ceilings. Annual budgetary procedure Eliminating distinction between compulsory and noncompulsory expenditure Giving either the Council or the EP the last word

-May lead to sovereignty concerns by Member States (MS). -Encourages the outvoted MSs to carry their discontent into the annual procedure Undermines the conﬂict reducing character of the FP. May increase the status quo bias. Increases vulnerability to the pressure exerted by other spending Councils. Does not allow solving budgetary and non-budgetary issues in a package agreement. The beneﬁts are undermined by the weak role of the European Parliament in the negotiations for the FP. Reduces the efﬁciency of the procedure to limit conﬂict among agents. Allows ‘common pool’ problems to emerge.

Close to current practices.

EP will have the last word on the annual budget.

Increases the efﬁciency of the annual process and the coherence of annual budgets. Avoids the ‘common pool’ problem.

Creates legitimacy concerns if the position of the other party is not taken into consideration

164

Margarita Katsimi

trade-off is depicted in Figure 1. Speciﬁcally, under the current political structure, strengthening supranational governance in the ﬁeld of budgetary policy will lead to reforms that may lead to a loss in accountability. This loss will emerge mainly because of reforms that aim to strengthen the role of the EP, thereby leading to the elimination of checks and balances, and allowing the emergence of a ‘common pool’ problem (see Persson et al., 1997). Though the political structure in the US can combine the beneﬁts of ﬁscal federalism with an institutional framework for the budgetary process that ensures accountability; in the EU, any beneﬁt from a more federal structure that assigns more responsibilities to supranational authorities will lead to accountability being lost. My ﬁnal comment relates to the assumption of an exogenous political integration process. Is the level of political integration exogenous, or can the proposed budgetary reforms initiate the need for a future reform of the political structure? If this is the case, then one should consider whether a higher level of political integration is optimal in economic terms. This question has been investigated by recent literature on the optimal number of nations (see e.g. Bolton and Roland (1997), Alesina and Spolaore (1997), Casella (2001) and Alesina et al. (2001)). The issue of the appropriate size of government and the number of nations is quite large and to a large extent is beyond the scope of the paper. However, if one believes that stronger political integration in the EU is desirable, the proposed budgetary reforms may be seen as a small ﬁrst step in this direction. References Alesina, A., I. Angeloni and F. Etro (2001), ‘‘The political economy of international unions’’, NBER Working Paper, No. 8645. Alesina, A. and E. Spolaore (1997), ‘‘On the number and size of nations’’, Quarterly Journal of Economics, Vol. 112, pp. 1027–1056. Bolton, P. and G. Roland (1997), ‘‘The breakup of nations: a political economy analysis’’, Quarterly Journal of Economics, Vol. 112, pp. 1057–1090. Casella, A. (2001), ‘‘The role of market size in the formation of jurisdictions’’, Review of Economic Studies, Vol. 68, pp. 83–108. Inman, R. and D. Rubinfeld (1996), ‘‘Designing tax policy in federalist economies: an overview’’, Journal of Public Economics, Vol. 60, pp. 307–334. Oates, W. (1999), ‘‘An essay on ﬁscal federalism’’, Journal of Economic Literature, Vol. 37, pp. 1120–1149. Persson, T. and G. Tabellini (1995), ‘‘Double-edged incentives: institutions and policy coordination’’, in: G. Grossman and K. Rogoff, editors, Handbook of International Economics, Vol. 3, Amsterdam: North-Holland. Persson, T., G. Roland and G. Tabellini (1997), ‘‘Separation of powers and political accountability’’, Journal of Political Economy, Vol. 112, pp. 312–327. Von Hagen, J. and I. Harden (1995), ‘‘Budget processes and commitment to ﬁscal discipline’’, European Economic Review, Vol. 39, pp. 771–779.

Part III: Fiscal and Financial Policies

This page intentionally left blank

CHAPTER 6

From the Stability and Growth Pact to a Sustainability Council for EMU$ Ju¨rgen von Hagen

1. Introduction: Europe’s fiscal framework under stress The European Monetary Union (EMU) includes a new framework for the ﬁscal policies of its member states. The need to create a genuine institutional framework to deal with public ﬁnances in the EMU was recognized in the Delors Report (1989), which called for institutional safeguards for ﬁscal discipline in the monetary union, arguing that a lack of ﬁscal discipline might undermine the stability of the new currency. Art. 4(3) of the Treaty on European Union (TEU) – i.e., the Maastricht Treaty and its successor, the Amsterdam Treaty – declares that ‘‘sound public ﬁnances’’ are a guiding principle of economic policy in the EU. Procedures of the EU with relevance to the conduct and coordination of ﬁscal policy are the Mutual Surveillance Procedure (Art. 99), the ‘‘No-bail-out clause’’ (Art. 103), the Excessive Deﬁcit Procedure (EDP, Art. 104), and the Stability and Growth Pact (SGP, Council Regulations 1466/97, 1476/97, Council Resolution 97/C236/ 01-02). Art. 99 holds that the member states of the EU regard their economic policies as a matter of common concern and coordinate them through the Council of Economic and Finance Ministers (Ecoﬁn) and on the basis of ‘‘broad economic guidelines’’. The no-bail-out rule relieves the Community and the

$ This paper draws from material presented by Antonio Fatas, Andrew Hughes Hallett, Rolf Strauch, Anne Sibert, and Ju¨rgen von Hagen in ‘‘Stability and growth in Europe: towards a better pact’’, Monitoring European Integration, Vol. 13, London: CEPR. An earlier version of this paper was presented at the CES-ifo conference in Delphi, ‘‘Designing the new EU’’, June 4–5, 2004. I am grateful to Helge Berger, Wolfgang Eggert, and the conference participants for useful comments.

CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79006-X

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

168

Ju¨rgen von Hagen

member states of responsibility for ﬁnancial liabilities of other members against their will. The EDP sets up a detailed process of monitoring the public ﬁnances of the member states with a view to ensuring that they remain sustainable. The SGP reﬁnes and concretizes the procedures of the EDP. This framework has been tested seriously in the economic downturn after 2001. An intense public debate over possible reforms has followed. In this paper, we review the background of this debate and develop our own proposal for reform. In section 2, we begin with a discussion of the institutional framework. In section 3, we turn to the empirical performance under this framework, focusing in particular on ﬁscal discipline and growth in the EMU. Section 4 presents our reform proposal, the creation of an independent Sustainability Council for the EMU. The ﬁnal section concludes. 2. Fiscal discipline in the EMU A basic belief underlying the framework of the EMU is that the stability of the common currency requires the sustainability of public ﬁnances. The fear that high and rising public debts would undermine the central bank’s ability to deliver price stability has left its mark on all important documents and political decisions on the way to the EMU. In terms of technical economic analysis, ﬁscal policy and monetary policy are indeed linked through the ‘‘intertemporal budget constraint’’, the requirement that, in the long run, the discounted sum of a government’s expected expenditures cannot exceed the discounted sum of its expected revenues.1 Given an expected stream of expenditures, governments in the EMU must adjust taxes to ensure that the intertemporal budget constraint holds, since they cannot use the printing press freely to monetize their debts. Otherwise, they will be forced at some point to default on their debts. A ﬁscal crisis would arise, and pressures would rise on the European Central Bank (ECB) to bail out the troubled government.2 It is far from certain that the ECB’s institutional independence would be sufﬁciently strong to withstand such pressures and put price stability ﬁrst. This reasoning leads to the conclusion that the EMU needs rules preventing the national governments from running up excessive debts that could threaten its common good, i.e., price stability. The tricky question is how to include this conclusion in a framework that guides and constrains the governments’ ﬁscal policies in the short run. Since the governments’ intertemporal budget constraint pertains to the long run, it has no strong implications for today’s budgetary

1

See, e.g., Sargent and Wallace (1981). Note that the ECB cannot legally bail out a government in ﬁscal difﬁculties by buying its debt directly, but it can still do so indirectly if it wants to. A bailout could be ex post, with the central bank buying up large amounts of government debt in the market, or ex ante, with the central bank holding down interest rates to reduce the government’s interest payments.

2

From the Stability and Growth Pact to a Sustainability Council for EMU

169

policies and ﬁscal ﬂows.3 In a world with perfect information and no transactions costs, one could adopt a ﬁscal policy rule stating in detail what governments should do under what circumstances to meet the intertemporal budget constraint. In reality, the world is too complex and uncertain. A simple ﬁscal rule limiting annual deﬁcits or debts of governments is of little use under such circumstances as it would constrain their ﬁscal policies either too much or too little in the short run. Either way, it would lack credibility: in the ﬁrst case, because it may force sovereign governments under some circumstances to adopt policies that are unreasonable or can damage their countries, in the second case, because it would not bind government actions sufﬁciently in the short run. Furthermore, simple rules are not adequate to assure sustainability, as in the European context of supranationality, a rule must treat all member states equally even if they are unequal. For example, a rule for the annual budget deﬁcit ratio must be based on an assumption about the long-run nominal growth trend of the member economies if it is to stabilize the debt ratio, G. Given a rate of inﬂation in the euro area of, say, 2 percent annually, the allowable deﬁcit ratio is d ¼ (a+0.02)G. With G ¼ 0.6, the allowable ratio is 4.8 percent for a country such as Finland, which is growing at 6 percent on average, but only 1.8 percent for Germany, whose trend growth rate is about 1 percent. Keeping Germany’s debt ratio at 60 percent would require a deﬁcit limit of 2.28 percent of the gross domestic product (GDP), which would bring Finland’s debt ratio to 28.5 percent in the long run. An important distinction exists between sustainable public ﬁnances and optimal public ﬁnances. Optimal public ﬁnances are the solution to an optimization problem that consists of a set of policy goals, political preferences regarding policy outcomes, resource constraints, and (assumptions about) laws describing the functioning of the economy. Designing optimal policies is, therefore, a political task. Sustainability is just one of the resource constraints that must be fulﬁlled in this task, i.e., all optimal policies are sustainable, but not all sustainable policies are optimal. This distinction is relevant for two reasons. First, governments in the EMU may choose to coordinate their ﬁscal policies to achieve more efﬁcient, jointly optimal outcomes. But policy coordination is logically different from sustainability, and is a political process requiring agreement on policy choices. Countries in the EMU may or may not decide to coordinate sustainable policies, but sustainability itself is neither a reason nor a justiﬁcation for coordination. Second, the more a framework is meant to achieve sustainability constrains short-term ﬁscal policies, the more likely it is to get in the way of optimal policy choices in the short run. Too much emphasis on the short run, therefore, will politicize the framework for sustainability to an

3

This is best seen in the fact that governments can always promise future actions to collect more revenues to compensate for current deﬁcits. See Perotti et al. (1998) for a more detailed discussion.

170

Ju¨rgen von Hagen

unnecessary extent. The proper response is to design a framework that combines guidelines for short-run budgetary policies, with proper judgment about current and future developments. 2.1. The Excessive Deficit Procedure and the Stability and Growth Pact The EDP is the cornerstone of the ﬁscal framework of the EMU. It combines the unconditional obligation on the part of the member states to avoid ‘‘excessive deﬁcits’’ with a procedure providing for a regular assessment of ﬁscal policies in the EMU and, if necessary, penalties for proﬂigate behavior (Art. 104 of TEU). The TEU charges the European Commission with the task of monitoring budgetary developments and public sector debt of the member states, checking in particular their compliance with two reference values for the ratio of the deﬁcit to GDP and the ratio of public debt to GDP. The two reference values are set at 3 and 60 percent, respectively (Protocol on the EDP). If a member state does not comply with these reference values, and unless the deﬁcit and the debt are approaching their reference values in a satisfactory way or the excess of the deﬁcit over the limit is exceptional and temporary, the Commission sends a report to the European Council, after ﬁnding out whether the deﬁcit exceeds public investment spending and taking into account ‘‘all other relevant factors, including the medium-term economic and budgetary position’’ (Art. 104(3)) of the country concerned.4 If the Commission feels that an excessive deﬁcit exists, it makes a recommendation to the European Council, which votes on it by qualiﬁed majority after taking note of any observations by the country concerned and the opinion of the Economic and Financial Committee (EFC), which advises the Council in these matters (Art. 114). The decision whether or not an excessive deﬁcit indeed exists is made by Ecoﬁn. In the context of the EDP, the numerical criteria for deﬁcits and debts thus serve as triggers for an assessment prepared by the European Commission and made by the European Council. They do not themselves deﬁne what an excessive deﬁcit is, nor does breaching them imply any sanctions per se. Since they merely serve as triggers for a more precise assessment of the situation, there is no need to make the criteria themselves responsive to economic circumstances, e.g., by redeﬁning them to exclude interest spending or cyclical effects on spending and revenues. These and other circumstances can be considered in the Commission’s analysis, the EFC’s opinion, and the Council’s judgment. In view of the need to balance long-term objectives with short-run constraints on actual policy, such a trigger-role is appropriate for the numerical criteria. However, the EDP suffers from a serious credibility problem. By assigning to Ecoﬁn the right to decide whether or not an excessive deﬁcit exists, the EDP

4

According to Art. 104(3), the Commission may also prepare a report if a member state complies with the criteria but the Commission sees the risk of an excessive deﬁcit nevertheless.

From the Stability and Growth Pact to a Sustainability Council for EMU

171

effectively makes a group of ‘‘sinners’’ judge the performance of fellow ‘‘sinners’’. Considering the ﬁscal performance of other governments, Council members have reason to be lenient and avoid actions that are politically costly for fellow members, anticipating that they might be in a similar position in the future. This makes political deals more likely than serious judgment and the application of sanctions by Ecoﬁn. The SGP modiﬁes the EDP in several ways.5 First, it sets up an early warning system, strengthening the surveillance of the public ﬁnances of the member states. Under the SGP, EMU member states submit annual Stability Programs to the European Commission and the European Council, explaining their intended ﬁscal policies and, in particular, what they plan to do to keep the budget close to the new and stricter medium-term objective of ‘‘close to balance or in surplus.’’ Implementation of these programs is subject to the scrutiny of the Council, which, based on information and assessments by the Commission and the EFC, can issue early warnings to countries that risk signiﬁcant deviations from the ﬁscal targets set out in their Stability Programs. The goal of the Programs is to achieve and maintain budgetary positions close to balance or in surplus. Second, the SGP clariﬁes the EDP by giving more speciﬁc content to notions of exceptional and temporary breaches of the 3 percent limit and by deﬁning the rules for ﬁnancial penalties, and speeds up the process by setting deadlines for individual steps. Third, the SGP gives political guidance to the parties involved in the EDP, urging them to implement the rules of the EDP effectively and in time. It commits the Commission, in particular, to using its right of initiative under the EDP ‘‘in a manner that facilitates the strict, timely, and effective functioning of the SGP.’’ This puts severe limits on the Commission’s right to exercise judgment in each case and situation, resulting in shifting that right to the Council. The rules of the SGP have been developed in a set of Ecoﬁn decisions on the format and content of the Stability Programs.6 In October 1998 Ecoﬁn endorsed a Monetary Committee (the precursor of the Monetary and Financial Committee) opinion, a ‘‘code of conduct’’ specifying criteria to be observed in the assessment of a country’s medium-term budgetary position, and data standards and requirements for the Programs. In October 1999, Ecoﬁn recommended stricter compliance with, and timelier updating of, the Programs. In July 2001 Ecoﬁn endorsed an appended code of conduct proposed by the EFC, reﬁning the format and the use of data in the Stability Programs, including a common set of assumptions about economic developments outside the EMU. The Commission

5 6

For an account of the genesis of the SGP, see Stark (2001). See European Commission (2002 p. 23).

172

Ju¨rgen von Hagen

(2000) has speciﬁed a detailed framework of interpretation of divergences from the targets set in the Stability Programs. Compared with the original EDP, the SGP has achieved two things. First, it has changed the nature of the ﬁscal framework from one based on informed judgment to a simple numerical rule restricting annual deﬁcits. Second, while the Maastricht Treaty gave the Commission considerable discretion in initiating the EDP and moving it forward, the SGP has reduced the Commission’s role and increased the importance of the Council’s judgments and decisions. In doing so, the SGP has shifted the balance of power in the ﬁscal policy framework from the institutional guardian of the Treaty to the representatives of the member states, thus politicizing the process and the relevant decisions more than the original EDP. Both changes have further eroded the credibility of the ﬁscal framework. Anticipating that the Council will tend to avoid serious judgment and the application of penalties, the European public will naturally dismiss arguments brought forth by the Council explaining why individual violations of the criteria and provisions should be accepted under special circumstances, even if these arguments are justiﬁed. As a result, the SGP has reduced the weight of sound economic judgment and increased the emphasis the public puts on the numerical criteria. Public opinion and ﬁnancial markets have begun to take the numerical deﬁcit rule more seriously than a sensible economic procedure would warrant. This has made the entire process more rigid than necessary. As a result, the ﬁscal framework is increasingly perceived as a straitjacket for policies, especially so in the large EMU countries. There is now a serious risk that the SGP will cause the opposite of what it intended: as the large EMU countries have decided – more or less openly – not to accept that perceived straitjacket any longer and ignore the rules, the EMU could be left with less protection against ﬁscal proﬂigacy.

2.2. Proposals for SGP reform Numerous reform proposals for the SGP have been presented after the framework came under stress in 2001. These proposals can be categorized in three groups: some authors insist that the current numerical rules must be preserved and the procedures strengthened to avoid crises. Others argue that the numerical rules be retained but with increased ﬂexibility. The Commission (2002) and Buti et al. (2003) propose to focus on structural balances rather than the actual deﬁcit, a suggestion that was incorporated in the 2001 Code of Conduct. Giavazzi and Blanchard (2002) call for a ‘‘golden rule’’ excluding investment spending from the budget. The German government even called for excluding member countries’ contributions to the EU from calculation of the deﬁcit. Coeure´ and Pisani-Ferry (2003) recommend that the emphasis be shifted to the debt ratio, and that countries with debt ratios below 40 percent be excused from

From the Stability and Growth Pact to a Sustainability Council for EMU

173

exceeding the 3 percent line for the deﬁcit ratio.7 Finally, there is the Commission’s proposal to increase coordination in return for greater ﬂexibility. None of these proposals will create a better framework for sustainability. Our analysis above suggests that the current crisis of the SGP has three causes. First, there is too much focus on numerical criteria for short-term ﬁscal ﬂows, although the goal relates to the long run only. Increasing ﬂexibility in a few ways suggested by the current proposals may provide some relief now, but it will not solve the basic problem of how to translate the long-run budget constraint into meaningful guidelines for current policies. That is, situations in which governments claim that the constraints are too rigid will recur even if some ﬂexibility is introduced now. But repeated revisions of the framework tailored to the policy problems of the day will damage its credibility. Second, the EMU has been unable to use the ﬂexibility already provided by the current framework, which is owing to the fact that the public does not regard governments as credible judges of their own policies. Softening the deﬁcit criteria would not solve that problem; on the contrary, it would be perceived by the public as an attempt to give more room for fudging budget data. Giving the Commission a greater role in monitoring budgetary developments would also not solve the problem; as a multitask organization that relies heavily on the cooperation of the member states in achieving its goals, it is too politically dependent on the member states to have the required credibility. 3. Fiscal performance under the EDP and the SGP 3.1. Experiences in the 1990s In 1992 the EU’s average debt ratio was almost 60 percent of GDP – hence the 60 percent limit foreseen in the Maastricht Treaty.8 It climbed to almost 75 percent in 1997, the base year for the May 1998 decision on which countries could enter the EMU. Since 1997, the average debt ratio has fallen to 63 percent. In contrast to what EU ofﬁcials and politicians like to say, the data do not show that the Maastricht process for ﬁscal consolidation was successful. Several qualiﬁcations apply. First, the increase in the average debt ratio was driven mainly by debt expansions in ﬁve states: Germany (from 44 percent to 61%), France (from 40 to 56 percent), Spain (from 48 to 70 percent), Italy (from 109 to 124 percent), and the UK (from 42 to 55 percent). While Belgium and Luxembourg almost stabilized their debt ratios, the Netherlands and Ireland enjoyed falling debt ratios during this period. The debt ratios of the other states

7

Note that this would not provide any relief for France, Germany, and Italy. The 3 percent deﬁcit limit under the EDP derives from the 60 percent debt limit, assuming an average nominal GDP growth rate of 5 percent in all EMU member states.

8

Ju¨rgen von Hagen

174

Table 1. Government debt since 1992 Period

1992–1997 1998–2003

Change in debt ratio (percent) All EU states 15.8 4.7

Large states 18.8 4.9

Intermediate states 4.1 10.5

Small states 3.3 7.1

Source: European Economy Statistical Appendix Spring 2002, Spring 2004.

stabilized or decreased after 19929. Do the EDP and SGP work more effectively in small EMU states than in the large ones? To answer this question, Table 1 reports the changes in the debt-GDP ratios for large states (whose GDP in 1997 was at least 7 percent of the EU GDP – Germany, Spain, France, Italy, and the UK), intermediate states (whose GDP was between two and seven percent – Belgium, the Netherlands, Austria, and Sweden), and small states (whose GDP was less than two percent of EU GDP – Denmark, Greece, Ireland, Luxembourg, Portugal, and Finland). The combined GDP of the large states is 80 percent of the EU GDP, and that of the intermediate states 13 percent. The small states have a combined GDP of 7.7 percent of the EU GDP. The table shows that the average debt ratio of the small states increased by only 3.3 percent between 1992 and 1997, at a rate lower than that of the large states (almost 19 percent). Between 1998 and 2003, the small states reduced their debt ratios by seven percent on average, compared with a reduction of 4.9 percent achieved by the large states. Intermediate states reduced their debt ratios by 10.5 percent on average. This suggests that the ﬁscal framework is indeed more effective in the small than in the large states, which implies that it is most effective where it matters the least, since a ﬁscal crisis in a small EMU member state would hardly threaten the stability of the common currency. In contrast, a ﬁscal crisis in a large state might do that, and there the ﬁscal rules seem much less effective. The second qualiﬁcation is that the observation of ﬁscal consolidations in some EU states during the 1990s cannot simply be attributed to the Maastricht process. Since most European countries had sizeable ﬁscal expansions during the 1970s and 1980s, a period of consolidation was expected in the 1990s anyhow. A study of European ﬁscal policy in the 1990s (Hughes Hallett et al., 2001) considers this argument in detail. It shows that the consolidations observed in the 1990s could well be expected by extrapolating patterns of ﬁscal behavior of EU states in the 1970s and 1980s. The evidence of a ‘‘Maastricht effect’’ speeding up or enforcing consolidations is weak at best.

9

Austria’s and Finland’s debt ratios increased after 1992, but these countries were not bound by the EDP at the time.

From the Stability and Growth Pact to a Sustainability Council for EMU

175

Figure 1. Fiscal stance 1998–2003 4 1998

1999

2000

2001

2002

D

EL

E

2003

3 2

Percent of GDP

1 0 -1 -2

B

DK

F

IRL

I

L

NL

A

P

FIN

S

UK EU-12

-3 -4 -5 -6 Year/Country

3.2. Fiscal policy stance 1998– 2003 As pointed out above, all EU countries enjoyed declining debt ratios between 1997 and 2001. But, the period from 1997 to 2000 was also one of relatively strong growth in Europe. Since ﬁscal performance is measured in terms of debt and surplus ratios relative to the GDP, it is not clear to what extent the observed reductions in government debt and deﬁcit ratios can be attributed to government policy as opposed to windfall gains from strong economic growth. In this section, the recent performance is assessed and an attempt made to separate policy from the effects of growth. A simple method of growth accounting is used. For each year, the change in the government surplus ratio due to economic growth and a ‘‘neutral’’ policy, deﬁned as one that keeps the average tax ratio and the ratio of government spending to potential GDP constant, is estimated. Subtracting the two from the observed change in the surplus ratio gives us an estimate of the active policy stance, see the Appendix for details.10 Figure 1 has our calculations for the years from 1998 to 2003. Since the decision on EMU membership was taken in 1998 on the basis of ﬁscal data for 1997, 1998 was the ﬁrst year after 1992 in which the governments of the EMU member states were no longer at risk of not making it into the monetary union

10

Alternatively, one might base similar calculations on the OECD’s cyclically adjusted budget balance and its estimates of changes in structural balances. These estimates, however, are based on data and policies in the 19870s and 1980s. If the 1990s brought about a change in the ﬁscal policy regime in Europe, they could be quite misleading.

Ju¨rgen von Hagen

176

Table 2. Preelection years in EMU Pre-election year

1998

1999

2000

2001

Country

Austria, Belgium, Finland, Luxembourg, l Portuga

Spain, Finland, Greece

Italy, Portugal

Germany, France, Portugal, Netherlands, Ireland

owing to excessively lax ﬁscal policies. In the ﬁgure, a negative number indicates a ﬁscal expansion, a positive number a ﬁscal contraction.11 The ﬁgure bears interesting observations. The ﬁrst is that on average, the EMU average policy stance was expansionary in 1998, very much so in 2000, neutral in 1999, 2001, and 2002 (despite the weaker economic performance that year), and restrictive in 2003 (again a year of weak growth). Thus, the ﬁscal framework of EMU seems to promote a pro-cyclical behavior of ﬁscal policies. The second observation is that ‘‘consolidation fatigue’’ – the loss of political interest in pursuing further consolidations – emerged in many countries in the ﬁrst year after the threat of their not being able to make it to the EMU had disappeared. The (non-weighted) average ﬁscal impulse among the EMU member countries in 1998 was –1.12 percent of the GDP, with a standard deviation of the mean of 0.24. This compares with an average ﬁscal impulse in all other country-years of –0.54 percent of GDP, with a standard deviation of 0.17. The t-test for equal means yields a statistic of t ¼ 3.2, which indicates that the 1998 ﬁscal impulses were signiﬁcantly more expansionary among the EMU member states. Thus, these countries used the ﬁrst opportunity to relax ﬁscal policies, although 1998 was a year of relatively strong economic growth. Interestingly, the countries that did not join the EMU in 1999 – Denmark, Greece, Sweden, and the UK – maintained tight or contractionary ﬁscal policies in 1998. The third observation emerges from considering the election dates in European countries in recent years. If governments use ﬁscal policies to improve the ruling parties’ chances of reelection, one should expect ﬁscal expansions in the year preceding an election. Table 2 indicates pre-election years in EU countries. Here, we count both parliamentary and presidential elections, where applicable. Collecting data from these country-cases, we ﬁnd that the (unweighted) average ﬁscal impulse in pre-election years is –0.96 percent of the GDP, with a standard deviation of the mean of 0.28. The average ﬁscal impulse in all other country-year cases is –0.65 percent of the GDP, with a standard deviation of the mean of 0.15. The t-test for the difference between the two averages is t ¼ 1.7, which is signiﬁcant at conventional levels. Thus, the data indicate that the ﬁscal

11

See Hallett et al. (2000) and Hallerberg, et al. (2001, 2002) for similar calculations and results.

From the Stability and Growth Pact to a Sustainability Council for EMU

177

strictures of the EDP and the SGP do not prevent governments from using ﬁscal policies to pursue electoral interests. Our estimates conﬁrm similar results in Hallerberg et al. (2001), who use a somewhat different methodology.

3.3. Patterns of fiscal adjustment in the EMU At the European Council in Lisbon in 2000, the EU called upon its membersto improve the ‘‘quality’’ of public ﬁnances. Without deﬁning precisely what ‘‘quality’’ of public ﬁnances meant, the Council accepted that the structure of public spending and taxation had important consequences for economic growth, and decided that the EU member states should aim at a more growth-friendly structure of public ﬁnances. Endogenous growth theory broadly suggests that a shift from taxing factor-incomes to taxing consumption, and a shift from public consumption and transfer spending to public investment have positive growth effects (Aghion and Howitt, 1998). Empirical results in this area are mixed, but they suggest that ﬁscal policies affect growth.12 Subsequently, the ﬁscal policies of EMU member states are characterized to assess the strength of this conjecture. This is done by a series of cross-section regressions focusing on the period since 1997. While cross-sections have obvious data limitations, the following bits of evidence add up to a picture that underscores the importance of the structure of ﬁscal adjustments and taxes and spending, more generally.13 The ﬁscal rules of the EDP and the SGP focus on a reference value for public debt relative to the GDP are noted. There are two ways of reducing this ratio – slow down the growth of nominal debt or speed up the growth of GDP. Since inﬂation is no longer under the control of domestic monetary policy, the latter is equivalent to speeding up real GDP growth. A ﬁrst question considers the choice of the EMU government between these two options. Let d ¼ B/Y be the ratio of public debt, B, to the GDP, Y. The relative contribution of growth in public debt and growth in real GDP to the change in this ratio in country i can be written as 1 þ bi C t ¼ 100 1 , ð1Þ 1 þ gi where b is the growth rate of nominal debt and g the growth rate of real GDP. If Ci>0, the growth of public debt contributed more to the change in the debt ratio than the growth of real GDP, otherwise, real GDP growth dominated.

12

See, e.g., Tanzi and Zee (1997), Fo¨lster and Henrekson (1999), and Kneller et al. (1999, 2000). To facilitate reading the following ﬁgures, note that an R2 of 0.20 in the following regressions corresponds to the 10 percent critical value, and an R2 of 0.26 to the 5 percent critical value of the F-distribution of a test for statistical signiﬁcance. 13

Ju¨rgen von Hagen

178

Figure 2 shows a plot of Ci against the real growth rates of the EU countries for two time periods, 1992–1997 and 1997–2003. Positive values on the x-axis indicate that the change in the debt ratio during the period considered was owing to growth rates of public debt in excess of the growth rate of real GDP. This was true in almost all EU countries in the ﬁrst period. In contrast, public debt grew less than real GDP in all countries since 1997. Signiﬁcantly, the ﬁgure also shows a strong correlation between the average real GDP growth rate over the post-1997 period, and the relative contribution of GDP growth to the change in the debt ratio. Such a relationship did not exist in the ﬁrst half of the 1990s. Figure 3 gives a plot of the relative contributions of debt and real GDP growth against the change in the debt ratio during the period under consideration. In the earlier period, debt ratios increased as debt was growing much faster than real GDP. In the later years, however, the pattern is reversed. Countries that achieved a large decline in the debt ratio are those that achieved high real-GDP growth rates relative to the growth rate of debt over this period. Countries that achieved little real growth relative to debt growth did not manage to reduce their debt ratios signiﬁcantly. The ﬁgure thus suggests that a successful strategy to reduce the debt ratio focuses on growing out of the debt burden rather than on slowing down the growth rate of debt and neglecting economic growth. Taking Figures 2 and 3 together, a clear message emerges: without reviving economic growth, a signiﬁcant reduction in the debt burden is unlikely. Taking the two periods together, another message is that rising debt burdens come from a lack of control over public-sector debt. But to reduce an excessive debt burden, controlling debt is a necessary condition. This suggests that the ﬁscal framework of the EMU is ill conceived. The focus on deﬁcit and debt Figure 2.

Fiscal adjustment

8 1997- 2003

1992-97

7 y = -0.0773x + 2.2667 R2 = 0.4219

6

Real GDP Growth

5 4 3 2 y = -0.0408x + 2.345 R2 = 0.208

1 0 -60

-40

-20

20 -1 Rel. Contribution of Debt and GDP

40

60

80

From the Stability and Growth Pact to a Sustainability Council for EMU

179

ratios alone would be justiﬁed if the EMU had started in a period in which public debt burdens were regarded as compatible with long-run equilibrium. Given that a reduction in the debt burden is necessary, particularly in the large countries, the policy framework pays too little attention to the role of economic growth in achieving sustainable public ﬁnances. Next, we turn to public-sector revenues and spending. In Figure 4, we look at the structure of public sector revenues and growth. The sum of taxes on income and wealth and social security contributions has been termed ‘‘direct’’ taxes. The Figure 3. Change in debt ratio 30 1997-2003

y = 0.4562x - 0.5017 R2 = 0.6998

1992-97

20

Change in B/Y

10 0 -60

-40

-20

0

20

40

60

80

-10 -20 y = 0.616x - 1.993 R2 = 0.6467

-30 -40 relative Contribution

Figure 4.

Revenues and growth

8

direct taxes in total revenue

7

real gdp growth

6 5 4 3 y = -0.2262x + 17.336 R2 = 0.2228

2 1 0 50

52

54

56

58

60 percent

62

64

66

68

70

Ju¨rgen von Hagen

180

ﬁgure shows a negative association between the share of direct taxes in total government revenues and real GDP growth over the period 1997–2003. Countries that relied heavily on direct taxes as a source of income had weaker growth than those that relied strongly on indirect taxation. In Figure 5, we look at the share of public investment and social transfers in total expenditures. The ﬁgure shows that countries with higher shares of public investment experienced higher growth rates, while those with higher shares of social transfers in total spending suffered from weak growth. Finally, in Figure 6, we look at the correlation between ﬁscal consolidation and real GDP growth. We do this by plotting the growth rate of public debt together with the growth rate of real GDP for the two time periods –1992–1997 and 1997–2003. The ﬁgure and the two regressions indicate that there is no signiﬁcant correlation between them. High growth rates of public debt in the early period apparently did nothing to stimulate economic growth, and lower growth rates in the latter period did not reduce growth. Nor does the ﬁgure give much credence to the concept of ‘‘non-Keynesian’’ effects of ﬁscal consolidations, i.e., the notion that a reduction in public debt would have positive growth effects by stimulating private investment and consumption (Giavazzi and Pagano, 1990). Such effects would lead us to expect higher growth rates for countries where public debt shrank in the period under consideration. Obviously, the present bivariate framework is not sufﬁcient to achieve a strong conclusion on the matter. Nevertheless, it is in line with the results from a larger econometric model presented by H. Hallett et al. (2001), which do not indicate ‘‘non-Keynesian’’ effects of the ﬁscal consolidations in Europe in the past decade. In passing, we note that our evidence here points to a methodological problem in earlier studies of such effects. Speciﬁcally, most studies Figure 5.

Spending structure

8

investment social transfers

7

real growth rate

6 y = -0.1539x + 12.938 R2 = 0.1828

5 4

y = 0.3224x + 1.1424 R2 = 0.4105

3 2 1 0 0

10

20

40 50 30 share of total spending

60

70

80

From the Stability and Growth Pact to a Sustainability Council for EMU

Figure 6.

181

Fiscal adjustment and growth

8 1992-97

1997-2003

7

real GDP Growth

6 5

y = -0.0317x + 3.717 R2 = 0.0028

4 3 y = -0.1449x + 2.6958 R2 = 0.1032

2 1 0

0

2

4

6

8

10

12

14

16

Growth of public Debt

identify ﬁscal consolidations as periods of signiﬁcant reductions in public debt or deﬁcit ratios and ‘‘non-Keynesian’’ effects as episodes where consolidations go along with vigorous economic growth. The European experience suggests that such episodes may have more to do with policies that succeed in stimulating growth by restructuring public spending and taxation, and reducing tax burdens, than with a reduction in public debt or deﬁcits. To summarize, the experience of the early years indicates that the main condition for achieving a sustainable reduction of the debt ratio is to achieve a sufﬁciently large rate of trend real growth. Fiscal policy can support this by lowering the tax burden on factor incomes and by shifting expenditures from transfers to public investment. Do these results matter for the EMU? After all, one might argue that the stability of the common currency depends only on the stability of public-sector debt ratios. How this stability is achieved might be left to the choice of the individual member states. The subsidiarity principle of the TEU would then suggest that the Union should not interfere with these choices. There are, however, at least two counterarguments. The ﬁrst is that, if Europeans truly believe that public-debt ratios must be low and sustainable, success in achieving this goal matters, and is a valid concern for the Union. From this perspective, the current ﬁscal framework is incomplete, as it does not give EMU member states enough guidance for the choice of a successful ﬁscal strategy. Countries should be encouraged to adopt more growth-friendly policies by restructuring their tax and expenditure systems. Second, it is necessary to recognize that the EMU did not start under conditions of a long-run equilibrium as far as public ﬁnances are concerned. The low growth rates in Germany, France, and Italy, in particular, are the result of overregulated economies plagued by high tax burdens and welfare systems that

182

Ju¨rgen von Hagen

discourage employment. The narrow focus of the EDP and the SGP on annual deﬁcits, however, may keep governments from adopting reform policies that might result in larger deﬁcits initially before the desired growth and employment effects kick in. If so, there is a risk that the current design of ﬁscal strictures will keep these countries in a state of low growth, in which they will be unable to make sufﬁcient progress in reducing their debts and deﬁcits, too. One may reasonably doubt that these large EMU states will continue to tolerate such a scenario, which is perceived as keeping them from adopting better economic policies for the sake of achieving some ﬁscal targets imposed by the EU. The recent episodes involving France and Germany clearly indicate that they will not. But if the outcome were that these countries continue to ignore the goals of the EDP and the SGP, other states would follow and the ﬁscal framework of EMU would fall apart. 4. A Sustainability Council for the EMU The experience recounted above and the institutional considerations discussed in Section 2 indicate that the ﬁscal framework of the EMU needs improvements in two directions: less focus on short-run ﬁscal ﬂows, and stronger surveillance and monitoring procedures. At the heart of the proposal below is the idea to replace the rigid rules by a judgmental assessment of the ﬁscal situation and outlook of each euro-area member state, and to entrust the judgment of sustainability to an independent institution, the European Sustainability Council. This would solve the basic credibility problem of the current framework. The proper link between the EMU’s long-run interest in sustainability and short-run constraints and exigencies on ﬁscal policy would be preserved by the independence of the Sustainability Council short-term political pressures.14 The Sustainability Council would be legally set up by the European Parliament, which would also provide its resources. Its members should be individuals of high public regard as experts in public ﬁnance or public ﬁnance management. Membership of the Council need not be a full-time activity for all members, although the chairman and the vice-chairman should be full-time professional appointments. The Council should have a small staff and secretariat, and guaranteed access to all relevant information at the national and Community levels. It should have the right to use the services of the European Commission and of the national government, and have courts support its work. The idea of creating yet another institution at the European level may seem unattractive to some. After all, the current European structure, with its network of overlapping policy processes and its opaque institutional setup, seems to call

14

Our proposal builds on work originally presented in von Hagen and Harden (1994), further elaborated in Eichengreen et al. (1998). A related proposal drawing on these ideas was recently presented by Wyplosz (2002).

From the Stability and Growth Pact to a Sustainability Council for EMU

183

for less rather than more policymaking bodies. Yet, the Sustainability Council would improve the transparency and visibility of the current institutional framework for public ﬁnances in the EMU. Some may ﬁnd the idea of delegating some authority over public ﬁnances – historically the core of parliamentary rights – to an independent body incompatible with modern democracy. Yet, as is argued below, a properly designed Sustainability Council will improve the functioning of democratic government rather than limit it (von Hagen and Harden, 1994). The delimitation of the Council’s authorities and competences is the key issue. In fact, several countries in the EU have already moved in this direction in recent years. 4.1. Mandate The Sustainability Council would have the sole statutory task of safeguarding the sustainability of public ﬁnances in the euro area. While this is the counterpart of the ECB’s principal task of maintaining price stability, the Sustainability Council has no need of a secondary objective – supporting the general economic policies in the euro area – since it would have no operative role in ﬁscal policy. The use of the instruments of ﬁscal policy would be left entirely to the national governments. Within the EMU framework, the function of the Council is to make the implications of the governments’ intertemporal budget constraint explicit. To fulﬁll its task, it must assess the ﬁnancial position of a government in all relevant aspects, produce forecasts of future ﬁnancial developments and, on this basis, evaluate the risk of future ﬁscal crises. Like ‘‘price stability’’, the empirical content of the concept of sustainability of public ﬁnances is rather vague, and it varies with economic circumstances over time and across countries. An important part of the Council’s task is, therefore, to develop a framework for the assessment of public ﬁnances and to make forecasts and judgments. This, again, is similar to the task of the ECB. Yet, it is unlikely that the Council will settle on a unique number in this deﬁnition as the ECB did for price stability. In fact, there is no need to do that, since, in contrast to the ECB’s case, the Council is not charged with the implementation of sustainability and, therefore, its deﬁnition of sustainability is not needed to hold it accountable for its actions in the short run. Thus, the Council is free to develop an empirical concept of sustainability that overcomes the basic problem, i.e., that general numerical limits are not meaningful in this context. That the Council would not have the authority to set taxes or expenditures for a country is the key to its democratic legitimacy. The Council’s mandate would be to make explicit the limits that monetary union imposes on national ﬁscal policy choices. This is a legitimate interest of the Union, and it can be assumed that national parliaments and governments, by agreeing to enter a monetary union dedicated to price stability, agreed to accept those limits. Thus, the creation of a Sustainability Council does not take away further sovereignty

184

Ju¨rgen von Hagen

from national governments compared with what is already implied by entering the EMU. Nevertheless, the Sustainability Council would hardly be able to judge the sustainability of a country’s public ﬁnances without taking into account a view of the size and structure of the public sector in that country. While the assessment of a country’s ﬁscal situation and outlook may contain recommendations regarding the total volume and distribution of public spending and revenues, ﬁxing these volumes would be beyond the mandate of the Council. Thus, it may ﬁnd itself in disagreement with national governments, precisely because the latter desire to expand the public sector. Governments could improve the process and ensure that the Council’s decisions were linked to democratic political choice by announcing, on their part, multi-annual targets for the size and structure of the public sector. Such targets, if announced by a new government, would most likely become part of political parties’ electoral platform and improve the democratic accountability of the government in public ﬁnances. 4.2. Method of operation The national governments would submit their annual and medium-term ﬁscal plans to the Sustainability Council, which would judge whether the implied change in general government debt is compatible with sustainability. The Council would make its judgment and the underlying reasoning public in a report to the European Parliament, and propose adjustments from the national governments. Note the substantial difference between the change in general government debt, the main focus of the Council, and the reference values of the Maastricht Treaty. The Treaty deﬁnes a deﬁcit on an accruals basis, implying that some items that the Council would cover are excluded, such as privatization receipts, capital (‘‘below-the-line’’) transactions, or changes in the value of foreign-currency denominated debt. The Council would, therefore, take a more encompassing view and judgment. Furthermore, the change in general government debt considered by the Council would set an unambiguous limit on government borrowing. In contrast to the EDP, there would be neither room nor a need for judgment once this has been established. The Sustainability Council could have procedures that allow for involvement with different countries in varying degrees of intensity. For example, it could apply a simple ﬁrst test using some rather broad-brushed analysis and turn to a more intensive investigation only in cases where the ﬁrst test has failed. The 3 and 60 percent criteria of the EDP are examples of a ﬁrst test. This would likely reduce the number of countries investigated intensively each year and, therefore, allow for a smaller administration. To fulﬁll its task, the Sustainability Council would produce an annual report on the sustainability of public ﬁnances in each member state and submit it to the European Parliament. In the preparation of the report, the Council would,

From the Stability and Growth Pact to a Sustainability Council for EMU

185

without prejudice to its independence, allow for participation of the national governments and other institutions, such as the ECB, in the process, e.g, by holding hearings with experts and representatives of relevant bodies. The Council should be able to obtain all relevant information for its tasks and be able to use it.As a general rule, the Council would present its report in the capital of the country concerned. This is important as a ‘‘general European public’’ still does not exist in the EU. Since the Council’s effectiveness relies on its ability to mobilize public opinion and debate, it is critical that it talk directly to the media and the public, especially when it has reason to criticize government policies. A voice from Brussels would hardly be heard in individual countries. Furthermore, the Council should seek opportunities to talk directly to the parliaments of countries where the sustainability of public ﬁnances is questionable. Where similar institutions exist at the national level, such as the High Council of Public Finances in Belgium, it would be natural and useful for the Council to talk to these institutions in the preparation of its report and presentation of the report to the public. 4.3. Enforcement Under the European Treaty, member states of the EMU have the unconditional obligation to safeguard the sustainability of public ﬁnances. If the Sustainability Council has the mandate of deﬁning and operationalizing what sustainability means and implies for national ﬁscal policies in the short run, this obligation implies that national governments are committed to implementing the Council’s judgments and prescriptions. The question how this commitment can be enforced remains. The current framework of public ﬁnances in the EMU relies on two enforcement mechanisms – peer pressure, and the possibility to impose ﬁnancial ﬁnes on countries with persistent excessive deﬁcits. But the effectiveness of these enforcement mechanisms remains very much in doubt. While peer pressure has not worked regarding the large states – Germany, France, and Italy, in particular – the effectiveness of the threat of ﬁnes remains to be tested in the EMU framework. But the lenience with which Germany’s ﬁscal developments were treated in 2002 and 2003 suggests that the European Commission and the Council wish to avoid the test. It is clear that the Council can rely neither on peer pressure nor on ﬁnes. As a Community institution, it does not talk to the national governments in the same way as Ecoﬁn does. At the same time, the Council cannot impose penalties on national governments, as its role is different from that of the European Court of Justice. Therefore, the Council would have to rely primarily on political pressures generated through public opinion to enforce its judgments. To do so, it must have the right to make its judgments and recommendations fully public in time. This includes the right to make press declarations on individual cases, and to educate public opinion through statements on the importance and proper interpretation and implementation of sustainability. It also includes the right to

186

Ju¨rgen von Hagen

talk to the European Parliament and national parliaments. The Council must have the right to make differentiated judgments on the ﬁscal situation of each member state in public, pointing to risks and problems, as it sees ﬁt. Finally, in order to create clear competences and avoid political haggling, the Council should have the sole right to recommend to the Ecoﬁn the imposition of ﬁnes under the EDP of the European Treaty, and the ECOFIN Council should be required to take a vote on that proposal. Enforcement in this way can only work if the public regards the Sustainability Council as an authority on this matter. A Council making unreasonable judgments or posing unreasonable demands to national governments frequently would soon lose attention in the public debate, as would one that makes its judgments on shaky analysis and questionable assumptions. The need to rely on public opinion, therefore, creates a strong incentive for the Council to exert good judgment, to use its public role carefully and to refrain from making public announcements and hinting at impending ﬁscal crises unless the situation is severe. Finally, governments in the EMU should have an opportunity to prepare a response in reasonable time and not be taken completely by surprise by announcements of the Sustainability Council. This can be achieved by demanding that the Council forwards its assessment of a country to the relevant government a few days before it is made public. 4.4. Independence, accountability, and transparency To fulﬁll its role properly and make unbiased judgments, the Sustainability Council must enjoy full political independence from the governments of the member states and of other Community institutions such as the European Committee. Like the independence of the ECB, that of the Council is determined in four statutory rules. First, a rule stating that the Council does not take directives from governments of EU member states, from other national institutions of these states, or from any Community institution. Second, a rule giving the Council the right to develop its own framework of analysis and its own operational concept of sustainability. Third, a rule determining the resources available to the Council, which should ﬁx its budget for a medium-term horizon, say, ﬁve years, and should be amendable only by a qualiﬁed majority of the European Parliament. Such a rule would shield the Council from shortsighted attempts of politicians to make it ineffective by draining it of resources. Fourth, the members of the Council should be personally independent from political pressures. Following the example of the European Central Bank, personal independence can be assured by giving the Council members a ﬁxed term and non-renewable appointments of sufﬁcient length, say, eight years, to acquire expertise and standing in the public debate. Appointments should be staggered to assure that thecomposition of the Council does not change entirely at the end of a year, thus assuring continuity in its views and judgments. It should be

From the Stability and Growth Pact to a Sustainability Council for EMU

187

impossible to dismiss members of the Council except for grave instances of unethical or unprofessional behavior, to ensure that they are not threatened with removal for decisions that are unpopular with governments. Salaries of Council members should be determined by a formula laid down by the Council’s statutes and linked to the salaries of comparable EU ofﬁces. Freedom from political pressure should be balanced by appropriate mechanisms of accountability. As indicated above, the Sustainability Council should report to the European Parliament. The Parliament should have the right to call the chairman of the Council for hearings and the right to dismiss the Council in toto by a qualiﬁed majority. Given the large publicity that such an action would have, the European Parliament would resort to such a step only in case of very poor performance by the Council. The Council’s independence should be balanced by a high degree of transparency of its operation. Limited transparency would only reduce the effectiveness of the Council’s public announcements, as they might raise doubts about the competence and fairness of its deliberations and judgments. This calls for the publication of all materials relevant to the Council’s decisions Council as well as the minutes of its meetings. However, there is no need for this immediately after ameeting or decision, a requirement that might affect the Council’s ability to obtain and process relevant information. The Council could choose a gap of, say, 6 to 12 months, within which all relevant information is published. Apart from that, Council members should be free to express their views on the sustainability of the public ﬁnances of individual countries. This would promote a public debate on relevant issues in member states where sustainability is indeed at risk, and, therefore, raise public pressures on the governments to correct the situation. 4.5. Appointments, composition, and resources In the context of ﬁscal policy, the role of the Sustainability Council would be fairly technical and important. Its members should have sufﬁcient experience. In some member states of the EU, academics with the necessary expertise would probably be regarded as appropriate candidates, while in other member states, individuals with careers in international institutions such as the IMF would be preferred. Since these appointments would be made by the European Parliament, the members of parliament could resolve national differences in preferences through the nomination procedure. For a European institution, the question of country representation naturally arises. Since the role of the Sustainability Council is not to make policy choices, representation of all countries at all times would not be important for its legitimacy. By decision of the European Parliament, the Council should be vested with the resources necessary to fulﬁll its task. This includes staff for analysis and a small secretariat. EMU member states must be required to give the Council full and timely access to all information requested.

188

Ju¨rgen von Hagen

5. Conclusions Creation of the Sustainability Council would improve the political process of public ﬁnances in the EMU. It would allow for greater ﬂexibility in the use of ﬁscal policy instruments, as it would be able to apply good economic judgment and would not focus narrowly on numerical criteria. This increase in ﬂexibility would not come at the cost of credibility, since the Council would have no need to take short-term political concerns into consideration for judging public ﬁnances. The review process of the Council would change public awareness of the problem by being more transparent and more focused on critical issues connected to sustainability. Like price stability, sustainability is a difﬁcult concept that needs continuous education of the public to build democratic support. The Council would be in a position to deliver the goods as sustainability would be its sole mandate. How does our proposal relate to the current framework for public ﬁnances in the EMU? One obvious question relates to the status of the Council within the EDP and the SGP. The existence of the Council along with the current framework could pose difﬁculties. The fact that the Maastricht criteria and the assessment of the Council do not coincide implies the possibility of a conﬂict between strict compliance with the rulings of the Council and the EDP. Situations could arise in which the Council declares that the sustainability of a member country’s public ﬁnances are at risk, while the actors under the EDP and the SGP come to the opposite conclusion. Political haggling involved in such situations would undermine the authority of all institutions and procedures, leaving the EMU with less protection against ﬁscal proﬂigacy. To avoid that risk, the provisions of the EDP and the SGP should be amended to clarify the authority of the Council in these matters. It is noteworthy that a number of EU countries already have institutions at the national level which resemble the Sustainability Council in some ways. In Belgium, the High Council of Finances coordinates ﬁscal policies between the national government, the regional governments, and the communities. In doing so, it assesses the sustainability of the public ﬁnances of the country and the subentities, and regulates their policies annually. In Austria, the reformed ‘‘Staatsschuldenrat’’ reports to the national parliament on issues related to the evolution of government deﬁcits and debt. The Swedish parliament recently decided to set up a similar council at the national level when Sweden enters the EMU. Thus, the principle has already gained some recognition in Europe. References Aghion, P. and P. Howitt (1998), Growth Theory, Boston: MIT Press. Delors Report (1989), Report on Economic and Monetary Union in Europe, Luxembourg: Ofﬁce for Ofﬁcial Publications of the EC, Committee for the Study of Economic and Monetary Union.

From the Stability and Growth Pact to a Sustainability Council for EMU

189

European Commission (2002), ‘‘Coordination of economic policies in the EU: a presentation of the key features of the main procedures’’, Euro Paper, DG ECFIN, p. 45. Fo¨lster, S. and M. Henrekson (1999), ‘‘Growth and the public sector: a critique of the critics’’, European Journal of Political Economy, Vol. 15, pp. 337–358. Giavazzi, F. and M. Pagano (1990), ‘‘Can severe ﬁscal contractions be expansionary?’’, NBER Macroeconomics Annual, pp. 75–116, Hallerberg, M., R. Strauch, and J. von Hagen (2001), ‘‘Forms of governance and the design of ﬁscal rules in the EU countries 1998-2000’’, Mimeo, Bonn: ZEI University. Kneller, R., M.F. Bleaney and N. Gemmel (1999), ‘‘Growth, public policy, and the government budget constraint: evidence from OECD countries’’, Journal of Public Economics, Vol. 74, pp. 171–190. Perotti, R., J. von Hagen and R. Strauch (1998), Sustainability of Public Finances, London: CEPR. Sargent, T. and N. Wallace (1981), ‘‘Some unpleasant monetarist arithmetic. Quarterly Review’’, Federal Reserve Bank of Minneapolis, Vol. 5(3). Stark, J. (2001), ‘‘Genesis of a pact’’, in: A. Brunila, M. Buti and D. Franco, editors, The Stability and Growth Pact, Houndmills: Palgrave. Tanzi, V. and H.H. Zee (1997), ‘‘Fiscal policy and long-run growth’’, IMF Staff Papers, Vol. 44, pp. 179–209. Appendix: calculation of fiscal stance Let the primary surplus ratio, st, be st ¼

Rt G t ¼ ðrt gt Þ Yt

ðA1Þ

where R denotes government revenues, G the non-interest government spending, and Y the GDP. The change in this ratio over time, then, is Dst ¼

DRt DG t DY t ðrt gt Þ Y t1 Y t1

ðA2Þ

where r ¼ R/Y, and g ¼ G/Y. We deﬁne a ‘‘neutral’’ ﬁscal policy as one that keeps the average tax rate, r, and the ratio of government spending to trend GDP constant. With this deﬁnition, the contribution of the neutral policy to the change in the surplus ratio is DsN t

DY t DY trend ¼ rt1 gtt Y Y t1

ðA3Þ

Ju¨rgen von Hagen

190

The contribution of economic growth to the change in the surplus ratio is deﬁned as "

trend # DY DY t DsG t ¼ gt Y Y t1

ðA4Þ

This is the change that would occur in addition to the neutral change if the government allowed economic growth above or below trend to change the expenditure ratio. The trend growth rate is estimated as the average real growth rate during the 1990s. Using these deﬁnitions, we obtain the policy-induced change in the surplus ratio as G DsPt ¼ Dst DsN t Dst

ðA5Þ

This part has been used as an indicator of ﬁscal policy stance, since it measures the active contribution of any policy actions against changes observed in the surplus ratio.

Comment Wolfgang Eggert

Ju¨rgen von Hagen points out that the Stability and Growth Pact (SGP) contains conventions that make limited economic sense. He proposes a Sustainability Council in the European Monetary Union (EMU) to correct the ﬂaws. The Council would replace rigid rules with more transparent policy recommendations tailor-made for the needs of the heterogeneous EMU member countries. However, it might also increase frustration about the working of EU institutions if the solution von Hagen proposes turns out to be impracticable, and if it can be undermined as much as the SGP by national governments. These are the costs that have to be weighed against the potential beneﬁts of the proposal. New transparency might help implement a more growth-oriented ﬁscal policy. I hope that the potential gains can be realized once the Council comes into being. In what follows, I will attempt to describe the problem that I think is addressed, and restate why the SGP seems to fail. I suggest that the proposed Council may have some similar conceptual problems as the SGP, which is unavoidable, at least to a certain degree, as long as EMU member countries are sovereign. The topic von Hagen addresses is particularly timely since the EMU countries, especially the larger ones, are increasingly acting against the substance of the SGP. Naturally, there may be many reasons that explain the failure of an agreement. An examination of the reasons for the failure of the SGP would, in my view, suggest that two might be of primary importance. Both arguments rely on the view that the SGP restricts the ability of countries to run risky policies to prevent inﬂationary debt bailouts, the costs of which are borne by the residents of all EMU countries, rather than only by those of the country responsible. According to the ﬁrst argument, restricting current public expenditures may cause an tightening of a country’s consumption possibility frontier. Countries may ﬁnd it in their national interest to act against the substance of the SGP. This view would lead to the conclusion that substantial participation in the SGP is

CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79014-9

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

192

Wolfgang Eggert

not advisable unless a supranational authority is established to enforce the Pact, or some large EU countries use their political power to enforce an international contract. In the latter case, it is, however, less clear whether the international contract should be welcomed from the aggregate perspective that considers the welfare of all EU member states. The second factor that may break the SGP is a coordination failure between countries. This scenario belongs to a class of economic environments where multiple equilibria in policy decisions exist and countries face an equilibrium selection problem. In this case, governments may shift the national consumption possibility frontier in two ways: by increasing the level of publicly provided consumption goods, or by increasing public investment, to increase the level of publicly provided consumption goods. Countries may still gain from imposing negative externalities on others. In contrast to the ﬁrst argument, however, there is a possibility that they will coordinate their expectations on a different equilibrium. The argument favouring an international institution or treaty among sovereign countries is convincing if the underlying problem is equilibrium selection. The multiplicity of equilibria can rationalize an international agreement that potentially serves as a coordination device. In this light, the SGP might be ill designed because failure suggests that it did not help coordinate expectations. It is a policy-relevant intellectual challenge to identify a new institution that is better suited for the task. Von Hagen gives empirical evidence in a very good attempt to illustrate the incentives to EU countries when they chose the structure and level of public expenditures in the past. In addition, it gives an idea about the potential gains from restructuring ﬁscal policy decisions. The following is an attempt to review some of the results that I think are essential as they provide insights about relevant economic mechanisms. Section 3.1 describes the experiences in the 1990s. To me, an important insight here is that the ﬁscal framework imposed by the excessive deﬁcit procedure (EDP) and the SGP works more effectively in the smaller EU countries than in the larger ones. The evidence provided suggests, moreover, that the ﬁscal framework is quite effective in countries that have limited ability to create negative externalities on others through strategic use of debt policy. Section 3 starts with the arguments that neither the EDP nor the SGP prevented governments from using policies to satisfy political interests in the past. The section continues by providing information about the patterns of ﬁscal adjustment in the EMU. The two main messages are as follows: ﬁrst, countries are able to grow out of debt, which gives rise to the perception that the focus on deﬁcits and public debt within the SGP is inappropriate to reduce the ratio of public debt to GDP over time. The discussion provides some evidence that the growth rate of public debt cannot explain the rate of economic growth. Second, repeated increases in tax rates, especially those on factor income, slow down economic growth. The negative effect of tax-increases seems to be more pronounced the more tax revenue is used to ﬁnance public consumption and the less tax-revenue goes to public investment.

Comment

193

Is there any way to correct the incentives of low-growth countries to use tax revenues and public debt to ﬁnance a high level of consumptive public expenditure instead of public investment? The answer of von Hagen, I think, is correct. The only way to escape the problem entirely is to change the possibly faulty perception in low growth countries that there is no alternative. In other words, one needs to create an institutionthat is able to coordinate expectations on an equilibrium which yields higher growth rates. Unfortunately, we do not have a fully articulated, completely reliable and universally accepted model, at least to my knowledge, that could show us how to establish such an institution. So, the problem here is to ﬁnd a practical solution. The Sustainability Council is a potential solution. In Section 4, von Hagen provides a very good start to characterize the conditions that the Council must satisfy in order to convince EU member states. First, it must solve the credibility problem in the SGP. Second, it should be able to enforce the commitment of national governments to implement the Council’s judgements. Of course, there is a good chance that national governments will be willing to accept the recommendations made by the Council if the underlying problem to be solved is one of coordination on an better equilibrium. Third, the Council should not directly interfere with the democratic institutions in nationstates. Fourth, the new institution should be independent and accountable, and its operation transparent. I think von Hagen nicely discusses these prerequisites for successful implementation of the Council. Considering that a potential lack of rationality for a purely numerical rule may well have been a major reason for the failure of the SGP, I think the Council may remove many defects in the Pact. Most importantly, the Council would replace rigid rules by more transparent policy recommendations that take into account the speciﬁc situation in each EMU member country. But important caveats remain. The ﬁrst condition mentioned above is credibility. This seems to be the most crucial to me, and I am slightly worried that the Council might fail in the end because of a lack of credibility resulting in an inability to change the expectations of voters concerning a country’s economic development. My own view is that credibility of the Council can potentially be destroyed by national governments. Conﬂicts of interests between national governments and the Council may arise if governments are shortsighted. Some degree of discretion on the part of governments has to be introduced in a model to explain that the electorate in a country gives more importance to recommendations and judgments made by the Council than on those made by the elected national government. Then, the Council may generate political pressure by creating public opinion, which is necessary to change ﬁscal policy decisions. The broad discussion that Germany experienced since 2002, when it became known that its deﬁcit ratio would again exceed the critical 3 percent ratio, was in the end dominated by declarations of the government. Counter arguments were not decisive. This outcome makes me a bit sceptical whether the new institution can change public

194

Wolfgang Eggert

opinion. But I might be overly pessimistic. The topic is complex since international policy problems coexist with difﬁculties to establish efﬁcient national policies. The Council von Hagen proposes would be the natural option to promote research and learn more about the interacting economic effects. The knowledge is essential to shape future policy.

CHAPTER 7

Supervision of the European Banking Market Martin Schu¨ler

1. Introduction The system for supervising ﬁnancial markets in Europe is changing. At the national level, countries such as the United Kingdom, Germany, and Austria recently installed integrated supervisory agencies, unifying their previously distinct authorities supervising banking, insurance, and securities. This raises the critical question of whether or not similar reform is needed at the European scale and, if so, how it should be structured. The Economic and Financial Committee (EFC) of the European Union (EU) recently submitted a report on ﬁnancial regulation, supervision and stability (EFC, 2002). This report, which was endorsed by the Council of Economic and Finance Ministers (Ecoﬁn), lays the basis for future supervisory and regulatory arrangements in the EU. According to the report, current securities regulatory mechanisms would be extended to the other ﬁnancial sectors based on existing inter-institutional agreements. However, the question remains whether these arrangements would be sufﬁcient for safeguarding the stability of ﬁnancial markets. Furthermore, there is controversy regarding the involvement of central banks in the supervision of ﬁnancial markets.1 Distinguishing between the cross-sector and cross-border dimensions of ﬁnancial markets, the focus in this paper is primarily on cross-border issues. Moreover, because ﬁnancial stability is still primarily a question of stability of credit institutions (Meister, 2002, p. 3), it seems appropriate to pay particular attention to supervision and regulation of banks.

1

For a good discussion of the pros and cons of an integrated ﬁnancial services supervision, see, e.g., Abrams and Taylor (2002) and Briault (2002).

CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79007-1

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

196

Martin Schu¨ler

There are two main arguments for the regulation and supervision of banks2: Consumer protection and managing the systemic risk inherent in the banking market. Several factors contribute to the need to protect consumers. Due to information asymmetries, consumers are not in the position to judge the safety and soundness of ﬁnancial institutions. Thus, it is necessary to protect individual depositors from institutional failures. Furthermore, agency problems may result in adverse behavior, i.e., a ﬁrm’s unsatisfactory conduct of business with clients.3 Systemic risk is the other reason why banks are regulated and supervised. It refers to the process whereby the failure of a single institution may lead to the failure of other institutions and to the breakdown of the entire system.4 Because individual banks are thought to be more susceptible to failure than other institutions (Kaufmann, 1995, 1996; Goodhart et al., 1998; De Bandt and Hartmann, 2000), the banking sector is viewed as particularly prone to systemic failure. There are two kinds of supervision: prudential and conduct of business (Llewellyn, 1999). Prudential supervision focuses on the solvency, safety, and soundness of ﬁnancial institutions, while conduct of business supervision looks at how ﬁnancial ﬁrms conduct business with their customers. In this study we concentrate exclusively on prudential supervision, and from here on the term supervision will refer to prudential supervision. Measures to facilitate the objectives of consumer protection and stability in the ﬁnancial markets can be viewed as a three-part structure. The ﬁrst part of this structure is regulation, which lays down the rules governing the behavior of banks and other institutions. The second part is monitoring, which ensures that banks are compliant with the regulations. The third part is the safety net, which limits the effects of bank failures on third parties. The safety net includes deposit insurance and the lender of last resort (LOLR) function, and is usually provided by the central bank. Supervision is primarily related to the regulation of ﬁnancial markets, but, more generally, it also refers to a more general observation of the behavior of ﬁnancial ﬁrms.5 The aims of this study are the following: First, to describe the existing supervisory arrangements both at the national and European levels as well as to propose

2

For a detailed discussion of the rationale for the regulation and supervision of banks, see, eg., Spong (1994), Goodhart et al. (1998), and Llewellyn (1999). For a discussion on whether the ﬁnancial system should be regulated at all, see, eg., Dowd (1996), Benston and Kaufman (1996), and Dow (1996). 3 Note that there is, therefore, a case for regulation and supervision even in the absence of systemic risk, e.g., in the case of insurance companies. 4 For a more detailed deﬁnition and survey of systemic, risk, see, e.g., De Bandt and Hartmann (2000). 5 Our distinction between regulation and supervision is in line with, e.g., Goodhart et al. (1998) and Llewellyn (1999). There is a widely used practice of referring to the authorities responsible for bank supervision interchangeably as supervisors and regulators. To avoid confusion, we will refer them as supervisors.

Supervision of the European Banking Market

197

a structure for a future supervisory arrangement. Second, these supervision arrangements are evaluated with respect to an increased systemic risk potential at the European level. I argue that cooperation between national supervisors – even in the new framework – will not sufﬁciently safeguard ﬁnancial stability. As a consequence, we argue in favor of a European observatory of systemic risk. The paper is organized as follows: The next section discusses the rationale for the need to centralize banking supervision in Europe, mainly by giving an overview of the development of the systemic risk potential in European banking; Section 3 describes the supervisory framework in the EU, including the recently implemented structural change; Section 4 assesses whether this framework is appropriate for safeguarding stability in the EU banking market; Section 5 outlines what else needs to be done, particularly with respect to crisis prevention and management; Section 6 offers conclusions. 2. The rationale for integration of banking supervision in Europe6 Before discussing reformation of the current, largely decentralized, European banking supervisory framework, it is necessary to analyze the rationale for an integration of supervision. One of the main reasons why banks are regulated and supervised is systemic risk. It is often argued that the ongoing integration of ﬁnancial markets in the EU – and not just since the introduction of the euro – has increased interdependencies among ﬁnancial institutions, and hence the potential of systemic risk (e.g., Aglietta, 2000; Speyer, 2001; ECB, 2001; PadoaSchioppa, 2002). The failure of a bank may not just have negative consequences for other banks in the same country, but may also result in breakdowns of banks in other countries. A national supervisory structure may lack the capability and incentives to assess negative cross-border externalities; if so, this system will not provide sufﬁcient supervision to prevent an international crisis. The empirical assessment of the current systemic risk potential in Europe is therefore essential to shaping and evaluating the future supervisory framework. There is extensive literature on theoretical systemic risk, starting with the classic bank-run models following Diamond and Dybvig (1983), to extensions of these models from single bank fragility to models of multiple bank systems, and ﬁnally to the modern bank contagion literature.7 However, so far there is little empirical evidence demonstrating the potential of systemic risk in Europe.8 Schu¨ler (2004), and Schro¨der and Schu¨ler (2003) try

6

This section draws partly on Schu¨ler (2002), and Schro¨der and Schu¨ler (2003). For a good survey on the theoretical as well as empirical literature on systemic risk, see De Bandt and Hartmann (2000). 8 One reason for this might be the lack of appropriate data on, e.g., interbank-lending for Europe. There are some studies (Angelini et al., 1996; Michael, 1998; Sheldon and Maurer, 1998) that try to assess the potential threat stemming from interbank lending at the national level. 7

198

Martin Schu¨ler

to assess the systemic risk potential in European banking by analyzing correlations between the stock returns of European banks and bank stock index returns, respectively. This approach to assessing the systemic risk potential follows De Nicolo and Kwast (2002). For an economic shock to become systemic, a negative externality must exist; a negative shock in a single bank is, perhaps, highly likely to have contagious effects on other ﬁnancial institutions. Such an externality can exist only if ﬁnancial institutions are interdependent in some way. Such interdependencies can either be direct, for example through direct exposures, or indirect, as when they arise from correlated exposures to non-ﬁnancial sectors and ﬁnancial markets. De Nicolo and Kwast (2002) measure total interdependencies by looking at the correlations of stock returns of large banks. Since stock prices reﬂect market participants’ collective evaluation of a ﬁrm’s prospects, they also include the impact of its interdependencies with other institutions. Consequently, one can assume that an observed increase in correlations among bank stock returns signals an increase in the systemic risk potential.9 No change in correlations or a decrease would, therefore, lead to the conclusion that the potential for systemic risk has not increased or has declined. Following this approach, we calculate rolling-window correlations between pairs of stock returns of the 54 largest European banks in 13 European countries. The analysis proceeds in four steps10: First, a return-generating model is estimated to control national factors in-

ﬂuencing stock returns.11 Second, from these regressions, residuals are calculated that contain the part of the returns that cannot be explained by national factors. These residuals should contain the part of the returns that is priced owing to international inﬂuences. Third, pairwise rolling-window correlations are calculated between these residuals using a 12-month backward-looking window. Fourth, the mean is calculated over these pairwise rolling-window correlations as an indicator of the interdependencies among European banks, and hence for the systemic risk potential in Europe.

9

Note that by using the notion ‘‘systemic risk potential’’ rather than ‘‘systemic risk’’, it is indicated that stronger interdependencies among European banks, i.e., higher correlations among bank stock returns, are a necessary – but not a sufﬁcient – precondition for systemic risk. 10 Correlations are calculated over the sample period from January 1986 to July 2001. For details concerning the methodology and data, see Schu¨ler (2002). 11 The return of the aggregate national market portfolio and the unanticipated interest rate change are used as determining factors. This step is necessary in an international context because estimating correlations between pairs of bank stock returns without controlling for common factors could result in incorrect conclusions with respect to interdependencies, and hence the systemic risk potential. An increase in correlations may result merely from an increase in the movement between the underlying common factors, which has nothing to do with the development of systemic risk.

Supervision of the European Banking Market

199

Figure 1 shows this mean series for the sample where only correlations between banks of different countries are included.12 There is a substantial temporary increase in mean correlation in the time following the stock market crash of 1987. There is a less pronounced temporary increase in the mean correlation in the early 1990s that is probably because of the banking crises in some European countries: Norway 1987–93, Sweden 1990–93, Finland 1991–94 and Italy 1990–94. A substantial temporary increase Figure 1. Mean rolling-window correlation of European bank stock returns. Mean of 1,216 pairwise monthly rolling-window correlations between the 54 largest European banks. Rolling-window correlations are calculated using a 12-month backward-looking window. Included are only combinations of banks from different countries. Included is also the result of a least-square regression of the mean correlation (mcorr) on a constant and a time trend. t-Statistics are indicated between parentheses. Two asterisks indicate significance at the 0.001 level. Moreover, the averages of mean correlation before and after January 1994 are indicated as horizontal lines

12

This gives a total of 1,216 pairwise combinations. For details, see Schu¨ler (2004).

200

Martin Schu¨ler

in mean correlation can be observed around 1999. This is probably the result of the near-failure of the hedge fund Long-Term Capital Management in September 1998 that followed the Russian crisis. Overall mean correlations have increased in the past 15 years. Figure 1 also shows a straight line representing the results of a simple least-squares regression of the mean correlation on a constant and a time trend. The coefﬁcient associated with the time trend is highly signiﬁcant, although very small in absolute terms. Furthermore, there seems to be a jump in mean correlations around 1994, which is probably a result of the completion of the internal banking market in the EU that resulted from the second banking directive implemented in 1993. A test for equality of means shows that the means before and after the implementation of the second banking directive vary signiﬁcantly. Before December 1993, the average mean correlation is 0.0057; after January 1994 it is 0.0832. These results are represented by the horizontal lines in Figure 1. In summary, the analysis provides some evidence that interdependencies among European banks of different countries have become stronger within the past 15 years. Part of this increase in interdependencies can be attributed to the completion of the internal banking market in the EU. Thus, the potential of systemic risk in banking has increased at the European level. Using a similar approach, Schro¨der and Schu¨ler (2003) also ﬁnd some evidence that the systemic risk potential at the European level has increased. They estimate correlations between excess returns of bank stock indices of EU countries using bivariate GARCH models, and apply several tests to assess changes in correlations over time. Our ﬁndings are in line with those of Gropp and Moerman (2003), and they empirically examine contagion in a sample of 67 EU banks using the distance to default method (introduced by Gropp et al., 2003) to measure bank risk. Among other things, they ﬁnd evidence for contagion between banks across borders.

Other justifications for the need to integrate the European supervisory system In addition to the systemic risk rationale, there are further reasons for the need to centralize banking supervision in the EU: As mentioned in Section 1, consumer protection is the other main argument for the regulation and supervision of banks. Bank customers have little individual incentive to perform the various monitoring functions (Dewatripont and Tirole, 1993). This phenomenon gives rise to a need for private or public representatives to protect depositors’ interest. Under this representation hypothesis, centralized supervision makes sense as it avoids the duplication of regulatory costs. In addition, centralization of supervision can limit regulatory arbitrage. Another reason to integrate supervision across borders is the increasing interEuropean and international mobility of ﬁnancial ﬁrms. Today, big European banks are in a position to ‘‘cherry-pick’’ their supervisors by selecting a country

Supervision of the European Banking Market

201

of residence – a practice that might lead to an unintended ‘‘race to the bottom’’ in terms of regulatory practice.13 Thus, there seems to be a need for the integration of banking supervision in Europe. The next section describes the current European supervisory framework, which already exhibits some degree of centralization. The ﬁndings of this section, however, pose the question whether or not the supervisory structure in the EU can safeguard ﬁnancial stability. Section 4 tries to answer the question. 3. Banking supervision in the EU In the EU the institutional arrangements for the supervision of ﬁnancial markets are based on the principles of home country control and mutual recognition, as well as cooperation among the national supervisory authorities (ECB, 2000; Lannoo, 2002).14 According to the home country principle, every bank has the right to do business in the whole euro area using a single license, under the supervision of the authority that has issued the license. Cooperation takes place both at bilateral and multilateral levels. 3.1. The national arrangements Institutional banking supervision arrangements in the EU Member States can be distinguished by the degree of cross-sectoral integration and central-bank involvement. Further, one can identify countries that have adopted a functional model of ﬁnancial supervision, which differentiates between conduct of business and prudential supervision.15 The recent establishment of integrated ﬁnancial supervisory agencies in Austria (April 2002), Germany (May 2002), Ireland (May 2003), and Belgium (January 2004) signiﬁcantly changes the balance of supervisory arrangements (see Table 1).16 A majority of the EU-15 member states have now united their ﬁnancial market supervision under a single agency. Only Greece, Portugal and Spain have three authorities each for the supervision of the banking, securities and

13

This ‘‘cherry-picking’’ of the supervisor may also be important with respect to crossing the line between the banking, insurance, and securities sector. The focus of this paper is, however, on banking. 14 Note at this point that we do not consider competition policy in the banking sector. For an examination of the relationship between competition policies and policies to preserve ﬁnancial stability see, e.g., Canoy et al. (2001), and Carletti and Hartmann (2002). 15 Note that in the literature the terms ‘‘functional approach to supervision’’ and ‘‘supervision by objective’’ are often used to express the same thing. Taylor (1995) uses the expression, ‘‘Twin Peak concept.’’ 16 Table A.1 gives a more detailed description of the national supervisory agencies in EU-15. Recent papers covering the national structures for ﬁnancial supervision are ECB (2002), Lumpkin (2002), and Schoenmaker (2003).

202

Martin Schu¨ler

Table 1. Institutional arrangements for the supervision of financial market at the national level Integrated Supervisor

Securities and Banking Supervisor

Specialized Banking Supervisor

Austria – NCB Belgium Denmark Germany – NCB Ireland – NCB The Netherlands – NCB (functional) Sweden United Kingdom

Finland – NCB France – NCB (functional) Italy – NCB (functional) Luxembourg

Greece – NCB Portugal – NCB Spain – NCB

NCB, National Central Bank at least partly involved in banking supervision; (functional) institutional separation between prudential and conduct of business supervision.

insurance sectors. The remaining countries have combined banking and securities sector supervision. In France, Italy, and the Netherlands, the supervisor responsible for prudential supervision differs from the one in charge of conduct of business supervision. Traditionally, the national central banks (NCBs) have played a signiﬁcant role in banking supervision and have often been solely responsible for it. All three member states with specialized banking supervisors delegate the supervisory authority exclusively to the NCBs. Moreover, the trend towards integrated ﬁnancial supervision did not always result in a separation of supervisory functions from the NCBs: The new Irish supervisor is a distinct component of the Central Bank of Ireland. In Germany and Austria, the central bank has kept its responsibilities in banking supervision. The prudential supervisors in the Netherlands, Italy, and France are built on the banking supervision and ﬁnancial stability function of their respective central banks. Often, the central bank is involved in banking supervision without being directly responsible. In these cases, close cooperation between the supervisor and the central bank remains in place. This cooperation often takes the form of board participation and regular committee meetings of high-level representatives from both institutions. In addition, central banks are closely involved in the operational conduct of supervisory duties. In Belgium, Denmark, Luxembourg, Sweden and the UK, the NCB is not involved in banking supervision but remains involved in the task of ensuring ﬁnancial stability. 3.2. The internationalization of banking supervision Clearly, as a consequence of the increase in the EU-wide systemic risk potential, addressing the threat of systemic risk by supervision at a national level alone no longer seems to be a viable approach, and further efforts to combine

Supervision of the European Banking Market

203

international banking regulations and foster some form of coordination of national supervisors seem advisable. A number of cross-national initiatives along these lines exist. For example, there is the Basel Committee on banking supervision, as well as cooperation at the bilateral and multilateral levels.17 Furthermore, initiatives to reduce the potential for crisis have been taken by international organizations: The Financial Sector Assessment Program (FSAP) is a joint initiative by the International Monetary Fund (IMF) and the World Bank that tries to increase the effectiveness of efforts to promote the soundness of ﬁnancial systems in member countries (Hilbers, 2001).

3.2.1. The Basel Committee on banking supervision The purpose of the Basel Committee on banking supervision is to set international standards and coordinate the work of national regulators. It does not possess any formal supranational supervisory authority. Rather, it formulates broad supervisory standards and guidelines and recommends statements of best practice in the expectation that individual authorities will take steps to implement them through detailed arrangements – statutory or otherwise – which are best suited to their own country’s national systems. In this way, the Committee encourages convergence towards common approaches and common standards without attempting detailed harmonization of member countries’ supervisory techniques. In 1988, the Committee introduced the so-called Basel Capital Accord, which has been introduced not only in member countries but also in virtually all other countries with active international banks. According to the Accord, banks are required to fulﬁll a minimum capital standard of 8%. It exempliﬁes the basic form of international banking regulation. In June 1999, the Committee issued a proposal for a New Capital Adequacy Framework (Basel II) to replace the 1988 Accord. The proposed capital framework consists of three pillars (BIS, 2003): minimum capital requirements, which seek to reﬁne the standardized rules set forth in the 1988 Accord; supervisory review of an institution’s internal assessment process and capital adequacy; and effective use of disclosure to strengthen market discipline as a complement to supervisory efforts. Following extensive interaction with banks and industry groups, a ﬁnal consultative document was issued in April 2003 with a view to introducing the new framework at the end of 2006.18

17

For a more detailed description of the supervisory authorities at the European level, see, e.g., Lannoo (2000), ECB (2000), Speyer (2001), and Lannoo (2002). 18 For an evaluation of the New Basel Capital Accord, see, e.g., Caruana (2003) and Large (2003), and of bank capital regulation, in general, e.g., Santos (2000).

204

Martin Schu¨ler

3.2.2. Memoranda of understanding Memoranda of understanding (MoU) provide the underpinning for cooperation between European supervisors at the bilateral level. MoU are bilateral agreements between banking supervisory authorities to safeguard ﬁnancial stability, and establish a practical framework for regular exchange of information and deﬁne procedures and reciprocal commitments. Nearly all EU Member States have signed MoU with each other. By the end of 1999, some 90 MoU were in place within the EU governing the exchange of information regarding the supervision of cross-border banking activities (Lannoo, 2002). At the multilateral level, there exist MoU that often deal with speciﬁc multinational banking institutions. In addition, in March 2003, an MoU came into effect between the European Central Bank (ECB), the NCBs and banking supervisory authorities of the EU on high-level principles of cooperation in crisis management situations (ECB, 2003). 3.2.3. Exchange of information in multilateral bodies At the European level there exist several committees to promote cooperation between supervisory authorities (ECB, 2000; European Commission, 2000a; Lannoo, 2002) – the Groupe de Contact (GdC), the Banking Advisory Committee (BAC), and the Banking Supervision Committee (BSC). The Groupe de Contact. The GdC, established in 1972, was the ﬁrst European banking supervisors’ committee, and is composed of mid-management banking supervisors from all 18 European Economic Area (EEA) Member States.19 The GdC, which meets at least thrice a year, was established as an informal group to promote practical cooperation and information exchange amongst national banking supervisors within the EEA. All members must be involved in the dayto-day supervision of banks. The GdC discusses developments in supervisory systems from the perspective of the hands-on supervisor. The Groupe de Contact is independent in setting its agenda. However, it may examine and prepare reports for the BAC and the BSC on its own initiative or upon request from those committees. The GdC prepares regular overviews of the solvency and proﬁtability of the EU banking system for the BAC and is assisting the committee in the review of the regulatory capital regime. The Banking Advisory Committee. The BAC was established ‘‘alongside the Commission’’ in 1978. Members include representatives from banking supervisory authorities, Finance Ministries of the member states and the Commission (Internal Market DG). It meets four times a year.20

19

The EEA includes, besides the 15 EU countries, Norway, Iceland and Liechtenstein. Clear indications regarding participants, meeting frequency, and tasks are given in the EU directive setting up the BAC (77/78/EEC).

20

Supervision of the European Banking Market

205

The BAC has to fulﬁll four speciﬁc tasks. First, it assists and advises the Commission on proposals regarding further coordination in the banking sector. This includes authorization requirements for banks, bank supervision, cooperation of supervisory authorities, procedures of cross-border banking activities, and deposit insurance schemes. Second, the BAC is responsible for establishing ratios for the solvency, liquidity, and proﬁtability of credit institutions (cf. GdC). Third, it assists in the implementation of EU Banking Directives. Fourth, the BAC advises the Commission on possible follow-up actions for the implementation of the directives. The BAC considers broad prudential regulatory issues; it does not examine speciﬁc problems of individual credit institutions. Unless the Commission decides otherwise, all discussions and conclusions remain conﬁdential. The Banking Supervision Committee. The BSC of the European System of Central Banks (ESCB) comprises high-level representatives from Member States’ NCBs and the banking supervisory authorities from States where the supervisory role is not fulﬁlled by the central bank. The ECB is represented. The Commission (Internal Market DG and Financial Institutions Directorate) and the Groupe de Contact participate as observers. The BSC, which meets ﬁve times a year, fulﬁlls two roles. First, it assists the ESCB in its statutory tasks in the area of banking supervision and stability of the ﬁnancial system. Second, the BSC is a forum where high-level EU banking supervisors exchange views and information regarding systemic issues and possible implications of the conduct of supervision. It examines the banking sector from a macro-prudential perspective. There is intense discussion – among economists and politicians alike – as to how international institutions cope with the increasing globalization in banking and ﬁnance (Goodhart et al., 1998; Aglietta, 2000; Speyer, 2001; Lannoo, 2002). Often, there are calls for a single European supervisor, or at least for greater cooperation between national supervisors. Political leaders have responded to the discussion by proposing a new supervisory arrangement in the EU.

3.3. Proposed reform of the supervisory arrangements in the EU At the moment the reform of ﬁnancial supervision at the European level is still pending. The start of the reform process marked a joint initiative by the ﬁnance ministers of Britain and Germany, Gordon Brown and Hans Eichel, in April 2002 (Brown and Eichel, 2002). Brown and Eichel initially proposed the creation of a modern, effective supervisory structure at the European level in which ﬁnal responsibility would lie with the national governments. This proposal was intensely discussed and criticized, especially by central bankers, who saw their inﬂuence on banking supervision reduced.

206

Martin Schu¨ler

Based on this proposal, Ecoﬁn mandated the EFC21 to assess and report on possible arrangements for ﬁnancial regulation and supervision. A ﬁnal report on ﬁnancial regulation, supervision and stability was submitted in October 2002 (EFC, 2002).22 This proposal has been approved by Ecoﬁn but still has not been adopted by the EU Parliament, which is insisting on the right to callback or to look again at implementing legislation not conforming to its wishes. Nevertheless, the EFC’s ﬁnal report will most likely lay the basis for future supervisory and regulatory arrangements in the EU. In fact, in November 2003, the European Commission launched a package of measures (including a proposal for a Directive and several Commission Decisions) based on the EFC report, and parts of this package have already been implemented. According to the EFC report, the arrangements that are already in place for securities regulation (Lamfalussy model) are to be extended to the other ﬁnancial sectors based on the existing inter-institutional agreements. Contrary to the initial Brown-Eichel proposal, this proposal includes the central banks in banking supervision, at least to a certain extent. According to the Lamfalussy model, legislation is split into framework principles (level 1) and implementing measures (level 2). In addition, cooperation between supervisors is strengthened to improve implementation (level 3). At level 4, the Commission checks Member States’ compliance with EU legislation and may take legal action against a State suspected of breaching Community law. According to the EFC proposal, the following supervisory structure will be implemented (see Figure 2). The three different sectors of ﬁnancial markets are represented by three sectoral committees each at levels 2 and 3: for banking, insurance (including pensions), and securities (including UCITS). In addition, a fourth committee at level 2 would be established to deal with speciﬁc issues concerning ﬁnancial conglomerates. At level 1, the Commission will adopt formal proposals for a directive/regulation after a full consultation process with the European Parliament and the Council of Ministers. The key political choices to be made by the Parliament and the Council on the basis of the Commission’s proposals are reﬂected in the framework principles. They determine the political direction and orientation, the fundamentals of each decision. At level 2, committees act as regulatory committees in line with the 1999 Council comitology decision.23 Level 2 committees also advise the Commission

21

The EFC was established in 1999. It advises the Commission and the Council on issues concerning the economic and ﬁnancial situation in the EU. 22 Note that the proposal of the EFC refers to supervision and regulation of ﬁnancial markets, comprising banking, insurance, and securities. 23 Comitology refers to the delegation of implementing powers by the Council to the Commission for the execution of EU legislation. Representatives of the Member States, acting through comitology committees (level 2 committees), assist the Commission in the execution of its powers.

Supervision of the European Banking Market

Figure 2.

207

EFC proposal for a new supervisory framework at the EU level. Source: EFC (2002)

on, for example, draft legislative texts. Each committee is chaired by the Commission, and each member state has one voting representative and one supporting technical expert, both nominated by the relevant Ministry. The chair of the respective level 3 committee has observer status without voting power. In the level 2 committees for banking and ﬁnancial conglomerates, additionally, the ECB has observer status. Level 2 committees can occasionally meet in a joint format at a high level to consider difﬁcult technical and cross-sectoral cases, and improve synergies and coherence of level 2 rules.24

24

According to the EFC proposal, the BAC is reformed into the level 2 banking committee. Functions of the current BAC that are not transferred to the level 2 committee are allocated to the level 3 banking committee. The above-mentioned Commission package from November 2003 includes two Commission Decisions (2004/10/EC; 2004/9/EC) setting up, respectively, the European Banking Committee (EBC) and the European Insurance and Occupational Pensions Committee (EIOPC) as the level 2 committees for banking and insurance (cf. Figure 2). The EBC, thus, replaces the BAC.

208

Martin Schu¨ler

At level 3, the committees fulﬁll three tasks25: ﬁrst, advise the Commission, in particular on its preparation of draft level 2 measures; second, promote consistent implementation of EU directives, supervisory convergence, and best practice in Member States; and third, provide an effective operational network to enhance day-to-day supervision, including exchange of information. Voting members are the respective national supervisory authorities of member states, which also chair committees and form the secretariat. Central banks without direct supervisory responsibilities, including the ECB, also participate in the level 3 banking committee. However, the respective national supervisory agency holds the vote. Representatives of the BSC and the GdC have observer status in the level 3 banking committee. The Commission has observer status in all level 3 committees. The GdC continues its activities and acts as the main working group of the level 3 banking committee, including for facilitating conﬁdential exchange of information. Besides the legislative process, a reconﬁgured Financial Services Policy Group (FSPG) provides political oversight of ﬁnancial market issues for the beneﬁt of the Ecoﬁn Council.26 It ﬁlls the gap between the political and technical regulatory levels, and provides for cross-sectoral strategic reﬂection. This group has a policy-shaping role, which includes deﬁning the medium- and long-term strategy for ﬁnancial services issues, such as for the post-FSAP27 period; considering ‘‘hot’’ short-term issues, such as terrorist ﬁnancing, reinsurance, and issues related to the current FSAP phase; and assessing progress and implementation, e.g., of the current FSAP. Political advice and oversight are provided on both internal issues (single market) and external issues (enlargement issues). The group reports to the EFC to assist the EFC in advising the Council. Members of the committee include one high-level representative of the relevant Ministry per member state. In addition, the level 2 committees chairs have full membership. The ECB and the level 3 committees chairs are observers. Additionally, the reconﬁgured FSPG contributes to the EFC’s work on issues related to ﬁnancial stability and related economic topics. To this end, the level 3 committee chairs, the ECB, and the Directorate General for Economic and Financial Affairs (DG Ecoﬁn) are granted full-member status in this body. The BSC chairperson participates as observer and regularly reports on macroprudential developments.

25

According to two Commission Decisions that are part of the Commission package from November 2003, the CEBS is established effective January 1, 2004 (Commission Decision 2004/5/EC), and the Committee of European Insurance and Occupational Pensions Supervisors effective November 24, 2003 (2004/6/EC) (cf. Figure 2). 26 The FSPG ﬁrst met in January 1999. In the context of work to establish the single market for ﬁnancial services, the Ecoﬁn Council asked the FSPG to examine where new legal initiatives would be required, where existing provisions had to be adapted to new developments, where existing provisions needed to be simpliﬁed, and where these should be more coherent. 27 The FSAP, laid out in 1999, seeks to push ﬁnancial market integration in over 40 areas.

Supervision of the European Banking Market

209

4. Evaluation of the supervisory arrangements A number of measures pertain to ensuring the safety and soundness of the ﬁnancial system. In general, prudential supervision comprises regulations, such as capital regulation, monitoring (i.e., the enforcement of banks’ compliance with the regulations), and crisis management (i.e., the safety nets deposit insurance and LOLR). The Lamfalussy framework on which the EFC proposal is based refers primarily to the legislative process and regulatory issues. Among other things, the goal is to speed up EU legislation so that ﬁnancial regulation is able to adapt quickly to new market developments and practices. In addition, the proposed framework aims at promoting convergence of supervisory practices and consistent implementation of EU directives. Furthermore, with regard to regulations, there are the Basel capital requirements that already constitute some form of international regulation. These will be revised by the New Capital Accord with the aim to set bank’s capital requirements closer to the actual economic risk that they face. Certainly, in terms of monitoring, supervision of individual institutions is best carried out at the level closest to the ﬁnancial intermediaries concerned, i.e., at the national level (Meister, 2002; Lejsek, 2002; EFC, 2002). National supervisors have sufﬁcient experience with the regional markets and the supervised entities. In a very practical sense, they can get in contact with the persons and institutions more easily. Nevertheless, increased interdependencies of banks across countries require enhanced coordination and cooperation between national supervisors. The proposed framework may help institutions share information between national supervisors and central banks, which is needed to monitor the amount of systemic risk at the EU level. Furthermore, the new framework may help align supervisory practices among countries, which is crucial for supervising multinational banks. Furthermore, cooperation currently exists at the bilateral and multilateral level due to the MoU and the multilateral bodies at the EU level (Sections 3.2.2 and 3.2.3). However, one should keep in mind that the motivation behind the (original) Lamfalussy model, which referred to securities markets, was the swift implementation of the FSAP. The objective was to smooth the functioning of European securities markets (Lamfalussy, 2004). The focus in banking, on the other hand, is on prudential issues, for which the Lamfalussy framework was not designed.28 Therefore, by extending the Lamfalussy model to banking, macroprudential issues – such as the LOLR issue and the involvement of the central bank – are not addressed. In fact, it is not the purpose of the Lamfalussy model,

28

See Allen and Herring (2001) for the differences between securities market and banking market regulation.

210

Martin Schu¨ler

and hence the framework proposed by the EFC to address such issues.29 Thus, open issues remain with respect to safeguarding ﬁnancial stability. 5. Open issues in crisis management Crisis management comprises deposit insurance and the LOLR function. In all EU countries, explicit deposit insurance schemes are in place and administered either ofﬁcially (by the government or the central bank) or privately.30 Deposit insurance schemes are compulsory for all banks in the EU, and it is clear who refunds deposits in the case of a bank failure. There is no difference between domestic and foreign depositors. Nevertheless, there are differences with respect to the scope of deposit insurance. However, these issues are minor compared with the LOLR issue, which is not clear-cut in the EU. 5.1. Lender of last resort and central bank involvement Who would give liquidity assistance in the case of a major bank failure that had systemic consequences at the European level, threatening banks in many countries? Does Europe need an explicit LOLR? Could the ECB possibly be such an LOLR, or is there already an implicit European LOLR?31 Before turning to the European situation, we sketch the pros and cons of the combination of central banking and banking supervision.32 Arguments in favor of combining central banking and banking supervision One advantage of having banking supervision within the central bank is to exploit synergies between central banking and prudential supervision. Synergies may arise in two respects. First, central banks are well informed about the overall country economy and ﬁnancial system, including information from running national or European payment and settlement systems, and monetary policy operations, both of which assist them in their supervisory tasks. Second, supervisory information can play an important role in the oversight of payment

29

One may thus argue that it is questionable whether the Lamfalussy model should be extended to the banking sector. However, in the context of the paper, the extension is taken as a given. The reform is already agreed upon, and – as mentioned above – the respective committees have already been set up. 30 For a detailed description of deposit insurance schemes, see Demirguc-Kunt and Sobaci (2000), Gropp and Vesala (2001), and Huizinga and Nicode`me (2002). 31 Note that so far we did not explicitly differentiate between Europe, the EU, and the eurozone. Throughout the paper, the notions Europe and EU are used interchangeably. Of course, the ECB is only responsible for the monetary policy of the eurozone. 32 For a more detailed discussion of the pros and cons of central bank involvement in banking supervision, see e.g., Goodhart and Schoenmaker (1993, 1995), Haubrich (1996), and Goodhart (2000).

Supervision of the European Banking Market

211

and settlement systems and of market infrastructures, and in managing liquidity crisis.33 Another argument is the central bank’s inherent concern for the systemic stability of the ﬁnancial system. There is a close relationship between prudential controls of individual institutions and the assessment of risks for the ﬁnancial system as a whole. In this context another point emerges. In order to fulﬁll its LOLR function, the central bank needs to have access to timely information on the solvency and liquidity of banks. This becomes particularly important in times of ﬁnancial distress.34 Furthermore, the protection of the payment system, which is the responsibility of the central bank, is a crucial tool for controlling systemic risk. Finally, there is the independence argument. Central banks tend to be independent, which may enhance the bank supervisor’s ability to enforce actions. Arguments in favor of separation The main argument against central bank involvement in banking supervision is a potential conﬂict of interest between the two functions. To avoid adverse effects on the ﬁnancial health of banks, and hence on ﬁnancial stability, the central bank may pursue a more accommodating monetary policy than warranted for the pursuance of price stability.35 Furthermore, the combination of functions raises a risk that the central bank’s credibility may be diminished. If bank failures occur, it could hurt the central bank’s global credibility. The private sector might conclude that the central bank’s monetary policy decisions may be inﬂuenced by ﬁnancial system stability considerations. Reduced credibility could result in higher inﬂation expectations, which in turn might increase inﬂation.36 Finally, some argue that banking supervision within the central bank results in an excessive concentration of power in the central bank.

33

For the United States, Peek et al. (1999) ﬁnd that supervisory information does affect monetary policy, and does so in the correct direction. Their results support the idea that supervisory information allows more accurate estimates of economic activity and inﬂationary pressures to be achieved, thereby favouring the choice of a more appropriate stance for monetary policy. 34 Goodhart and Schoenmaker (1995) ﬁnd that in countries in which the central bank is the supervisory authority there were less bank failures. 35 Note that the argument that there is a conﬂict of interest between banking supervision and the conduct of monetary policy breaks down to a potential conﬂict of interest between ﬁnancial and monetary stability. Consequently, if such a conﬂict existed, it would be irrespective of whether the central bank is involved in banking supervision. For a discussion of the relationship between monetary and ﬁnancial stability, see e.g., Crockett (2001). Empirical evidence by Herrero and Rio (2003) suggests that focusing on price stability reduces the likelihood of a banking crisis. 36 Di Noia and Di Giorgio (1999) ﬁnd empirical evidence that the inﬂation rate is higher and more volatile in countries where the responsibility of banking supervision is entirely placed within the central bank.

212

Martin Schu¨ler

5.2. The European situation – do we need a European LOLR? With respect to the involvement of the central bank in banking supervision, Europe is unique in the sense that, since the introduction of the euro, the jurisdiction of monetary policy and banking supervision does not overlap. According to the ECB (2001), this has changed the balance of the arguments considerably. Arguments against central bank involvement have lost much of their weight, while those in favor have become even more prominent. In particular, the increasing focus on the potential for systemic failures supports the involvement of the central banks. NCBs may beneﬁt from their traditional concern for systemic stability and from their composite nature – they are components of the Eurosystem and, simultaneously, a national institution. Furthermore, conﬂicts of interest and concentration of power are not a real concern owing to the institutional separation of monetary and supervisory jurisdiction. Regarding crisis management, the LOLR function is of crucial importance. The ECB does not carry out banking supervision nor does it have a mandate as a European LOLR. The contribution of the ECB to ﬁnancial stability is mentioned under Article 105(5) of the Treaty, according to which, ‘‘The ESCB shall contribute to the smooth conduct of policies pursued by the competent authorities relating to the prudential supervision of credit institutions and the stability of the ﬁnancial system.’’ This language is ambiguous. The responsibility is diffused and the ‘‘competent authorities’’ are not named. The ambiguity regarding crisis management has also lead to the so-called Brouwer reports (European Commission, 2000b, 2001), which were prepared by the EFC under the chairmanship of Henk Brouwer from De Nederlandsche Bank. The conclusion of these reports was that the current system based on national responsibility is appropriate, but there is a need to strengthen crossborder and cross-sectoral cooperation between supervisors, to enhance convergence of supervisory practices, and to reinforce collaboration and cooperation between supervisors and central banks. These tasks may be achieved by the proposed framework. There is criticism that the absence of a clear and transparent LOLR that would act in an emergency raises doubts in the markets about the ability of the Eurosystem to handle crisis situations (e.g., Aglietta, 2000; Prati and Schinasi, 2000). According to Padoa-Schioppa (1999), this criticism is not justiﬁed. His argument has three parts. First, this criticism reﬂects a notion of LOLR operations that is largely outdated. Padoa-Schioppa argues that due to deposit insurance, large exposure rules, and other supervisory measures, contagion from an insolvent to a solvent institution has become very unlikely. In addition, a rapid outﬂow of uninsured inter-bank liabilities would be absorbed by the width and depth of today’s inter-bank market. Thus, it is extremely unlikely that a solvent bank would lose liquidity at the same time that it lacks sufﬁcient collateral to obtain regular central bank funding. Second, he argues that the criticism underestimates the Eurosystem’s capacity to act. Not to declare explicitly

Supervision of the European Banking Market

213

beforehand the procedures of emergency actions may even be advisable, because it may help reduce the moral hazard associated with the safety net. This policy of ‘‘constructive ambiguity’’ was also stressed by Wim Duisenberg (1999), former ECB president. And third, the criticism represents too mechanistic a view of how a crisis should be managed. Deviation from normal-time procedures and rules is allowed and even required in an emergency. Furthermore, Padoa-Schioppa (1999) stresses that besides the central bank money solution to a ﬁnancial crisis, there are alternative arrangements to provide money. In particular, there are private money solution and funds raised with taxes. There is no doubt that the LOLR must be conﬁned to play a minor role. However, these alternative solutions ‘‘are costly and can in general be at least part of a solution in which the central bank is also involved’’ (Vives, 2001: p. 76). There exists no explicit European LOLR, and as argued by Padoa-Schioppa (1999), there may also be no requirement for it – at least not at the present time. Nevertheless, there may already exist an implicit European LOLR. The 1999 Annual Report of the ECB37 disclosed that a bank, in an emergency, could rely on the so-called emergency liquidity assistance (ELA) (Scacciavillani et al., 2002). The ELA comprises the support given by the NCB to temporarily illiquid institutions, and is only given ‘‘in exceptional circumstances and on a case-bycase basis.’’ The NCBs take the decision concerning the provision of ELA to an institution operating in its jurisdiction. All costs associated with the provision of the ELA must be borne by the NCBs. Any provision of ELA that would endanger price stability could be sterilized by the ECB. In its Annual Report of 1999, the ECB stresses: ‘‘Central bank support should not be seen as a primary means for ensuring ﬁnancial stability, since it bears the risk of moral hazard’’ (p. 98). However, it is clearly stated: ‘‘If and when appropriate, the necessary mechanisms to tackle a ﬁnancial crisis are in place’’ (p. 98). This is in line with a statement given by Wim Duisenberg (1999) and the view of Padoa-Schioppa (1999) that the Eurosystem would have the necessary capacity to act when it needed to, and that a clear reassurance about the capacity to act should be sufﬁcient for the markets. This view is also supported by the reactions to September 11, 2001, where the ECB acted to guarantee the availability of dollars to cover needs arising in the ﬁnancial system (Scacciavillani et al., 2002). Vives (2001, p. 70) argues that ad hoc coordination during crises will not be sufﬁcient and may endanger the stability of the system. As a consequence, he urges for a centralized LOLR authority, namely the ECB. The discussion of whether or not there should be more explicit and transparent arrangements for a European LOLR breaks down to this question: What amount of ambiguity in LOLR rules is still ‘‘constructive’’? There may be no doubt that the Eurosystem has the necessary capacity and willingness to act

37

See ECB (2000b)

214

Martin Schu¨ler

when needed. And, of course, it must not be too easy to forecast the ﬁnal decision on granting LOLR. However, the decision process itself has to be transparent (Enoch et al., 1997; Bruni and De Boissieu, 2000; Fellgett, 2003). Another important issue arising in the context of crisis management is failure or insolvency of banks with cross-border effects. Several questions arise regarding this issue: How will the amount of a rescue be decided, and who will pay for a failed insolvent transnational institution that has gone bankrupt after being helped? How will the losses be eventually shared among the ﬁscal authorities? Since a discussion of these ﬁscal issues is beyond the scope of this paper, the interested reader is referred to the existing literature.38 With respect to crisis management, the recently signed memorandum of understanding between the ECB, the NCBs, and the banking supervisory authorities establishes principals and procedures that aim at enhancing the practical arrangements for handling crises at the EU level.39 These principles and procedures deal with the identiﬁcation of the authorities responsible for crisis management, the information ﬂows between the involved authorities, and a logistical infrastructure to support the enhanced cross-border cooperation in crisis situations. 5.3. A European observatory of systemic risk Coming back to monitoring in a broader sense, there is a further issue that should be considered. A supervisor should observe systemic risk in general. Thus, with respect to the increased Europe-wide systemic risk potential and the prevention of a potential cross-border systemic crisis, there is a need for a ‘‘European observatory of systemic risk’’ (EOSR) (European Shadow Financial Regulatory Committee, 1998; Aglietta, 2000; Meister, 2000). Simple cooperation between national supervisors, even in the new framework, will not be sufﬁcient. The newly established committees at the European level, for example, do not have any power to request information from the national authorities. The EOSR should be endowed with a permanent staff to fulﬁll its tasks. It must be in a position to obtain substantive, mandatory, and timely information transmitted by the national supervisory agencies and central banks. It would aggregate the data and coordinate the multilateral information ﬂow with the national authorities and the ECB. This committee does not necessarily possess decision-making power. However, it would follow, understand, and interpret Europe-wide market developments, and alert national and European authorities

38

See, e.g., Prati and Schinasi (2000), and Schoenmaker (2000, 2003). Baxter et al. (2004) discuss international insolvency systems in a more general manner. 39 Note that this memorandum of understanding, as well as other MoU, is not a public document. Information on the content given here is drawn from the ECB’s Press Release (ECB, 2002). Consequently, there still remain questions about the transparency of the decision process in a crisis.

Supervision of the European Banking Market

215

to exposures to a potential systemic impact. Thus, the national and European authorities (in particular the ECB as the potential European LOLR) could make their decisions based on better information. Primarily, this committee would be essential for crisis prevention (Aglietta, 2000). In normal times it would work as a warning agency and a research center on systemic risk issues. In addition, it could be valuable for optimizing the use of the LOLR when it becomes unavoidable. In times of crisis, it could operate as a task force for the ECB board and the council of governors to restore ﬁnancial stability efﬁciently. The necessary proximity to the ECB and the ESCB would be fulﬁlled, e.g., by a reformed BSC. Besides these tasks, the EOSR could also ensure common supervisory and transparency standards (Bruni and De Boissieu, 2000). In this respect, there would be an overlap with the tasks of the newly established Committee of European Banking Supervisors (CEBS; cf. Section 3.3). However, the focus of the EOSR would clearly be on prudential issues.

6. Conclusions The integration of ﬁnancial markets in the EU – not just since the introduction of the euro – has increased the systemic risk potential at the European level. The divergence between this increase in EU-wide systemic risk and the current nation-based supervisory structure calls for a reform of the European supervisory framework. A milestone in the political discussion has been a report by the EFC, which will probably lay the basis for a future supervisory structure in the EU. According to this report, the Lamfalussy framework is to be extended to the banking sector based on the existing international agreements. The new structure aims at speeding up EU legislation, strengthening cooperation between national supervisors, and promoting convergence of supervisory practice. Supervision of individual institutions is best carried out at the level closest to the ﬁnancial intermediaries concerned, i.e., the national level. However, the increased EU-wide systemic risk calls for greater cooperation between national supervisors. This task may be fulﬁlled by the new framework. With regard to crisis management, although there is no explicit European LOLR, the Eurosystem can be regarded as an implicit LOLR having the capacity and willingness to act when needed. Furthermore, the memorandum of understanding between the potential parties involved in crisis management may help enhance practical arrangements for handling crises at the EU level. Nevertheless, more transparency concerning the decision-process seems preferable. In addition, EU-wide systemic risk calls for a EOSR. A revised and more powerful BSC could fulﬁll this task. Such a committee would be essential with regard to crisis prevention and could also be valuable for optimizing the use of the LOLR when it becomes unavoidable.

216

Martin Schu¨ler

References Abrams, R.K. and M.W. Taylor (2002), ‘‘Assessing the case for uniﬁed ﬁnancial sector supervision’’, London School of Economics Financial Market Group Special Paper, No. 134. Aglietta, M. (2000), ‘‘A lender of last resort for Europe,’’ pp. 31–67 in: C. Goodhart, editor, Which Lender of Last Resort for Europe? London: Central Bank Publication.Allen, F. and R. Herring (2001), ‘‘Banking regulation versus securities market regulation’’, Wharton Financial Institutions Center Working Paper, No. 01-29. Angelini, P., G. Maresca and D. Russo (1996), ‘‘Systemic risk in the netting system’’, Journal of Banking & Finance, Vol. 20(5), pp. 853–868. Bank for International Settlement (2003), ‘‘Overview of the new Basel Capital Accord, consultative document’’, Basel Committee of banking supervision, April 2003, Basel. Baxter, T.C. Jr., J.M. Hansen and J.H. Sommer (2004), ‘‘Two cheers for territoriality: an essay on international bank insolvency law’’, American Bankruptcy Law Journal, Vol. 78(Winter), pp. 57–91. Benston, G.J. and G.G. Kaufman (1996), ‘‘The appropriate role of bank regulation’’, The Economic Journal, Vol. 106(May), pp. 688–697. Briault, C. (2002), ‘‘Revisiting the rational for a single national ﬁnancial services regulator’’, FSA Occasional Paper Series, No. 16, London. Brown, G. and H. Eichel (2002), ‘‘Structures for European ﬁnancial stability and supervision of ﬁnancial institutions’’, Draft joint letter from Gordon Brown and Hans Eichel to Rodrigo Rato, April 11, 2002. Bruni, B. and C. De Boissieu (2000), ‘‘Lending of last resort and systemic stability in the eurozone’’, pp. 175–196 in: C. Goodhart, editor, Which Lender of Last Resort for Europe?, London: Central Bank Publication. Canoy, M., M. van Dijk, J. Lemmen, R. De Mooij and J. Weigand (2001), ‘‘Competition and stability in banking’’, CPB Document, No. 015, The Hague. Carletti, E. and P. Hartmann (2002), ‘‘Competition and stability: what’s special about banking?’’ ECB Working Paper, No. 146. Caruana, J. (2003), ‘‘The importance of transparency and market discipline approaches in the New Capital Accord’’, Keynote speech at the Market Discipline Conference, Federal Reserve Bank of Chicago and BIS, November 1, 2003, Chicago. Crockett, A. (2001), ‘‘Monetary policy and ﬁnancial stability’’, Speech given at the Fourth HKMA Distinguished Lecture, February 13, 2001, Hong Kong. De Bandt, O. and P. Hartmann (2000), ‘‘Systematic risk: a survey’’, ECB Working Paper, No. 35, Frankfurt. Demirguc-Kunt, A. and T. Sobaci (2000), ‘‘Deposit insurance around the world: a data base’’, Mimeo, World Bank.

Supervision of the European Banking Market

217

De Nicolo, G. and M.L. Kwast (2002), ‘‘Systemic risk and ﬁnancial consolidation: are they related?’’, Journal of Banking & Finance, Vol. 26(5), pp. 861–880. Dewatripont, M. and J. Tirole (1993), The Prudential Regulation of Banks, Cambridge/London: MIT Press. Diamond, D.W. and P.H. Dybvig (1983), ‘‘Bank runs, deposit insurance, and liquidity’’, The Journal of Political Economy, Vol. 91(3), pp. 401–419. Di Noia, C. and G. Di Giorgio (1999), ‘‘Should banking supervision and monetary policy tasks be given to different agencies?’’, International Finance, Vol. 2(3), pp. 361–378. Dow, S.C. (1996), ‘‘Why the banking system should be regulated’’, The Economic Journal, Vol. 106(May), pp. 698–707. Dowd, K. (1996), ‘‘The case for ﬁnancial laissez-faire’’, The Economic Journal, Vol. 106(May), pp. 679–687. Duisenberg, W. (1999), ‘‘Introductory statement delivered on the occasion of the presentation of the ECB’s Annual Report 1998 to the European Parliament’’, Strasbourg, October 26. Economic and Financial Committee (2002), EFC report on ﬁnancial regulation, supervision and stability, revised version, October. Enoch, C., P. Stella and M. Khamis (1997), ‘‘Transparency and ambiguity in central bank safety net operations’’, IMF Working Paper, No. WP/ 97/138. European Central Bank (2000), ‘‘EMU and banking supervision’’, ECB Monthly Bulletin, April 2000, Frankfurt, pp. 49–64. European Central Bank (2001), ‘‘The role of central banks in prudential supervision’’, April 2001, Frankfurt. European Central Bank (2002), ‘‘Structural analysis of the EU banking sector: year 2001’’, November 2002, Frankfurt. European Central Bank (2003), Memorandum of Understanding on high-level principles of cooperation between the banking supervisors and central banks of the European Union in crisis management situations, ECB Press Release, March 10, 2003, Frankfurt. European Commission (2000a), ‘‘Institutional arrangements for the regulation and supervision of the ﬁnancial sector’’, Brussels. European Commission (2000b), ‘‘Report on ﬁnancial stability’’, Economic Papers, No. 143. European Commission (2001), ‘‘Report on ﬁnancial crisis management’’, Economic Papers, No. 156. European Shadow Financial Regulatory Committee (1998), EMU, ‘‘The ECB and ﬁnancial supervision’’, Statement No. 2, October. Fellgett, R. (2003), ‘‘Comments on Freixas, X., Crisis Management in Europe’’, pp. 120–122 in: J.J.M. Kremers, D. Schoenmaker and P. Wierts, editors, Financial Supervision in Europe, : Cheltenham.

218

Martin Schu¨ler

Goodhart, C. (2000), ‘‘The organisational structure of banking supervision’’, Financial Stability Institute Occasional Papers, No. 1, Bank for International Settlements, Basel. Goodhart, C., P. Hartmann, D. Llewellyn, L. Rojas-Sua´rez and S. Weisbrod (1998), Financial Regulation: Why, how and where now?, London: Routledge. Goodhart, C. and D. Schoenmaker (1993), ‘‘Institutional separation between supervisory and monetary agencies’’, LSE Financial Market Group, Special Paper, No. 52. Goodhart, C. and D. Schoenmaker (1995), ‘‘Should the functions of monetary policy and banking supervision be separated?’’, Oxford Economic Papers, Vol. 47(4), pp. 539–560. Gropp, R. and G. Moerman (2003), ‘‘Measurement of contagion in banks’ equity prices’’, ECB Working Paper, No. 297, Frankfurt. Gropp, R. and J. Vesala (2001), ‘‘Deposit insurance and moral hazard: does the counterfactual matter’’, ECB Working Paper, No. 47. Gropp, R., J. Vesala and G. Vulpes (2003), ‘‘Market indicators, bank fragility and indirect market discipline’’, in FRBNY Economic Policy Review, forthcoming. Haubrich, J.G. (1996), ‘‘Combining bank supervision and monetary policy’’, Federal Reserve Bank of Cleveland, Economic Commentary Series, November 1996. Hilbers, P. (2001), ‘‘The IMF/World Bank ﬁnancial sector assessment program’’, IMF Economic Perspectives, February 2001. Huizinga, H. and G. Nicode`me (2002), ‘‘Deposit insurance and international bank deposits’’, European Commission Economic Papers, No. 164. Kaufmann, G.G. (1995), ‘‘Comments on systemic risk’’, pp. 47–54 in: G. Kaufman, editor, Banking, Financial Markets, and Systemic Risk, Greenwich: Jai Press. Kaufmann, G.G. (1996), ‘‘Bank failures, systematic risk, and bank regulation’’, The Cato Journal, Vol. 16(1). Lamfalussy, A. (2004), Keynote speech on European bond markets at the Symposium of the ECB-CFS research network on ‘‘Capital Markets and Financial Integration in Europe’’ hosted by the European Central Bank, May 10–11 2004, Frankfurt. Lannoo, K. (2000), ‘‘Challenges to the structure of ﬁnancial supervision in the EU’’, Report of a CEPS Working Party, Brussels. Lannoo, K. (2002), ‘‘Supervising the European Financial System’’, CEPS Policy Brief, No. 21. Large, A. (2003), ‘‘Basel II and systemic stability’’, Speech at the British Bankers’ Association – Basel II/CAD 3 Conference, March 13,2003, London. Lejsek, A. (2002), ‘‘Financial services in the era of the euro and e-commerce: does home country control work?’’ Austrian Report to the International Federation for European Law. Llewellyn, D. (1999), ‘‘The economic rationale for ﬁnancial regulation’’, FSA Occasional Paper Series, No. 1. Lumpkin, S. (2002), ‘‘Supervision of ﬁnancial services in the OECD area’’, Financial Market Trends, No. 81, Paris: OECD.

Supervision of the European Banking Market

219

Meister, E. (2000), ‘‘Challenges for regulators in Europe’’, Speech at the Joint Bundesbank/BIS conference on ‘‘Recent developments in ﬁnancial systems and the challenges for economic policy’’, Frankfurt, September 28–29, 2000. Meister, E. (2002), European Parliament hearing, ‘‘After Enron: ﬁnancial supervision in Europe’’, Statement, July 10, 2002, Brussels. Michael, I. (1998), ‘‘Financial interlinkages and systemic risk’’, Financial Stability Review, Vol. 4(Spring), pp. 26–33. Padoa-Schioppa, T. (1999), ‘‘EMU and banking supervision’’, International Finance, Vol. 2(2), pp. 295–309. Padoa-Schioppa, T. (2002), Statement on ‘‘EU structures for ﬁnancial regulation, supervision, and stability’’, Public hearing on ‘‘After Enron: ﬁnancial supervision in Europe’’, at the Committee on Economic and Monetary Affairs of the European Parliament, Brussels, July 10, 2002. Peek, J., E.S. Rosengren and G.M.B. Tootell (1999), ‘‘Is bank supervision central to central Banking?’’, The Quarterly Journal of Economics, Vol. 114(2), pp. 629–653. Prati, A. and G.J. Schinasi (2000), ‘‘Financial stability in European Economic and Monetary Union’’, pp. 71–173 in: C. Goodhart, editor, Which Lender of Last Resort for Europe?, London: Central Banking Publications. Santos, J.A.C. (2000), ‘‘Bank capital regulation in contemporary banking theory: a review of the literature’’, BIS Working Paper, No. 90, September 2000, Basel. Scacciavillani, F., D. Schumacher, and K. Daly (2002), ‘‘Focus: Euroland, meeting the threat of ﬁnancial instability’’, Goldman Sachs European Weekly Analyst, October 11, 2002. Schoenmaker, D. (2000), ‘‘What kind of ﬁnancial stability for Europe?’’, pp. 215–223 in: C. Goodhart, editor, Which Lender of Last Resort for Europe?, London: Central Banking Publications. Schoenmaker, D. (2003), ‘‘Financial supervision: from national to European?’’, Financiele & Monetaire Studies, Vol. 22(1). Schro¨der, M. and M. Schu¨ler (2003), The systemic risk potential in European banking – evidence from bivariate GARCH models, : mimeo. Schu¨ler, M. (2002), ‘‘The threat of systemic risk in European banking’’, Quarterly Journal of Business and Economics, Vol. 41(3/4), pp. 145–165. Sheldon, G. and M. Maurer (1998), ‘‘Interbank lending and systemic risk: an empirical analysis for Switzerland’’, Swiss Journal of Economics and Statistics, Vol. 134(4.2), pp. 685–704. Speyer, B. (2001), ‘‘Internationalisation of banking and banking supervision’’, Deutsche Bank Research Notes in Economics & Statistics, No. 01–7, Frankfurt. Spong, K., (1994). Banking Regulation: Its Purposes, Implementation, and Effects, (4th Edition), Federal Reserve Bank of Kansas City, Kansas City. Taylor, M. (1995), Twin peaks: a regulatory structure for the new century, London: Centre for the Study of Financial Innovation. Vives, X. (2001), ‘‘Restructuring ﬁnancial regulation in the European Monetary Union’’, Journal of Financial Services Research, Vol. 19(1), pp. 57–82.

220

Appendix Table A.1.

Banking supervisory agencies in the EU Member States

Competent banking supervisory agency

Scope of supervision

Notes

Web address

Austria

Financial Market Authority (Finanzmarktaufsicht, FMA)

Banking, insurance & securities

The FMA is an independent institution under public law, and its independence is secured by constitutional provision. The FMA is legally required to cooperate with the Central Bank (Oesterreichische Nationalbank) and reports to parliament.

www.fma.gv.at

Belgium

Banking, Finance and Insurance Commission (CBFA) Danish Financial Supervisory Authority (Finanstilsynet) Financial Supervision Authority (FSA)

Banking, insurance & securities

www.cbfa.be

Banking, insurance & securities

www.ftnet.dk

Commission Bancaire based at the Banque de France (NCB) and Autorite´ des marche´s ﬁnanciers (AMF)

Banking & securities

Denmark

Finland

France

Banking & securities

The FSA operates in connection with the central bank, but is an independent decision making organization. The members of the Commission Bancaire are appointed by the Minister For Economic Affairs, Finance, and Industry. It is chaired by the governor of the Banque de France. The Banque de France provides staff and funds for the operations of the Commission Bancaire.

www.rata.bof.ﬁ

www.commissionbancaire.org; www.amf-france.org

Martin Schu¨ler

Country

Germany

Banking, insurance & securities

Banking Banking, insurance & securities

Italy

Banca d’Italia (NCB) and Commissione Nazional per le Societa` e la Borsa (CONSOB)

Banking & securities

Luxembourg

Commission de Surveillance du Secteur Financier (CSSF) De Nederlandsche Bank (NCB) and Autoriteit Financiele Markten (AFM)

Banking & securities

The Netherlands

Banking, insurance & securities

The Financial Services Regulator is a distinct component of the Central Bank and Financial Services Authority of Ireland, with clearly deﬁned regulatory responsibilities. Functional approach: The Banca d’Italia is responsible for the prudential supervision of banks and securities ﬁrms. CONSOB is responsible for transparency and proper conduct

www.baﬁn.de

www.bankofgreece.gr www.ifsra.ie

www.bancaditalia.it; www.consob.it

www.cssf.lu

www.dnb.nl; www.autoriteit-fm.nl

221

Functional approach: De Nederlandsche Bank is responsible for prudential supervision, whereas the Autoriteit Financiele Markten is responsible for conduct of business supervision

Supervision of the European Banking Market

Greece Ireland

Federal Financial Supervisory Authority (Bundesanstalt fu¨r Finanzmarktaufsicht, BAFin) Bank of Greece (NCB) Irish Financial Services Regulatory Authority

Functional approach: The Commission Bancaire is also responsible for the prudential supervision of the securities market. The Autorite´ des marche´s ﬁnanciers (AMF) established in August 2003 is responsible for conduct of business supervision. The Bundesbank is involved in banking supervision on behalf of the Federal Financial Supervisory Authority.

222

Table A.1 (Continued) Country

Competent banking supervisory agency

Scope of supervision

Portugal

Banco de Portugal (NCB) Banco de Espan˜a (NCB) Financial Supervisory Authority (Finansinspektionen) Financial Services Authority (FSA)

Banking

www.bportugal.pt

Banking Banking, insurance & securities

www.bde.es www.ﬁ.se

Banking, insurance & securities

www.fsa.gov.uk

Spain Sweden

United Kingdom

Notes

Web address

Martin Schu¨ler

Source: ECB (2002), Lumpkin (2002), Schoenmaker (2003), as well as the websites and the annual reports of the respective national banking supervisory agencies.

Comment Tuomas Takalo1

1. Introduction In contrast to monetary policy, the supervision and crisis management of the ﬁnancial services sector in the Economic and Monetary Union is a responsibility of national authorities. Over the past few years, however, a pressure to reform ﬁnancial market regulation and supervision has been mounting and, accordingly, a number of reforms are ongoing. In his clear and concise essay Martin Schu¨ler argues that because of the increased systemic risk potential in Europe, the pending reforms are needed and further centralization of supervision and crisis management might be justiﬁed in some areas, e.g., in the monitoring of soundness of the European banking system. The structure of Schu¨ler’s essay is logical. It begins with an argument that the systemic risk potential in banking has increased in Europe. This gives a rationale for a further integration of ﬁnancial market regulation and supervision. Against this rationale, Schu¨ler describes and evaluates the current supervisory framework and the Economic and Financial Committee’s (EFC) proposal for a new framework. Finally, the open issues in the management of a systemic crisis are discussed. The discussion includes a proposal for a European observatory of systemic risk. Schu¨ler’s essay is also remarkably well written. Given the complexity of the current supervisory arrangements and the EFC’s proposal, the value of their clear description cannot be underestimated. I have only two substantive critical remarks that I will elaborate in what follows. Anyone writing on the reforms of ﬁnancial market regulation in the

1

I thank David Mayes, Emmi Martikainen and Elina Rainio for useful discussions when preparing this comment.

CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79015-0

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

224

Tuomas Takalo

European Union (EU) faces the challenging task of improving upon Vives (2001) which, at least to me, forms a landmark study among the non-technical evaluations of the European ﬁnancial market regulation. Because Vives’s article was written before the EFC’s proposal, there would be room for a careful evaluation of the proposal but, unfortunately, Schu¨ler is a bit sketchy in his evaluation. My second concern is that I am neither sure that the systemic risk potential has increased in Europe nor that this still should be the principal justiﬁcation for ﬁnancial regulation in the EU. I expand upon my comments in the reverse order, beginning with the systemic risk rationale for the regulation. 2. Systemic risk potential in Europe and the rationale for banking regulation Different approaches to measure systemic risk have been tried over recent years, but the literature has neither been able to identify an ideal measure nor coordinate a standard practice (see De Bandt and Hartmann (2001) for a survey, and Gropp and Moerman (2004) for a recent attempt). Building on the measurement methodology developed by De Nicolo and Kwast (2002), Schu¨ler argues that the systemic risk potential has signiﬁcantly increased in Europe since the mid-1980s. The systemic risk potential is measured by the correlations between pairs of stock returns of the 54 largest European banks in 13 countries, after controlling for national macroeconomic factors. The measurement practice raises the question of what the captured substantial increase in the correlations between stock returns really tells us. For example, there is evidence of integration of the European banking sector over the same period (Beale et al., 2004). Integration increases the correlation of the stock returns almost by deﬁnition. Schu¨ler recognizes the measurement difﬁculties and carefully interprets the increased correlation as a sign of increased systemic risk potential. In my view, nonetheless, an increase in such systemic risk potential provides a rather weak case for the reform of banking supervision and regulation in the EU. Although systemic risk has traditionally been seen as a key justiﬁcation for banking regulation and supervision, the question of whether systemic risk has increased or integration has deepened or both is beside the point regarding the need of a regulatory reform. Even if the ﬁnancial market integration decreased systemic risk, there would be a rationale for a centralized regulatory and supervisory authority. As Dell’Ariccia and Marquez (2003) show, a centralized regulator is beneﬁcial only if ﬁnancial markets are sufﬁciently integrated. Moreover, the representation hypothesis of bank regulation (Dewatripont and Tirole, 1994) suggests a single regulator because it avoids the duplication of regulatory costs. In line with this argument, Nordea, the leading ﬁnancial services group in the Nordic and Baltic Sea region will change its organization from a subsidiary structure to a European company (Societas Europaea). According to Nordea, the move to a branch structure is made to overcome obstacles created by current regulations that subject Nordea to four sets of legislation, supervised by four regulatory systems and supervisory authorities.

Comment

225

Even if the right justiﬁcation for the regulation is immaterial for the need of regulatory reform, it could be relevant for the details of the reform agenda. I will next suggest that taking the representation hypothesis as the starting point would not essentially change Schu¨ler’s conclusions but rather support them. 3. Prevailing financial market supervision in the EU and the EFC proposal for reform As mentioned, the supervision and crisis management of ﬁnancial markets in Europe are responsibilities of national authorities. Although there is some tendency towards a uniﬁed national authority that combines the supervision of banking, securities and insurance sectors, there is a wide variety of national arrangements among the EU countries.2 The national authorities cooperate to take care of cross-border dimension of supervision and crisis management. The cooperation is based on home country and mutual recognition principles, taking place on multiple platforms, in particular, through Basel Committee on banking supervision, Memoranda of Understanding, the Groupe de Contact, the Banking Advisory Committee, and the Banking Supervision Committee. From Schu¨ler’s description it becomes quite clear that the current regulatory and supervisory framework is too convoluted to successfully deal with Europewide supervision and crisis management. The problems have been recognized at a high level, and the EFC’s proposal for reforming regulation and supervision in the EU is a result of these concerns. The EFC’s recommendation is that the Lamfalussy model of securities market regulation and supervision should be extended to the banking and insurance markets. As a result, the new supervisory and regulatory framework will be a complicated multilevel arrangement based on cooperation of the national authorities in various committees (comitology). The proposed framework introduces a number of new committees without abolishing existing cooperative organizations. 3.1. Evaluation of the EFC’s proposal Since systemic risk is taken as the rationale for banking regulation, it is logical to ask how the proposed new framework will cope with systemic risk. Given the laborious comitology structure of the new framework, the answer is not obvious. As Schu¨ler points out, the new framework, like its prototype, the Lamfalussy model of securities market regulation and supervision, mainly aims at governing legislative regulation. For this purpose, the new framework may be suitable: for

2

Schu¨ler deals only with the 15 countries that previously formed the EU. Of the current 25 members, 11 have a uniﬁed supervisory authority, but in the rest of the countries supervision is divided among 2–6 authorities. The national supervisory arrangements further differ in the degree of the central bank involvement, and in some countries prudential and business conduct supervision are separated.

226

Tuomas Takalo

example, the extensive cooperation of the national authorities will certainly promote the common implementation of regulation at the national level. However, the ability of the new framework to react to market developments is much in doubt, even in restricting the attention to preventive legislative actions, not to mention the actions needed in a crisis. The main rationale for securities market regulation has traditionally been investor protection, not systemic risk, as Allen and Herring (2001) argue. It is therefore questionable whether the Lamfalussy model should be extended to banking and insurance markets. In contrast, if the protection of small depositors was assumed as the rationale for banking regulation in line with the representation hypothesis, the Lamfalussy model would be a natural candidate for banking and insurance regulation and supervision. In that case it would be valuable to know how well the Lamfalussy model is working in securities markets. The new proposal leaves monitoring of ﬁnancial institutions at the national level. Schu¨ler supports this arrangement on the grounds of the informational advantage of the national authorities over the local supervised institutions. Schu¨ler’s view here differs from Vives (2001) who makes a political economy argument in favor of an independent European Financial Services Authority. Schu¨ler’s view can be defended by making an analogy to banking. Degryse and Ongena (2003) show that national borders drastically increase the monitoring difﬁculties banks encounter with their customers. I see no reason why the relationship between supervisors and banks should be different. One can, however, expect that technological and regulatory progress will soon reache a stage where monitoring information can be codiﬁed and to be transformed across the borders. Once we experience a rise of truly European banks, further centralization of monitoring of ﬁnancial institutions from the national to European level will be warranted. At this stage of the development, where national borders still hinder the hardening of information, the representation hypothesis of bank regulation dictates a need to further cut down branches of the National Central Banks. Schu¨ler argues that in contrast to the monitoring of single institutions, there would be a need for a European monitor of systemic risk. Such an institution would be responsible for assessing the soundness of the entire European ﬁnancial system. I share this view in principle, but many questions remain unanswered in Schu¨ler’s essay. For example, why cannot the European Central Bank (ECB) act as the monitor? The ECB is certainly already taking care of this task (see, e.g., the ECB reports ‘‘EU banking sector stability’’ and ‘‘Structural analysis of the EU banking sector’’). If the European observatory of systemic risk is not the ECB, what would be its mandate and method of operation? To whom should it be accountable? As to deposit insurance, Schu¨ler argues that there is no need for a European deposit insurance scheme, because national schemes are mandatory and have clear jurisdictions in case of bank failures. I do not necessarily agree, as there is a large variation in the scope of deposit insurance in the EU. In the case of major

Comment

227

cross-border banks such as Nordea, there are also problems in transferring funds between national deposit insurance schemes. But I too think that the open issues concerning deposit insurance are minor compared with other areas of crisis management. For example, the lender-of-last-resort arrangement in the Monetary Union is essentially based on the subtle notion of ‘‘constructive ambiguity’’. It is inherent to the constructive ambiguity policy that it is neither transparent nor ofﬁcial. Schu¨ler nonetheless calls for increased transparency of the lender-of-last-resort arrangements in Europe, and argues that transparency is compatible with constructive ambiguity. I ﬁnd the argument interesting and novel, but it is not well articulated (see Winkler (2000) for a conceptual framework distinguishing various aspects of transparency).

4. Future work Designing a reform agenda for the European ﬁnancial market regulation and supervision is an overwhelming task for a single scholar. The rather restricted focus of Schu¨ler’s study is therefore understandable. However, as Schu¨ler takes the systemic risk rationale as the point of departure in the argumentation, the absence of discussions of competition and bank closure policies is surprising. Vives (2001) and Carletti and Hartmann (2002) suggest that there is a need to clarify the relationship between competition and supervisory authorities in the European banking market. The optimal Europe-wide bank closure policy is an even more tedious issue. Prominent failures of ﬁnancial institutions, such as Herstatt, Drexell Burnham Lambert, BCCI, Barings, and LTCM, have shown that there is ambiguity about the allocation of business units and assets of a failed bank to legal entities and regulatory authorities even within a country. In the EU there are also different national insolvency regimes that have conﬂicts of interest in case of failures. Moreover, given the liquidity of the modern interbank markets, the traditional lender-of-last-resort doctrine has become moot (Freixas et al., 2003). This further suggests that instead of the lender-of-last-resort arrangement, a need for transparent and credible European bank closure procedure should be assessed. If, instead of systemic risk, the representation hypothesis of bank regulation is taken as the starting point, competition and bank closure policies are perhaps less relevant, save deposit insurance schemes. But more attention to the reform of regulation and supervision of ﬁnancial conglomerates and insurance sector should be directed. The eventual reforms of the pension systems in Europe will most likely involve at least partial privatization and deregulation of pension insurance schemes. Privatization and deregulation of the banking sector in the 1980s and the telecommunications sector in the 1990s displayed holes in the regulatory frameworks that destabilized economies worldwide. If ﬁnancial conglomerate and insurance sector regulation and supervision are not updated before the reforms of European pension systems, a new boom-to-bust is likely.

228

Tuomas Takalo

References Allen, F. and R. Herring, 2001, ‘‘Banking market versus securities market regulation’’, Wharton Financial Institutions Center Working Paper, No. 01–29. Beale, L., A. Ferrando, P. Ho¨rdahl, E. Krylova, and C. Monnet, 2004, ‘‘Measuring ﬁnancial integration in the euro area’’, ECB Occasional Paper, No. 14/ April 2004. Carletti, E. and P. Hartmann, 2002, ‘‘Competition and stability: what’s special about banking?’’ ECB Working Paper, No. 146. De Bandt, O. and P. Hartmann (2001), ‘‘Systemic risk: a survey’’, pp. 249–298 in: C. Goodhart and G. Illig, editors, Financial Crisis, Contagion and the Lender of Last Resort. A Book of Readings, London: Oxford University Press. Degryse, H. and S. Ongena, 2003, ‘‘The impact of technology and regulation on the geographical scope of banking’’, Working Paper, Tilburg University. De Nicolo, G. and M.L. Kwast (2002), ‘‘Systemic risk and ﬁnancial consolidation. Are they related?’’, Journal of Banking & Finance, Vol. 26, pp. 861–880. Dell’Ariccia, G. and R. Marquez, 2003, ‘‘Competition among regulators and credit market integration’’, Working Paper, International Monetary Fund. Dewatripont, M. and J. Tirole (1994), The Prudential Regulation of Banks, Cambridge, MA: The MIT Press. Freixas, X., B.M. Parigi, and J.C. Rochet, 2003, ‘‘The lender of last resort: a 21st century approach’’, ECB Working Paper, No. 298. Gropp, R. and G. Moerman (2004), ‘‘Measurement of contagion in banks’ equity prices’’, Journal of International Money and Finance, Vol. 23, pp. 405–459. Vives, X. (2001), ‘‘Restructuring ﬁnancial regulation in the European Monetary Union’’, Journal of Financial Services Research, Vol. 19, pp. 57–82. Winkler, B. 2000, ‘‘Which kind of transparency? On the need for clarity in monetary policy-making’’, ECB Working Paper, No. 26.

Part IV: One Money, One Economy?

This page intentionally left blank

CHAPTER 8

Reasons and Implications of Inflation Differentials Within the European Monetary Union Ignazio Angeloni and Michael Ehrmann

1. Introduction A central motivation for introducing a single currency in Europe was to create an area of monetary stability with low and stable inﬂation in all participating countries, where disruptive ﬂuctuations of relative prices and inﬂation rates, such as those experienced in the 1970s and 1980s, would be avoided. To achieve this, many euro candidates embarked on wide-ranging monetary and ﬁscal stabilisation programs in the 1990s, with the aim of bringing inﬂation in line with that of Germany, the best performer in this respect. This effort was remarkably successful: when the Economic and Monetary Union (EMU) started, in 1999, the mean absolute deviation of national inﬂation rates in the euro area was 0.8 percent, the lowest level in post-war Europe and remarkably low by any standard among 11 nations with largely independent economic policies and distinct (though partly integrated) product and labour markets. This remarkable success in reaching inﬂation convergence in a short time perhaps concealed, to the eyes of some, that the conditions experienced in 1999 were rather exceptional, and that the re-emergence of signiﬁcant inﬂation differentials in the euro area was entirely possible even after the euro was in place. In fact, euro area inﬂation rates started to diverge again shortly thereafter, much to the surprise and dismay of many observers and policymakers. As the international experience demonstrates, however, the divergence of local and sectoral inﬂation and price levels is an integral part of a well-functioning adjustment mechanism of any economy. The euro area – a single economy from a monetary perspective but, in other respects, a combination of still partly separate entities – can only be an illustration of this, perhaps a rather extreme one. The central CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79008-3

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

232

Ignazio Angeloni and Michael Ehrmann

questions one needs to ask in this context are the following: How smoothly, and how fast, does the international adjustment process operate within the euro area now? Without exchange rate changes between the economies of the euro area, will the interplay of aggregate demand and supply at the national level, coupled with wage and price ﬂexibility, be such to avoid those wide and disruptive gyrations of external competitiveness that the EMU was designed to avoid in the ﬁrst place? And if the answer to this last question is a partial no, what policies are needed to smoothen and enhance the adjustment mechanism? In a sense, asking these questions is like asking again, ex post, whether the euro area does indeed constitute an ‘‘optimum currency area’’.1 In this chapter we approach these questions from a new angle and offer some tentative answers. Our approach is to use a stylised aggregate supply and demand model of the 12 countries currently constituting the euro area, embodying the main mechanisms that have a bearing on the functioning of the euro area internal adjustment process. We draw on recent research at the European Central Bank (ECB) (mainly ECB, 2003; Angeloni and Ehrmann, 2004 – henceforth AE (2004); and Altissimo et al., 2005), as well as a number of other papers that have recently looked at the issue of inﬂation differentials in the EMU. Our plan is the following: ﬁrst, in Section 2, we review the main stylised facts concerning post-EMU inﬂation differentials2 and discuss them in the light of the main alternative explanations offered in the literature. Our discussion in this section is qualitative, but it also helps to collect elements that are important for more formal modelling. Second, in Section 3, we brieﬂy outline the structure of a model, drawn from AE (2004), of national inﬂation and business cycle differentials in the EMU. Third, in Sections 4 and 5, we put the spotlight on two speciﬁc factors having a bearing on inﬂation differentials. The ﬁrst is inﬂation persistence at the national level; the reasoning behind is that, if national inﬂation processes are highly persistent, then even small economic shocks occurring nationally will bring inﬂation rates to diverge markedly and for long periods of time. The second factor is the single monetary policy; the idea here is that, if national monetary policy transmission is uneven, as argued by several authors (e.g., Cecchetti, 1999), then a single area-wide ECB policy can be a source of national differentials. Then, perhaps, different ways to conduct such policy (different central bank ‘‘policy rules’’) can mitigate or exacerbate the problem. The main message from these analyses, summarised in the concluding Section 6, is that for plausible model speciﬁcations the impact of persistence in the national

1 This was a major focus of research in the 1990s. See Angeloni and Dedola (1999) for a discussion and a list of the main references. 2 Strictly speaking, the term EMU refers to the whole process of preparation and implementation of the European Economic and Monetary Union. In practice many authors refer to the EMU as the culminating act of such process, the introduction of the euro in 1999. We follow this simpler verbal convention here.

Reasons and Implications of Inflation Differentials Within the European Monetary Union

Figure 1.

233

Inflation dispersion in the euro area and the US, 1990–2003

7

6

5

4

3

2

1

Jul-03

Jul-02

Jan-03

Jul-01

Jan-02

Jul-00

Jan-01

Jul-99

Jan-00

Jul-98

Jan-99

Jul-97

US (14 MSAs)

Jan-98

Jul-96

Jan-97

Jul-95

Jan-96

Jul-94

Euro Area

Jan-95

Jul-93

Jan-94

Jul-92

Jan-93

Jul-91

Jan-92

Jul-90

Jan-91

Jan-90

0

US (4 Census Regions)

Note: The chart shows the unweighted standard deviation of national inﬂation rates in the euro area, of inﬂation rates in 14 metropolitan statistical areas, and of 4 census regions in the US.

inﬂation processes is potentially very important. Here, we do not provide substantive new evidence in favour or against the existence of inﬂation persistence in the euro area. What we say is simply that, in the range of degree of persistence normally found in most recent analyses, its potential consequence for inﬂation differentials is large, probably more than that coming from any other factors we consider. On the contrary, the impact of monetary policy, even postulating somewhat extreme forms of asymmetry in monetary policy transmission across countries, does not seem very important. If anything, the ECB can contribute to contain inﬂation differentials by tightly controlling aggregate inﬂation around its target. 2. Stylised evidence and interpretations Figures 1 and 2 document the convergence of inﬂation that occurred in Europe in the 1990s, the subsequent modest divergence, and its stabilisation at a level of dispersion slightly above its minimum in late 1999, comparing these developments with those in other economies.3 Inﬂation differentials declined steadily in

3

The sources of all data used in this chapter are described in AE (2004).

Ignazio Angeloni and Michael Ehrmann

234

Figure 2.

Inflation dispersion in the euro area and within three euro area countries, 1990–2003

7

6

5

4

3

2

1

Spain (18 Aut. Communities)

Germany (12 States)

Jul-03

Jul-02

Jan-03

Jul-01

Jan-02

Jul-00

Jan-01

Jul-99

Jan-00

Jul-98

Jan-99

Jul-97

Jan-98

Jul-96

Jan-97

Jul-95

Jan-96

Jul-94

Jan-95

Jul-93

Jan-94

Jul-92

Euro Area

Jan-93

Jul-91

Jan-92

Jul-90

Jan-91

Jan-90

0

Italy (12 Cities)

Note: The chart shows the unweighted standard deviation of national inﬂation rates in the euro area and of regional inﬂation rates in Spain, Germany and Italy.

1990–1999 among prospective euro area countries. The unweighted standard deviation of national inﬂation rates reached a minimum of below 1 percent in 1999, after which it inverted its trend and started to edge up again slightly; since 2000, it has ﬂuctuated around 1 percent. The vertical line in Figures 1 and 2 denotes the start of the EMU, after which the euro area countries shared a common monetary policy. To put into perspective the level of dispersion that has been observed since, it is useful to compare it to the one observed in other economies with a common monetary policy. In the US, the inﬂation divergence among the 14 US ‘‘metropolitan areas’’ (Bureau of Labor Statistics (BLS) data) is remarkably stable around 1 percent. The analogy between European Union (EU) countries and US cities is probably misleading, however: US cities are much smaller than EU nations, and their price indices tend to be more volatile. A probably more appropriate comparison is that with the US Census regions (large aggregations including many states; the BLS publishes data for four macro-regions). According to this measure, inﬂation discrepancy in the US is much lower, typically around 0.5 or less. The proper term of comparison for the euro area countries is probably something between US cities and US macro regions, possibly closer to the latter. An alternative comparison is that with European regional data; as shown in Figure 2, inﬂation divergence among regions within individual European states tends to be smaller than that observed among euro area nations, though regions are obviously smaller and therefore the

Reasons and Implications of Inflation Differentials Within the European Monetary Union

235

discrepancy could be expected to be more volatile. This conﬁrms the evidence of Beck and Weber (2001), showing that the volatility of relative price indices is strongly affected by borders and currency arrangements. Another distinctive feature of euro area inﬂation differentials is their persistence. Table 1 shows annual deviations of national inﬂation rates from the euro area average, as well as those of the four US macro-regions from the US average. Remarkably, of 12 countries now constituting the euro area, eight have remained on the same side relative to the area average in each year since 1999. In six of these eight cases, these countries had also remained on the same side of the mean in all the preceding six years. Changes in the position of US regions are much more frequent: each region has changed side within the last few years. A simple counting exercise can facilitate the comparison. For each, the euro area and the US, Table 1 contains 12 cases where countries move from one side of the mean to the other. Given that there are 12 countries in the euro area, and four US regions, such changes are three times more frequent in the US than in the euro area. The fact that the inﬂation differentials in the euro area are somewhat larger and more persistent than in other currency unions, at least for the time being, does not seem particularly surprising. Product and labour markets are still only partially integrated. Integration is proceeding, but it is not complete yet, and until then the euro area countries should remain more prone to different price and output developments than are, for example, US States and regions. Moreover, many of the factors giving rise to differences in market structures depend on policies largely under national control (budgetary and tax policies; Table 1. Inflation differentials with respect to the euro area/US average

Euro area Austria Belgium Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain United States North East Midwest South West

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

0.2 0.9 0.1 1.2 0.2 10.0 1.6 1.1 3.4 1.8 2.5 1.5

0.1 0.4 1.2 1.1 0.6 7.4 0.1 1.4 2.8 0.7 2.1 1.8

0.9 1.3 2.1 0.8 1.3 6.4 0.3 2.8 2.6 1.2 1.4 2.0

0.5 0.5 1.2 0.2 1.1 5.6 0.1 1.7 1.1 0.9 0.6 1.3

0.5 0.2 0.4 0.4 0.1 3.8 0.4 0.2 0.3 0.2 0.2 0.2

0.3 0.3 0.2 0.5 0.6 3.4 1.0 0.8 0.2 0.6 1.1 0.6

0.6 0.0 0.2 0.6 0.5 1.0 1.3 0.5 0.1 0.9 1.0 1.1

0.2 0.6 0.8 0.3 0.7 0.8 3.1 0.5 1.7 0.2 0.7 1.4

0.1 0.1 0.3 0.6 0.5 1.3 1.6 0.0 0.0 2.8 2.0 0.5

0.6 0.7 0.3 0.3 0.9 1.6 2.5 0.3 0.2 1.6 1.4 1.3

0.8 0.6 0.8 0.1 1.0 1.4 1.9 0.7 0.5 0.2 1.2 1.0

0.2 0.1 0.2 0.0

0.2 0.3 0.2 0.3

0.2 0.3 0.2 0.2

0.2 0.1 0.1 0.3

0.1 0.1 0.2 0.1

0.2 0.1 0.3 0.3

0.1 0.1 0.2 0.5

0.0 0.0 0.2 0.1

0.0 0.1 0.5 0.9

0.5 0.4 0.3 0.3

0.5 0.4 0.0 0.2

236

Ignazio Angeloni and Michael Ehrmann

competition policies in both the wholesale and the retail sectors; labour market regulations; and so on). There is no reason to believe that these policies should systematically be conducted so as to smooth price or output differentials. Since countries have entered the EMU with signiﬁcantly different initial price levels, a high persistence of the inﬂation differentials in the initial years of the monetary union could derive from the fact that countries with below-average price levels experience systematically higher inﬂation until the level has converged. In fact, for several countries there is a correspondence between the initial price gap (relative to the area average) and the inﬂation gap in the initial years of the EMU, as we will show below.4 However, there are cases (e.g., the Netherlands, Ireland, Finland) where convergence does not explain the persistence of inﬂation differences, conﬁrming that the persistent differentials seen in Table 1 cannot be explained by convergence effects alone. Essentially, three types of explanations of these differentials have been put forth in recent literature. The first relates to the possibility of systematic deviations from the law-of-one-price across nations, owing to the existence of segmentations across national markets. If this is part of the explanation, one would expect inﬂation and price level differentials to be more signiﬁcant in these sectors where international arbitrage is less strong, hence particularly in the less traded segments of the consumer price index (CPI). Nonetheless, one could expect differences in other compartments, normally considered subject to international competition, if they incorporate some local non-traded input.5 Moreover, one would expect differentials to be related to domestic aggregate demand and supply developments. The second explanation is linked to the ‘‘BalassaSamuelson’’ theory, according to which inﬂation differentials are the reﬂection of a catch-up process in productivity levels, resulting in inﬂation differentials in non-traded domestically produced goods and hence in the national CPI as a whole.6 Another explanation of similar nature is that inﬂation rates may diverge because of initial differences in price levels. The introduction of a single currency

4

See Figure 4, which uses comparative price levels computed by the Organisation for Economic Cooperation and Development (OECD) based on purchasing power parity. The calculation of these price levels is explained in the OECD Main Economic Indicators (e.g., http://www.oecd.org/dataoecd/48/18/18598721.pdf). 5 The effects of less than fully integrated markets are studied in Andres et al. (2003), who, building on work by Bergin (2003), developed a two-country model where each country produces differentiated goods traded in monopolistically competitive markets. Price discrimination occurs due to differentiated demand conditions and price adjustment costs: the key parameters in the model are the elasticity of demand and the slope of the Phillips curve, both of which are country-speciﬁc. The authors suggest that in their currency area inﬂation differences depend more on the characteristics of local demand than on price inertia. Corsetti and Dedola (2002) derive a model in which international price discrimination is optimal for monopolistic ﬁrms and feasible because distributing traded goods to consumers requires non-tradables. 6 See Sinn and Reutter (2001), Alberola (2000) and Ortega (2003).

Reasons and Implications of Inflation Differentials Within the European Monetary Union

237

may even increase the inﬂation differentials in this case, by accelerating the catch-up process and by preventing price-level convergence from happening hrough changes in the nominal exchange rate. Finally, a third explanation, advanced recently by Honohan and Lane (2003, 2004), is that inﬂation differentials in the euro area may depend on external shocks coming from international prices or the exchange rate. The argument is that, if euro area countries are characterised by different exposure to such shocks – say, because of different trade shares with the rest of the world – then ﬂuctuations in the exchange rate of the euro can produce temporary, but sizeable, differences in national inﬂation rates. Figure 3 shows that there is a correlation between the inﬂation rate and the growth rate of output across countries, on average, in 1999–2003. The result survives even after the exclusion of Ireland, which is an outlier in the chart, though the signiﬁcance of the slope coefﬁcient obviously falls. A natural interpretation is that aggregate demand developments at the national level are one (but not the only) factor behind inﬂation performance, something that lends support to the ﬁrst explanation. Moreover, as documented extensively in ECB (2003), different inﬂation rates at national CPI level tend to correlate with other price and cost growth measures, like for example in gross domestic product (GDP) deﬂators or unit labour costs. This suggests that inﬂation differences are well rooted in the whole chain of costs and prices at the national level. To further verify the existence of deviations from the law-of-one-price, Table 2 reports the decomposition of CPI inﬂation differentials into two Figure 3.

GDP growth and inflation, 1999–2003

4.5 y = 0.3325x + 1.645 R2 = 0.529

HICP growth average

4.0

IE

3.5 ES

PT

GR

NL

3.0

IT

2.5

LU FI

2.0 FR

BE AT

1.5 DE 1.0 1.0

2.0

3.0

4.0 5.0 6.0 Real GDP growth average

7.0

8.0

Note: The chart shows average GDP growth and average HICP inﬂation over the ﬁve-year period.

238

Table 2. Decomposing inflation differentials: index composition versus sectoral differentials Country

Average deviation Composition effect

Effect from different sectoral inﬂation rates

Residual

Unprocessed food Processed food Energy Other industry Services

Unweighted standard deviation

0.85 0.59 0.59 0.69 1.12 0.73 1.40 0.05 0.02 0.48 0.73 0.57

0.01 0.01 0.05 0.01 0.00 0.03 0.15 0.09 0.04 0.03 0.04 0.06

0.99

0.12

0.05 0.04 0.08 0.01 0.11 0.12 0.03 0.06 0.00 0.00 0.03 0.19

0.13 0.12 0.25 0.08 0.14 0.20 0.36 0.06 0.14 0.00 0.02 0.04

0.09 0.05 0.03 0.12 0.15 0.03 0.02 0.06 0.01 0.25 0.05 0.04 0.94

0.20 0.02 0.16 0.19 0.30 0.23 0.18 0.20 0.01 0.11 0.30 0.21

0.38 0.34 0.11 0.48 0.72 0.24 1.01 0.10 0.12 0.15 0.47 0.23

0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.06

Note: This table shows the average deviation of national inﬂation rates from the unweighted country average in the euro area for the period January 1999 to July 2004, based on monthly y-o-y rates, as well as their decomposition into composition effects owing to the different composition of the national basket relative to the area and the discrepancy effects, given by the deviation of a country’s sectoral inﬂation relative to the area’s sectoral inﬂation. This decomposition is derived as: Dk ¼ pk Sk pk =K ¼ Sj wj;k ðpj þ dj;k Þ Sj wj pj ¼ Sj ðwj;k wj Þpj þ Sj wj;k dj;k , where dj,k is the deviation of country k’s sectoral inﬂation in sector j from the area-wide inﬂation in sector j and wj,k denotes the weight of category j in the basket of country k. The ﬁrst term, Sj ðwj;k wj Þpj , measures the composition effect (CE), the second, Sj wj;k dj;k , the discrepancy effect (DE). The last row shows the average unweighted standard deviation of qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ ﬃ ﬃ qﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ national HICP inﬂation rates over the same sample and its decomposition into s ¼ Sk D2k K ¼ ðSk CE 2k þ Sk DE 2k þ Sk Residualk Þ K .

Ignazio Angeloni and Michael Ehrmann

Austria Belgium Finland France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain

Reasons and Implications of Inflation Differentials Within the European Monetary Union

239

components: a ‘‘composition effect’’ and a ‘‘discrepancy effect’’ (the methodology for calculating the decomposition is explained in the note to the table). The rationale behind this is that differences in aggregate inﬂation across countries may depend on two factors. One is given by the fact that the composition of national indices differs, hence divergent national price dynamics can arise simply because sectoral price dynamics differ (‘‘composition’’). The other is given by deviations from the law-of-one-price in the individual compartments. The calculation in the table is done illustratively for each country and the euro area as a whole, and for ﬁve sectors of the harmonised index of consumer prices (HICP): processed and unprocessed food; energy; other industry and services. The table shows three things. First, the inﬂation differences depend mainly on deviations from the law-of-one-price within sectors, and only to a little extent on ‘‘composition effects’’. Second, the differentials tend to have the same sign across sub-indices; this is consistent with the predominance of factors of national origin, as opposed to sector-speciﬁc origin. Third, we observe larger discrepancies in the service sector, which tends to be less internationally traded. Again, this accords well with the idea of nationally originated differentials that extend across all sectors but relatively more in those less exposed to external competition. But other evidence suggests that the other two explanations are relevant as well. Figure 4 plots the residuals from the interpolating line shown in Figure 3 against estimates of the deviations of consumer price levels from the law of one price, published by OECD (2003). There is a clear negative relation, as one Figure 4. 1.0 0.8

PT

Deviations from law of one price in 1999 and inflation in the euro area countries y = -0.0332x - 0.0578 R2 = 0.6845

HICP spread

NL

ES

0.6

IT

0.4

GR

0.2 IE

0.0 -0.2

AT

-0.6 -0.8 -30

FR

BE

-0.4

LU

FI

DE

-25

-20

-15 -10 -5 0 5 1999 Comparative price deviations

10

15

Note: The chart plots the unexplained part of average inﬂation in Figure 3 against the deviations of consumer price levels from the law of one price in 1999.

Ignazio Angeloni and Michael Ehrmann

240

Figure 5.

Exchange rate developments and inflation in the euro area, 1999–2003

4.0

25

3.5

20 15

3.0

10

2.5

5

2.0 1.5

0 -5

Mean Median US$ lagged

1.0 0.5

-10 -15 -20

19 99 Q 1 19 99 Q 3 20 00 Q 1 20 00 Q 3 20 01 Q 1 20 01 Q 3 20 02 Q 1 20 02 Q 3 20 03 Q 1 20 03 Q 3

0.0

1.6

25

1.4

20 15

1.2

10

1.0

5

0.8

0 -5

0.6

St Dev

-10

0.4

US$ lagged

-15

03

Q

1 Q 03

20

3 20

02

Q

1 Q 02

20

3 20

Q 01 20

01

Q

3 20

00

Q

1 20

00

Q

3 20

99

Q

1 19

Q 99 19

3

-20 1

0.2

Note: The chart shows the mean, median and standard deviation of national HICP inﬂation rates (left axis) and the annual rate of the exchange rate change of the euro against the US dollar lagged 2 quarters (right axis). The chart updates comparable charts in Honohan and Lane (2003).

would expect in the presence of a catch-up process. This conﬁrms that the dynamics of inﬂation in the initial years of the EMU were also driven by a convergence mechanism. The third explanation emphasises, as we have mentioned, the role of import prices. There are two dimensions to be considered here. First, since countries differ in their openness to international trade, any given change in import prices (expressed in domestic currency, hence including exchange rate changes) impacts differently on national inﬂation rates. Moreover, the composition of imports differs, potentially giving rise to further asymmetries. It turns out that euro area countries’ inﬂation rates are indeed correlated with their degree of external openness and with the changes in import prices (see ECB, 2003). This effect is also found in Honohan and Lane (2003, 2004), who use the substantial swings in the exchange rate of the euro to explain the national inﬂation differentials, based

Reasons and Implications of Inflation Differentials Within the European Monetary Union

241

on the idea that the euro area economies have different exposures to exchange rate shocks. They ﬁnd signiﬁcant explanatory power of the exchange rate both for the early months of the EMU, when the euro depreciated substantially (Honohan and Lane, 2003), as well as later, when the euro recovered (Honohan and Lane, 2004). Figure 5, replicated and updated from Honohan and Lane’s ﬁrst paper, provides visual support for this idea for the ﬁrst years of the EMU. The exchange rate appears to correlate not only with inﬂation in the euro area (upper panel), but also with its dispersion (lower panel). Whereas the former is easily understood, the latter is more surprising. It appears that the large depreciation of the euro in the initial years of EMU led to an increasing inﬂation dispersion, where more open economies like, e.g., Ireland, experienced aboveaverage inﬂation, whereas the subsequent appreciation of the euro implied a correction, which helped bring down inﬂation differentials. However, it appears that the co-movement of the series has changed considerably in recent years. To explore this hypothesis further, we have updated Honohan and Lane’s (2003) panel estimates including another year of data (plus data revisions intervened in the meantime), and also checked the sensitivity of their results to the exclusion of Ireland (an outlier, with very high inﬂation and growth in the sample period). The results, in Table 3, conﬁrm that an exchange rate effect exists, but its statistical signiﬁcance is rather weak, particularly if one extends the sample. Conversely, the signiﬁcance of the output gap and of lagged prices (seen as a proxy for initial deviations from purchasing power parity (PPP)) is strengthened if one adds more recent data. The fair conclusion from this review of stylised evidence is that all three interpretations receive some support from the data, and hence should ideally be part of a comprehensive model of inﬂation differentials in the EMU. A model that attempts to combine all three interpretations is described in the next section. 3. A model of inflation differentials in EMU AE (2004) describe a simple model of the 12 euro area countries and use it to examine the dynamic properties of the euro area internal adjustment process and to study what factors may be behind the observed patterns of inﬂation differentials. In this section we brieﬂy describe the model: readers interested in more detail may refer to AE (2004). Each national economy is represented by two simple aggregate supply and demand equations. The aggregate supply equation links consumer price inﬂation with past and expected future inﬂation and the domestic output gap. This ‘‘hybrid’’ Phillips curve is augmented by a ‘‘pass-through’’ mechanism from the nominal effective exchange rate to inﬂation, thereby introducing in the model a mechanism akin to that discussed by Honohan and Lane (2003). The nominal effective exchange rate for each country is derived from the changes of the euro exchange rates with non-euro area currencies, weighted by the country-speciﬁc

242

Table 3.

Output gap Fiscal stance Lagged price level

Honohan/Lane

Replacing inﬂation

Replacing output gap

Dropping Ireland

Extending to 2002

0.28** 2.71 0.22** 2.65 0.02 0.32 0.03** 2.88

0.17 1.68 0.27*** 3.19 0.05 0.59 0.04*** 3.36

0.18 1.43 0.26*** 3.06 0.06 0.91 0.03*** 3.06

0.38 1.54 0.19 1.18 0.01 0.06 0.03*** 2.84

0.02 0.31 0.41*** 5.95 0.11 1.68 0.04*** 4.03

Note: The dependent variables are the differentials of national HICP inﬂation rates from the euro area average, annual data. Estimates are obtained in pooled Ordinary Least Squares (OLS) regressions with time-ﬁxed effects. Numbers in italics are t-statistics based on White-corrected standard errors. *, **, *** denote signiﬁcance at the 90%, 95% and 99% level, respectively. Column (1) replicates the results in Honohan and Lane (2003), using their original data; sample period: 1999–2001. Columns (2) and (3) replace HICP inﬂation or the output gap with a later vintage of data, maintaining the same sample period as in Honohan and Lane (2003). Column (4) uses the original dataset by Honohan and Lane (2003), but drops Ireland from the sample. Column (5) uses a later vintage of all data, extending the sample to 1999–2002 (including Ireland). Sources: Effective exchange rate: IMF International Financial Statistics; Output gap and Fiscal stance: OECD Economic Outlook; Lagged price level: OECD Main Economic Indicators.

Ignazio Angeloni and Michael Ehrmann

Lagged change in effective exchange rate

Replicating Honohan and Lane (2003)

Reasons and Implications of Inflation Differentials Within the European Monetary Union

243

trade shares. Aggregate demand consists in an output gap equation, which is a function of past domestic output gaps, a short-term real interest rate, and the real effective exchange rate, which measures the country’s external price competitiveness. Each country’s output gap equation contains the domestic real interest and exchange rates: the common nominal interest rate is converted to a real rate by means of the domestic inﬂation rate, and real effective exchange rates are constructed from the relevant bilateral intra- and extra-area exchange rates, weighted by the appropriate trade shares. The precise formulation of the model, its lag structure, and the choice of country-speciﬁc parameters follow the restrictions obtained from the estimation:7 pt;j ¼ c1 þ a1 pt1;j þ ð1 a1 ÞE t ptþ1;j þ bgapt2;j þ gDneert;j þ 1t;j

ð1Þ

gapt;j ¼ c2 þ a2 gapt1;j þ dj ðrt1 pt1;j Þ þ lj reert1;j þ 2t;j

ð2Þ

The index j represents a generic country; neer and reer denote the nominal and the real effective exchange rate of each country; all other symbols are self-explanatory. This model contains various factors that are relevant for the evolution of inﬂation differentials among countries. First, in the absence of fully integrated product markets, we assume that the relevant inﬂation rate that agents use to calculate the real short-term interest rate is given by the domestic rate. This means that an above-average inﬂation in country j automatically results in a below-average real interest rate, since money market rates are equalised under a common currency. This effect tends to amplify, from the demand side, the original inﬂation differentials. Second, a country with above-average inﬂation experiences over time a cumulative appreciation of the real exchange rate. This appreciation of the real exchange rate reduces aggregate demand for homeproduced goods, with a resulting deﬂationary effect at home that eventually re-equilibrates the inﬂation differentials. Third, if the initial price levels (and thus the real exchange rates) are not in steady-state equilibrium, a convergence process takes place, during which national inﬂation rates temporarily differ – a catch-up mechanism of the type described in the preceding section. Finally, inﬂation and output persistence (that is, their dependence on past values) constitute a propagating effect, through which cross-country differences are carried partially through time. AE (2004) report estimates of the model obtained using panel instrumental variables on quarterly data for the EMU period. The scarcity of data for this recent period severely limits the degree of parameter differentiation across countries we can allow for. As an initial benchmark, we assume full symmetry across countries except for the degree of external openness (reﬂected in the value

7

For example, we have tested for the presence of a forward-looking component in the aggregate demand equation, which was, however, not statistically signiﬁcant.

244

Ignazio Angeloni and Michael Ehrmann

of l) and the country-speciﬁc trade weights that enter in the construction of reert,j. We then proceed experimenting with other types of cross-country differentiation, using a mixture of calibration and estimation. Simulations of the model in AE (2004) show that it generates economically signiﬁcant inﬂation and output differentials, but that the economy moves back to an equilibrium without inﬂation and output differentials for the benchmark parameters.8 4. The role of inflation persistence In this section we use the model to analyse the effect of changes in one of the key model parameters – that expressing the degree of inﬂation persistence (a1). We do so by subjecting the model to a variety of shocks and examining how large and long-lived the inﬂation differentials across countries are for different values of the inﬂation persistence parameter within plausible ranges. Since we plan to examine the effect of monetary policy separately in the next section, the model we use here is closed with a very simple monetary policy rule, one where the interest rate reacts only to expected inﬂation as follows: rt ¼ 1:5E t ptþ1 . We also assume that the exchange rate of the euro remains constant compared with non-euro area currencies. We consider four types of shocks: demand or supply shocks concentrated in one country (namely, Germany) alone, as well as common demand or supply shocks hitting all countries simultaneously by the same amount. We summarize the results using two parameters: the ‘‘maximum dispersion’’ (peak value of the impulse response of the unweighted standard deviation of euro area inﬂation differentials following the shock) and the ‘‘half life’’ (number of quarters it takes for the cumulated response to reach one half of its ﬁnal value). Concerning the persistence parameter, a1, we perform three illustrative experiments. First, we keep it uniform across countries and let it vary around its point estimate (0.46). Note that the standard error of this estimate reported by AE (2004) is 0.21; hence, for example, a 90 or 95 percent conﬁdence interval would contain a large range of parameter values including most of those reported in recent literature.9 Second, we introduce differentiation across countries in the value of a1, by distinguishing two groups, high persistence (ah) and low

8

In a stochastic simulation of the model, using the estimated residual covariance, we were able replicate the historical mean size of inﬂation differentials when setting the parameter of inﬂation persistence, a1, to 0.3, a value not statistically different from the point estimate of 0.46, and well within the range of estimates reported in the literature. 9 Smets and Wouters (2003) obtain 0.31 in their area-wide stochastic dynamic general equilibrium (SDGE) model estimated with 1980–1999 quarterly data with Bayesian techniques, whereas Smets (2003) obtains 0.48 in a simple two-equation model estimated with annual 1980–1999 data using generalised method of moments (GMM). Galı´ et al. (2001) obtain a lower level of persistence: between 0.04 and 0.27. Benigno and Lo´pez-Salido (2002) report estimates of a1 that differ across European countries; they rank Germany at the lower end with 0.04, and classify Italy as the most persistent country with an estimate of 0.55.

Reasons and Implications of Inflation Differentials Within the European Monetary Union

245

persistence (al), respectively. Tentatively, we included France, Greece, Ireland, Italy and Portugal in the ‘‘high persistence’’ group, following Nicoletti et al. (1999), who identify these countries as those with stricter product market regulation. Third and ﬁnal, we consider the case in which only one country’s (namely, France’s) inﬂation persistence varies, while in all others the parameter remains constant at the estimated value (0.46). The results of these experiments, reported in the three panels of Table 4, are remarkable as they indicate that even small increases in persistence from the benchmark value, well within the conﬁdence range, can increase inﬂation divergence sharply and make the system dynamically unstable. The threshold value of a1 where unstable solutions arise is just below 0.5, close to the benchmark value of 0.46. This happens both when all a1 are equal, and when they are differentiated in two groups, or even when it is only one country’s persistence that increases. 5. Effect of changes in the monetary policy rule So far we have assumed the simplest possible form for our monetary policy rule rt ¼ 1:5E t ptþ1 , i.e., monetary policy reacts to expected inﬂation one period ahead. We now consider how inﬂation differentials react to different and more realistic speciﬁcations for the monetary policy rule. We consider extensions of our basic rule that take the form rt ¼ rrt1 þ ð1 rÞ½jp E t ptþk þ jy gapt , where r measures the inertia of the nominal short-term interest rate, jy measures the interest rate response to the area-wide output gap, and k measures the extent to which the policy rule is forward-looking in inﬂation. The simple rule considered earlier is a special case, with r ¼ 0, jp ¼ 1.5, k ¼ 1 and jy ¼ 0. Table 5 shows the results of changing r and jy while leaving k ¼ 1 and jp constant at 1.5. Note that the results are rather insensitive even to signiﬁcant changes in these parameters. As inertia increases, the size of the differentials decreases slightly; the duration increases slightly under idiosyncratic shocks and decreases slightly under common shocks. Inﬂation differentials (size and duration) remain virtually identical for all values of jy. More interesting are the results in Table 6, where the policy horizon is extended up to three years into the future (the two panels in this table should be compared with the benchmark case, reported on top of Table 7). Our analysis here is related to that of Batini et al. (2004), who study the performance of forward-looking rules in open economies with ﬂoating exchange rates; they ﬁnd that such rules generate indeterminacy when the forecast horizon lengthens beyond a certain threshold. We are interested in seeing if this result carries over to our model of the euro area. In the table we see that the benchmark results remain essentially unchanged if k rises from 1 to 6 quarters, but change signiﬁcantly if we move up to 12 quarters. Highly forward-looking rules tend to generate indeterminacy, the more the higher is the value of jp.

246

Table 4.

Common demand shock

Idiosyncratic cost-push shock: DE

Common cost-push shock

Max dispersion

Half-life

Max dispersion

Half-life

Max dispersion

Half-life

Max dispersion

Half-life

0.04 0.07 0.08 0.11 0.17 0.17 0.18 inf

7 7 7 8 16 18 26 inf

0.01 0.01 0.02 0.03 0.07 0.08 0.09 inf

18 15 15 14 17 19 21 inf

0.28 0.34 0.36 0.42 0.48 0.49 0.49 inf

2 4 4 6 14 16 24 inf

0.01 0.03 0.03 0.06 0.15 0.17 0.19 inf

17 14 13 12 16 18 20 inf

0.11 0.10 0.09 0.09 0.08 0.08 0.07 inf

8 8 8 8 8 9 12 inf

0.03 0.06 0.09 0.13 0.16 0.20 0.24 inf

14 12 12 12 13 15 19 inf

0.42 0.40 0.39 0.38 0.37 0.36 0.35 inf

6 5 5 5 5 6 8 inf

0.06 0.14 0.22 0.30 0.39 0.48 0.57 inf

12 10 10 10 11 13 18 inf

a1 0.00 0.20 0.25 0.35 0.45 0.46 0.47 0.50 ah, ala 0.35, 0.38, 0.40, 0.42, 0.44, 0.46, 0.48, 0.50,

0.35 0.32 0.30 0.28 0.26 0.24 0.22 0.20

Ignazio Angeloni and Michael Ehrmann

Idiosyncratic demand shock: DE

Model simulations: changing inflation persistence

ah, alb 0.35 0.32 0.30 0.28 0.26 0.24 0.22 0.19

0.11 0.10 0.09 0.09 0.08 0.08 0.07 0.07

8 8 8 8 8 8 11 227

0.03 0.04 0.06 0.08 0.10 0.12 0.15 0.24

14 12 12 13 14 16 21 242

0.42 0.40 0.39 0.38 0.37 0.36 0.35 0.34

6 5 5 5 5 5 7 222

0.06 0.09 0.13 0.18 0.23 0.28 0.34 0.46

12 10 10 11 12 14 19 241

Note: ‘‘inf’’ denotes explosive solutions, ‘‘ind’’ indeterminate ones. ‘‘Max dispersion’’ refers to the peak response of the unweighted standard deviation of national inﬂation rates following the speciﬁed shock. ‘‘Half-life’’ refers to the quarter in which the cumulated response of the unweighted standard deviation of national inﬂation rates reaches 50% of the total accumulated response. a France, Greece, Ireland, Italy, and Portugal are more persistent. b France is more persistent.

Reasons and Implications of Inflation Differentials Within the European Monetary Union

0.35, 0.38, 0.40, 0.42, 0.44, 0.46, 0.48, 0.51,

247

248

Table 5.

Model simulations: sensitivity to the monetary policy rule (I): changing the inertia of interest rates and the response to the output gap Idiosyncratic demand shock: DE

Common demand shock

Idiosyncratic demand shock: DE

Common cost-push shock

Half-life

Max dispersion

Half-life

Max dispersion

Half-life

Max dispersion

Half-life

0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17

18 18 18 18 18 18 19 19

0.08 0.07 0.07 0.06 0.05 0.04 0.03 0.03

19 18 18 18 18 18 18 18

0.49 0.49 0.49 0.49 0.49 0.49 0.49 0.49

16 16 16 16 17 17 17 17

0.17 0.16 0.16 0.15 0.14 0.13 0.11 0.11

18 17 17 17 17 17 17 17

0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17

18 18 18 18 18 18 18 18

0.08 0.08 0.08 0.08 0.08 0.08 0.07 0.07

19 18 18 18 18 18 18 18

0.49 0.49 0.49 0.49 0.49 0.49 0.49 0.49

16 16 16 16 16 16 16 16

0.17 0.17 0.18 0.18 0.18 0.18 0.19 0.19

18 18 18 17 17 17 17 17

q 0.00 0.30 0.40 0.60 0.70 0.80 0.90 0.95 uy 0.00 0.50 1.00 1.50 2.00 2.50 3.00 4.00

Note: See Table 4.

Ignazio Angeloni and Michael Ehrmann

Max dispersion

Idiosyncratic demand shock: DE

Common demand shock

Idiosyncratic cost-push shock: DE

Common cost-push shock

Max dispersion

Half-life

Max dispersion

Half-life

Max dispersion

Half-life

Max dispersion

Half-life

0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17

18 19 19 19 18 18 18 19

0.14 0.15 0.15 0.14 0.14 0.13 0.11 0.09

28 24 22 20 19 18 18 19

0.49 0.49 0.49 0.49 0.49 0.49 0.49 0.49

16 17 17 17 17 16 17 17

0.23 0.26 0.26 0.27 0.27 0.27 0.25 0.23

27 23 21 19 18 17 17 19

0.17 0.17 inf 0.18 0.18 0.18 ind ind

18 37 inf 64 25 19 ind ind

0.14 0.20 inf 0.28 0.25 0.22 ind ind

28 58 inf 82 33 22 ind ind

0.49 0.49 inf 0.49 0.49 0.49 ind ind

16 31 inf 53 22 17 ind ind

0.23 0.30 inf 0.46 0.43 0.40 ind ind

27 55 inf 79 30 20 ind ind

/p(t+6) 0.00 0.50 1.00 1.50 2.00 3.00 5.00 10.00 /p(t+12) 0.00 0.50 1.00 1.50 2.00 3.00 5.00 10.00

Note: See Table 4.

Reasons and Implications of Inflation Differentials Within the European Monetary Union

Table 6. Model simulations: sensitivity to the monetary policy rule (II): changing the forward-lookingness of the monetary policy rule

249

250

Table 7. Model simulations: sensitivity to the monetary policy rule (III): changing the response to the expected inflation Idiosyncratic demand shock: DE

Common demand shock

Idiosyncratic cost-push shock: DE

Common cost-push shock

Max dispersion

Half-life

Max dispersion

Half-life

Max dispersion

Half-life

Max dispersion

Half-life

0.17 0.17 0.17 0.17 0.17 0.17 0.17 0.17

18 18 18 18 18 18 18 19

0.14 0.12 0.09 0.08 0.08 0.07 0.05 0.03

28 21 20 19 18 18 18 18

0.49 0.49 0.49 0.49 0.49 0.49 0.49 0.49

16 16 16 16 16 16 17 17

0.23 0.21 0.18 0.17 0.17 0.16 0.14 0.11

27 20 18 18 17 17 17 18

inf inf 76 33 22 16 14 12

inf inf 0.23 0.20 0.19 0.17 0.15 0.12

inf inf 119 65 42 27 19 15

inf inf 0.28 0.28 0.28 0.28 0.28 0.28

inf inf 66 25 13 8 6 5

inf inf 0.47 0.47 0.47 0.46 0.45 0.43

inf inf 118 63 40 25 17 13

/p(t+1)

/p(t+1) proportional weightsa 0.00 0.50 1.00 1.50 2.00 3.00 5.00 10.00

inf inf 0.04 0.04 0.04 0.04 0.04 0.05

Note: In all countries but France a1 ¼ 0.0; for France, a1 ¼ 0.5. a See Table 4.

Ignazio Angeloni and Michael Ehrmann

0.00 0.50 1.00 1.50 2.00 3.00 5.00 10.00

Reasons and Implications of Inflation Differentials Within the European Monetary Union

251

As a ﬁnal step, we combine the analysis of this section with that of the preceding one, and study how monetary policy affects inﬂation differentials in the presence of inﬂation persistence. As before, we distinguish the case where all countries have equally persistent inﬂation from the case where persistence is differentiated across countries. Considering that interest rate inertia and response to the output gap seem to matter little in our case, we return to our simple rule rt ¼ jp E t ptþ1 and examine the combination of alternative degrees of persistence and policy ‘‘activism’’ (with jp ranging from zero to large positive numbers). The top panel of Table 7 reports the results if inﬂation persistence (i.e., the parameter a1) is set to our benchmark value of 0.46 in all countries. It turns out that a high degree of monetary activism reduces inﬂation dispersion (both its size and its half-life) under common shocks, while it is not inﬂuential under idiosyncratic shocks. This suggests that there may be complementarity between aggregate inﬂation and inﬂation differences under common shocks, where a stronger reaction of monetary policy to expected inﬂation developments helps reduce both aggregate inﬂation and the cross-country dispersion. This is also the case when persistence is differentiated across countries. The lower panel shows the results for the experiment where a1 is set to zero for all countries but one, for which it is set to 0.5. However, as one would expect, given the high asymmetry introduced in the model, the level of inﬂation dispersion rises dramatically. Note that the system is unstable if monetary policy does not react strongly enough to inﬂation (i.e., jp at 0 or 0.5). In this case, a higher jp both stabilises the model and reduces inﬂation differentials under common shocks. 6. Conclusions In this chapter we have ﬁrst reviewed the development of inﬂation differentials in the euro area since 1999 and examined in parallel some potential explanations for them. Our goal in this ﬁrst part of the paper is to see how the main contending explanations ﬁt the facts, thus providing input for the speciﬁcation of a model. Next, we have introduced a simple multi-country model of the euro area and used it to examine the role played in shaping such differentials by two key elements: the degree of inﬂation persistence existing in the national economies and the type of monetary policy rule conducted by the ECB. In the ﬁrst part of the paper, our main conclusion is that a satisfactory account of inﬂation differentials in the European currency area in the EMU’s early years must include a combination of at least three elements:

the interplay of aggregate demand and aggregate supply at the national level, the convergence of prices from different initial levels, the effect of external shocks (import prices and exchange rate changes).

The model we used in the second part of the paper includes all three elements. We ﬁnd that the degree of inﬂation persistence prevailing in the individual

252

Ignazio Angeloni and Michael Ehrmann

countries can, within plausible parameter ranges, increase the size and duration of inﬂation differentials sharply. It needs emphasising that we do not provide evidence in favour or against the existence of a high degree of inﬂation persistence here, but simply ﬁnd that, if persistence increases within reasonable bounds, inﬂation differentials increase sharply and in the limit become unbounded. We also ﬁnd that a more aggressive monetary policy, aiming at stabilising aggregate area-wide inﬂation, can reduce inﬂation differentials when inﬂation persistence is high. Together, this evidence suggests that the best policy to reduce the size and duration of inﬂation differentials in the euro area is a mix comprising structural measures to reduce the persistence of the inﬂation process at the national level and a monetary policy geared at controlling area-wide inﬂation tightly. Acknowledgement We are grateful to an anonymous referee, seminar participants at the CESifoDelphi Conference on ‘‘Designing the New EU’’, and especially Helge Berger, Ju¨rgen von Hagen, Andrew Hughes Hallett, Thomas Moutos, and Gerard Roland for useful comments. The views expressed in this paper are not necessarily those of the institutions with which the authors are afﬁliated. References Alberola, E. (2000), ‘‘Interpreting inﬂation differentials in the euro area’’, Banco de Espan˜a Economic Bulletin, April, Vol. 61–70. Altissimo, F., P. Benigno and D. Rodriguez Palenzuela (2005), ‘‘Long-run determinants of inﬂation differentials in a Monetary Union’’, CEPR Discussion Paper, No. 5149. Andres, J., E. Ortega and J. Valle´s (2003), ‘‘Market structure and inﬂation differentials in the European Monetary Union’’, Banco de Espan˜a Documento de Trabajo, No. 0301. Angeloni, I. and L. Dedola (1999), ‘‘From the ERM to the euro: new evidence on economic and policy convergence among EU countries’’, ECB Working Paper, No. 4. Angeloni, I. and M. Ehrmann (2004), ‘‘Euro area inﬂation differentials’’, ECB Working Paper, No. 388. Batini, N., P. Levine and J. Pearlman (2004), ‘‘Indeterminacy with inﬂationforecast-based rules in a two-bloc model’’, ECB Working Paper, No. 340. Beck, G.W. and A.A. Weber (2001), ‘‘How wide are European borders? On the integration effects of monetary unions’’, CFS Working Paper, No. 2001/07. Benigno, P. and D. Lo´pez-Salido (2002), ‘‘Inﬂation persistence and optimal monetary policy in the euro area’’, ECB Working Paper, No. 178. Bergin, P. (2003), ‘‘A model of relative national price levels under pricing to market’’, European Economic Review, Vol. 47(3), pp. 569–586.

Reasons and Implications of Inflation Differentials Within the European Monetary Union

253

Cecchetti, S. (1999), ‘‘Legal structure, ﬁnancial structure, and the monetary policy transmission mechanism’’, pp. 170–194 in: Deutsche and Bundesbank, editors, The Monetary Transmission Process. Recent Developments and Lessons for Europe, Houndmills: Palgrave. Corsetti, G. and L. Dedola (2002), ‘‘Macroeconomics of international price discrimination’’, ECB Working Paper, No. 176. European Central Bank (2003), ‘‘Inﬂation differentials in the euro area: potential causes and policy implications’’, www.ecb.int/pub/pdf/other/ inﬂationdifferntialreporten.pdf Galı´ , J., M. Gertler and D. Lo´pez-Salido (2001), ‘‘European inﬂation dynamics’’, European Economic Review, Vol. 45(7), pp. 1237–1270. Honohan, P. and P. Lane (2003), ‘‘Divergent inﬂation rates in EMU’’, Economic Policy, Vol. 18(37), pp. 358–394. Honohan, P. and P. Lane (2004), ‘‘Exchange rates and inﬂation under EMU: an update’’, Economic Policy Web Essay (http://www.economic-policy.org/ responses.asp). Nicoletti, G., S. Scarpetta and O. Boyland (1999), ‘‘Summary indicators of product market regulation with an extension to employment protection legislation’’, OECD Economics Department Working Papers, No. 226. OECD (2003), Main Economic Indicators, December. Ortega, E. (2003), ‘‘Persistent inﬂation differentials in Europe’’, Banco de Espan˜a Documento de Trabajo, No. 0305. Sinn, H.W. and M. Reutter (2001), ‘‘The minimum inﬂation rate for Euroland’’, NBER Working Paper, No. 8085. Smets, F. and R. Wouters (2003), ‘‘An estimated dynamic stochastic general equilibrium model of the euro area’’, Journal of the European Economic Association, Vol. 1(5), pp. 1123–1175. Smets, F. (2003), ‘‘Maintaining price stability: how long is the medium term?’’, Journal of Monetary Economics, Vol. 50(6), pp. 1293–1309.

This page intentionally left blank

Comment Andrew Hughes Hallett

In a discussion of this kind, one can take one of two positions. I could comment on the details of the model – how it is speciﬁed and estimated, and what you can learn from the estimates or certain simulations. That is what the authors have done very well for themselves. Alternatively, I could comment on the general framewok of analysis and provide an interpretation, in terms of the insights and improvements offered, compared with existing literature. This is what I shall do here. To set the scene, the results in this paper can be summarized as follows:

There are persistent inﬂation and inﬂation differentials in the Euro zone These inﬂation differentials have a stable character, meaning, countries mostly remain on one side of the mean of the inﬂation distribution or the other. The distribution itself therefore persists, and real convergence is very slow. The authors demonstrate that inﬂation differentials decreased in 1999, but then increased again thereafter. Inﬂation persistence itself appears to have increased those differentials. That suggests market failures somewhere. The size and persistence of country-speciﬁc shocks have had the same effect, but to a lesser extent. The asymmetric transmission of common shocks and monetary policy is therefore more important than the differences between countries per se. The differences between Phillips curve parameters also play a role – greater than asymmetric shocks, but less than inﬂation persistence. Output persistence has less inﬂuence on the euro zone economies unless it becomes very responsive to monetary conditions or to exchange rate movements.

CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79016-2

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

256

Andrew Hughes Hallett

Monetary policy likewise has had smaller effects, in particular where the responsiveness to output gap varies. However, this result changes rapidly if the horizon of the inﬂation targeting exercise changes. This is a common ﬁnding: increase one of the model’s parameters beyond a certain threshold, and the model will often become unstable, or produce different solutions or no solution at all. But you can usually offset that effect by making the policy rule more responsive to inﬂation or inﬂation differentials (e.g., increasing the weights on the worst performers). That suggests stabilisability is possible, but the outcomes are very sensitive to changes in the underlying parameters – with the possibility that the economy itself could become unstable in certain cases. Are these observations reasonable? Do they ﬁt in with our existing knowledge? 1. The authors document the persistence of inﬂation and its differentials; and also the persistence in that persistence (i.e., that certain countries are always on the upside of the average, and others on the downside). But they do not explain why these things happen. Instead, they show that they could result from a variety of sources: (a) asymmetric monetary transmissions – but do they, in reality?; or (b) from asymmetric shocks (but they would have to be systematically biased to produce the observed ‘‘persistence in persistence’’); or (c) from different rigidities in different places (this must be the case since the degree of persistence in asymmetric shocks is not reported, and it is difﬁcult to see how the similar sized asymmetries can produce the same inﬂation differentials every time) ; or (d) from steeper Phillips curves in some places, so that the more ﬂexible markets must carry the burden of adjusting for the others1; or (e) from a increasing persistence in inﬂation itself, while differentials fall with the elasticity of output to the real exchange rate. The last two points suggest that market rigidities lie at the bottom of these results. This is important because slow real adjustments would cause the European Union (EU) to have more persistent inﬂation or inﬂation differentials,2 and the process of monetary union will presumably have made this worse by making the economies more closed to the outside world. 2. The Hughes Hallett–Piscitelli ‘‘cyclical divergence’’ result3 stands conﬁrmed: cyclical convergence will peak as integration proceeds beyond a certain threshold value. This appears to have happened in 1998–1999. And, if the Hughes Hallett–Piscitelli model is correct, this result will not go away although it may be weakened by asymmetries in the underlying volatility, size, or industrial structure of the economies concerned. All of these features are

1

See Hughes Hallett et al. (2004) for an explicit demonstration of that proposition. There is some evidence that the important transmission asymmetries are mostly on the real side (in wage bargains, asset effects, and income responses; see Hughes Hallett and Piscitelli, 2002a). 3 See Hughes Hallett and Piscitelli (2002b). 2

Comment

257

present here, given that whatever happens to the output cycle will also be reﬂected in the inﬂation cycle. Of course, it will be the larger EU economies that drive these results. But they are precisely the ones that have gone beyond that integration threshold. 3. It is also possible to tie in the Benigno (2004) argument that welfare improves if monetary policy attaches a higher weight to the higher inﬂation countries. On the face of it, this proposal seems to make little sense since all it does is shift the mean of the inﬂation distribution up, and hence the implicit inﬂation target for the usual case of pro-portional weights down, without taking the dispersion of inﬂation rates into account. But there is always a nonlinearity in any government’s evaluation of the available inﬂation-output trade-offs, which makes the aggregate outcomes greater than their average.4 So, all that Benigno is doing is replace the average of the outcomes by the correct aggregate in order to get better outcomes in terms of those evaluations. Put another way: because there is no ‘‘representative’’ country in this case, targeting average inﬂation (as if a representative country existed) will produce inferior outcomes and larger differentials if the Phillips curves are nonlinear or have different slopes. Indeed, there is a danger that to do so will leave the steady state equilibrium path undeﬁned. I return to that point below. 4. Another interesting result is the sensitivity of the model’s outcomes to the policy rule’s horizon. Equilibria are possible if interest rates target expected inﬂation in the next quarter; but explosive behaviour is observed if inﬂation is targeted four quarters ahead. Similar results are obtained in work by Batini and Nelson (2001), and Batini et al. (2004). The explanation is to be found in Salmon (1982): the solution to any dynamic targeting problem requires the policy rule to follow a difference equation process which is of order one greater than that driving the target variable you are attempting to reach. Hence, if the target variable follows a difference equation system of order p, the policy rule must be of order at least p+1 if that target is to be reachable in expectation. If policy follows a p-order rule instead, the target failures remain constant. If it follows a rule of order p1, you diverge from the target path and get explosive behaviour. Here, inﬂation and output follow a second order scheme (see equation (3) below) – and a third order system in the empirical implementation (see Table 3). Hence the highest order of policy rule that can be derived from the model will be 2, or 3, whereas the uncontrolled targets will follow a kth order scheme where k is the number of periods that inﬂation is projected forward before the policy rule kicks in. Consequently we get wellbehaved solutions if k ¼ 1, but no solutions at all if k ¼ 4 or higher. That explains one source of sensitivity in the simulation results.

4

See Hughes Hallett and Demertzis (1998)

258

Andrew Hughes Hallett

5. A second source lies in the possibility that controllability might not be guaranteed. In many of the numerical experiments, it is clear that unstable, volatile, or explosive solutions can be avoided if the policy response parameters become large enough. This suggests that the model easily becomes unstable with small changes in its parameter values; but that it remains stabilizable in the sense that feedback parameters large enough can always be found for the policy rule such that pre-assigned target values become reachable again. I show this below. It means that the desired targets can be reached even if the model’s parameter values stray into a region in which the model, left to itself, would show unstable behaviour. But to achieve this may require large changes in the policy rule or outcomes. Hence the sensitivity found in many of the numerical exercises here. 6. A third possibility is that changing parameters generate unstable equilibrium paths. This could happen two ways. First, it could imply the model itself becomes unstable or close to unstable. One example of this (see below) is when a2o1db or a1>db, i.e., when the impact of inﬂationary expectations becomes less than the impact of the real interest rate on inﬂation, or when persistence becomes larger than the impact of real interest rates. That would explain the appearance of explosive behaviour as d increases; or as b increases; or as a1 increases (and a2 decreases). All these outcomes appear in the numerical results. Second, it could also come from the aggregation effects within the Union where, for example, Phillips curves are nonlinear or linear with different slopes. But that is just one possibility. It will happen more generally if equal and opposite shocks in different countries, would drive inﬂation up in one country by more than they reduce it in another. Depending on how wage bargains are struck, that would mean that the natural (equilibrium) rate of unemployment for the union is undeﬁned.5 In this case, the equilibrium path is undeﬁned – you get a different path for each possible distribution of aggregate demand across countries. So, what drives the sudden changes and instability here is that each change in parameter values generates a new pattern of demand, and hence a different equilibrium path. Aggregate performance is then a function of the aggre-gation biases: in effect, the short-run Phillips curve has been pushed to the right in inﬂation-unemployment space, so that the trade-offs are worse than they would be, or need to be, had the dispersion of unemployment and inﬂation rates been minimized across countries. This has three consequences: (a) that the current (aggregate) policy rules are ‘‘inefﬁcient’’: that is, less seffective than they should be; (b) lower inﬂation and lower inﬂation differentials are indeed complements, as we see here; (c) Benigno’s weighting rule suddenly makes sense because, in order to get these better results, you need to replace the average inﬂation target with the proper (i.e., lower) aggregate target.

5

This result is due to Brechling (1973), but a demonstration for the present case is given in Hughes Hallett (2000). Equilibrium requires the variance of inﬂation (or unemployment rates) to be minimised.

Comment

259

So, it seems the fragility in the numerical results that we see here is as much the result of the equilibrium path not being tied down in the current system of European policymaking – as it is by a lack of controllability or stability. The implication is we need ﬁscal policy not so much to manage aggregate demand in the traditional sense, as to manage the distribution of that demand across countries or regions. This is needed to stabilize the underlying equilibrium and produce more effective policies. The controllability issue A model containing n equations of the form Y t ¼ AY t1 þ BX t þ vt

ð1Þ

is said to be stabilisable if it is also controllable. And it is controllable, meaning that speciﬁc values for the nendogenous variables E t Y tþk can be reached exactly in expectation for some k in the future, given appropriate choices for the policy variables X t , if the following matrix is of full rank: rank ½B; AB; A2 B; . . . ; Ak B ¼ n.

ð2Þ

If (1) is controllable in this sense, then some matrix K t in the feedback rule X t ¼ K t Y t1 þ kt implying Y t ¼ ðA þ BK t ÞY t1 þ vt þ Bkt can always be found such that all the roots of (A+BKt) lie within the unit circle. Applying such a policy rule means that the model has been stabilised – and hence controlled – even if the original model itself was unstable because some of the roots of A lie beyond the unit circle. In this case, we can take the model in Section 4, equations (1) and (2), renormalise on period t+1 and take expectations, to get ytþ1 A1 A2 yt ð3Þ ¼ þ Brt þ rest yt I 0 yt1 where the second-order difference equation system has been rewritten as a ﬁrst order one with the following deﬁnitions:

yt ¼

pt gapt

!

" ; A1 ¼

#

1=a2

b=a2

d=ða2 a4 Þ

ð1 db=a2 Þ=a4 0 1 0 B C B d=a4 C B C B¼B C B 0 C @ A 0

" A2 ¼

a1 =a2

0

da1 =ða2 a4 Þ

a3 =a4

#

Andrew Hughes Hallett

260

From here it is easy to compute the controllability matrix given in (2) above for the case of k ¼ 3. It takes the form 2

0 6b 6 1 6 40 0

a2 b2

a3 b3

0 b1

a2 b2

3 a4 b4 7 7 7 a3 5 b3

The elements of this matrix are deﬁned as a2 ¼ bd/(a2a4), a3 ¼ a2, a4 ¼ ða2 db=ða2 a4 Þ þ 1=a4 a3 =a4 ba1 =a2 Þbd=ða2 a4 Þ;and b1 ¼ d/a4, and b2 ¼ dðdb=a1 1Þ=a24 ; b3 ¼ ðd2 b=a2 dð1 a3 ÞÞ=a24 ; and b4 ¼ ðdb=a22 ð1 db=a2 a4 Þ2 = a4 ðdb=a2 1Þa3 =a4 dbða1 =a2 þ a3 =a4 Þ=a2 þ a3 =a4 Þd=a24 It is now easy to check that the controllability matrix is of full rank as long as b1 a0; i.e., as long as we have at least one effective instrument (d6¼0), and that the impact of an anticipated output gap is not zero (a46¼0). If that is true, the model is controllable and stabilisable, but only with a horizon of four quarters ahead. Otherwise, this economy would not be stabilisable, even with a lead time of one year. Finally, returning to (3), we can calculate the roots of this model. For stability before control, they all need to lie within the unit circle: the usual condition of having as many unstable roots as expectations having been reversed by the process of normalising on period t+1. Evidently, a sufﬁcient condition for at least one root to be outside the unit circle is 1db>a2 , or dboa1, if we use the restriction a2 ¼ 1a1 referred to in the paper. Since do0, the model is unstable without policy, but controllable with policy. This has some very real implications for European policymaking, all of which are clearly visible in the numerical results of this paper: (a) For a ﬁxed policy rule, increasing the strength of the transmissions will make the model more unstable, and hence increase inﬂation differentials or inﬂation persistence (b) Increasing internal persistence (market rigidities) will do the same. Hence, negative shocks in a less ﬂexible economy will impose a greater adjustment burden on the more ﬂexible economies (Hughes Hallett, 2003). (c) As rigidities and transmission effects strengthen, you need to target your target variables more aggressively. (d) The feedback parameter in the Taylor rule in equation (1.5) may have been enough to make the targets achievable in the baseline. But as the model’s parameters are increased, the Taylor rule’s parameters may need to be strengthened to values that might be thought implausible in a conventional setting.

Comment

261

References Batini, N. and E. Nelson (2001), ‘‘Optimal horizons for inﬂation targeting’’, Journal of Economic Dynamics and Control, Vol. 25, pp. 891–910. Batini, N., P. Levine and J. Pearlman (2004), ‘‘Indeterminacy with inﬂation forecast based rules in a two bloc model’’, ECB Working Paper, No. 340, Frankfurt. Benigno, P. (2004), ‘‘Optimal monetary policy in a currency area’’, Journal of International Economics, Vol. 63, pp. 293–320. Brechling, F. (1973), ‘‘Wage inﬂation and the structure of regional unemployment’’, Journal of Money Credit and Banking, Vol. 5, pp. 555–583. Hughes Hallett, A. (2000), ‘‘Aggregate Phillips curves are not always vertical: heterogeneity and mismatch in a multi-region or multi-sector economy’’, Macroeconomic Dynamics, Vol. 4, pp. 534–546. Hughes Hallett, A. (2003), Asymmetries and asymmetric policy transmissions in the Eurozone Submissions on EMU from Leading Academics, London: HM Treasury, HMSO. Hughes Hallett, A. and M. Demertzis (1998), ‘‘Asymmetric transmission mechanisms and the rise in European unemployment’’, Journal of Economic Dynamics and Control, Vol. 22, pp. 869–886. Hughes Hallett, A. and L. Piscitelli (2002), ‘‘Does one size ﬁt all? A currency union with asymmetric transmissions and a stability pact’’, International Review of Applied Economics, Vol. 16, pp. 71–96. Hughes Hallett, A. and L. Piscitelli (2002), ‘‘Does trade integration cause convergence?’’, Economics Letters, Vol. 75, pp. 165–170. Hughes Hallett, A., S.E.H. Jensen and C.R. Richter (2004), ‘‘The European economy at the cross roads: structural reforms, ﬁscal constraints and the Lisbon agenda’’, in: E. Hochreiter, editor, Macroeconomic Performance and Productivity under the Lisbon Agenda, Vienna: Austrian National Bank. Salmon, M. (1982), ‘‘Error correction mechanisms’’, Economic Journal, Vol. 92, pp. 616–629.

This page intentionally left blank

CHAPTER 9

Black Tie Required? How to Enter a Currency Union Volker Nitsch

1. Introduction One of the most dramatic changes in the international monetary system is the recent revival of monetary integration. For most of the post-war period, countries strongly preferred to have their own national currencies. When a territory gained political independence, the currency union link with the former colonizer was often quickly dissolved. More importantly, between 1950 and 1995, only two sovereign countries were willing to abandon their independent monies: Mali became again a member of the CFA franc zone in 1984 after it left the arrangement in 1962, and Equatorial Guinea joined the CFA franc zone in 1984. Since the mid-1990s, however, currency unions have become increasingly attractive again. In Europe, 12 nations have formed the European Monetary Union (EMU) and almost the same number of countries is expected to join the EMU within a few years. Ecuador and El Salvador have unilaterally adopted the US dollar as legal tender, possibly providing a blueprint for other Latin American countries. Given this growing interest in common currencies, it is noteworthy that the recent transitions to monetary union have been extremely diverse. The 12 EMU member countries, for instance, ﬁrst had to undergo a yearlong process of convergence before the monetary union was established. Ecuador, in contrast, has abandoned its national currency in a rather spontaneous and unplanned step.1

1

Barry Eichengreen (2002) refers to these schemes as the ‘‘coronation’’ approach and the ‘‘just do it’’ approach, respectively.

CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79009-5

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

264

Volker Nitsch

In this paper, I examine whether the entry strategy matters for monetary uniﬁcation. Most of the previous work on currency union formation discusses the costs and beneﬁts of adopting another currency and then aims at identifying potential candidates (partner or anchor countries) for successful monetary integration; see, for instance, the contributions in HM Treasury (2003) for an extensive discussion on whether the UK should join the EMU and A. Alesina et al. (2002) for a recent application of the cost-beneﬁt analysis to a much larger set of countries. Here, I ask a different question: given that a country has already decided to join a monetary union, how should the new currency be implemented? To analyze this issue, I explore the (few) different experiences in monetary union formation in the post-war period in more detail. More speciﬁcally, I analyze whether the transition approach has had a measurable effect on the subsequent macroeconomic performance of the country that has entered a monetary union. This issue seems to be of particular interest for EU accession countries, which currently face the option of convergence (to the Maastricht criteria) or unilateral adoption of the euro (‘‘euroization’’). Previewing the main results, I ﬁnd that there is no systematic association between the degree of pre-entry convergence and a country’s later economic performance within a currency union; there is little evidence that a long period of preparations signiﬁcantly improves economic growth. Based on these results, the economic rationale for requiring an extensive transition period (including participation in the exchange rate mechanism ERM II for a minimum period of two years) before recent EU entrants are allowed to join the euro appears weak. The remainder of the paper is organized as follows. The next section presents two recent country cases of currency union entry. In Section 3, I discuss some theoretical background. Section 4 describes the empirical setup and the data. Section 5 presents the results, followed by a discussion of some robustness checks. Section 7 provides a brief conclusion. 2. Implementing a monetary union: two case studies To illustrate the possible variation in the adoption of a foreign currency, I discuss two recent country cases in more detail: the accession of Greece to the EMU in 2001 and ofﬁcial dollarization in Ecuador in 2000. 2.1. Greece On February 7, 1992, the member countries of the European Union signed the Maastricht Treaty. Having eliminated all restrictions to the internal movement of capital, goods, and people, and sharing a strong desire for further economic integration, they agreed to establish a monetary union. With the treaty, the countries installed an institutional framework (including a detailed timetable) for the introduction of a common currency, and set in motion a gradual process

Black Tie Required? How to Enter a Currency Union

265

of convergence. More directly, based on the idea that a monetary union is only sustainable if the member countries share some common macroeconomic characteristics, the treaty stipulated a number of convergence criteria that needed to be fulﬁlled by a country that wanted to join the monetary union.2 When the treaty was signed, Greece was in serious divergence from the other member countries of the European Union. Unlike Portugal and Spain (both joined the EU in 1986), Greece had experienced no improvement in its macroeconomic performance after its accession to the EU (in 1981); for more than a decade, weak growth and high inﬂation continued.3 As a result, when it was decided in May 1998 that 11 countries would adopt a single currency, Greece was not able to join the EMU as it failed to meet any of the criteria. Beginning in 1994, however, Greece had implemented a convergence program. Adjustment measures included structural reforms (e.g., the abolishment of wage indexation), a ‘‘hard drachma’’ policy (in which the drachma was allowed to depreciate only by less than the inﬂation differential), and ﬁscal consolidation (which was achieved, among others, by widening the tax base). Owing to these policies, Greece’s economic performance improved markedly in the late 1990s; real GDP growth jumped to an average rate of 3.2 percent between 1995 and 2000, about 0.6 percent above the EU average. By 2000, ﬁscal deﬁcit was reduced to 0.8 percent of GDP (from about 13.5 percent in 1993); the debt-to-GDP ratio declined from a peak of 111.3 percent in 1996 to 103.9 percent; and inﬂation decreased to 3.3 percent. In June 2000 Greece was ﬁnally admitted into the euro area. On January 1, 2001, almost nine years after signing the Maastricht Treaty and two years after the formation of the monetary union, Greece became the 12th country to adopt the euro. 2.2. Ecuador In remarkable contrast to this experience, Ecuador’s decision to dollarize was taken in haste, and without preparations, when the country was in the midst of an economic and political crisis. The decision was announced by then president Jamil Mahuad on January 9, 2000. On January 21, Mahuad was ousted in a civilian-military coup, but his successor, former vice president Gustavo Noboa, opted to stick with dollarization. On March 13, 2000, Ecuador abandoned the sucre as national currency and adopted the US dollar as legal tender. Economic conditions in Ecuador had gradually begun to deteriorate in the mid-1990s when the country was hit by a series of external shocks: the El Nin˜o

2

An excellent early assessment of the process towards the EMU is provided in Charles Wyplosz (1997). 3 For the period 1980–94, real GDP growth in Greece averaged 0.8 percent (compared with an average rate of 2.0 percent in the EU) and inﬂation was on average 18.3 percent (compared with 6.4 percent in the EU). More details are documented in R. Bryant et al. (2001).

266

Volker Nitsch

ﬂoods in 1997 destroyed crops and infrastructure worth about 13 percent of the GDP; the fall in world oil prices in 1998 cut public sector revenues by about 3.5 percent of the GDP; and due to the spillover of the Russian ﬁnancial crisis into Latin America, external credit lines for the banking system were substantially reduced. Domestic policies, instead of countering these shocks, further weakened the Ecuadoran economy. Problems included political instability (with frequent changes of government and accusations of corruption and cronyism), chronic inﬂation, and social discontent. In 1999, the situation worsened dramatically when problems in an already fragile banking sector escalated. The closure of a bank in April 1998 triggered deposit runs on other banks and, after a series of ad hoc actions by the government (including a deposit freeze, bank holidays, and government bailouts), the banking system collapsed. At the same time, Ecuador’s ﬁscal position, which was already weak with a deﬁcit of about 5 percent of the GDP, rapidly deteriorated, thereby aggravating the ﬁnancial crisis. Finally, as the demand for dollars gained momentum, the sucre depreciated dramatically, pushing up inﬂation. By the end of 1999, output had dropped by about 7.5 percent; the annual inﬂation rate had accelerated to 60 percent; the sucre had lost about 80 percent of its dollar value; and the ratio of total public debt to the GDP was at 130 percent.4 Against this background, Ecuador adopted the US dollar as legal tender as a substitute to the sucre. S. Fischer (2001, p. 7), then ﬁrst deputy managing director of the International Monetary Fund (IMF), notes: ‘‘The decision to dollarize was taken in desperation. y If they had asked us, we would have said that the preconditions for making a success of dollarization were not in place.’’ Nonetheless, the economy stabilized after dollarization, and the ﬁnancial system in Ecuador recovered. 3. Background and literature Despite these large differences in regime transitions, it is a priori unclear which entry-approach to monetary union is preferable. Theory appears to provide no conclusive answer. According to the standard literature on optimum currency areas (OCAs), as developed by R. Mundell (1961), P. Kenen (1969), and R. McKinnon (1963), a monetary union is desirable when territories display a high degree of economic integration. Speciﬁcally, it is argued that a common currency is suitable if the members of a potential OCA satisfy the following criteria: a high degree of bilateral trade integration; a strong correlation of domestic business cycles; free mobility of labor; and a system of ﬁscal transfers. Accordingly, there should have been substantial convergence between potential member countries before a monetary union is entered.

4

A more detailed account of the Ecuadoran crisis is given in Luis and Jacome (2004).

Black Tie Required? How to Enter a Currency Union

267

Alternatively, it is possible to emphasize the integration effects of a common currency. While OCA literature focuses mainly on (minimizing) the costs of monetary integration, this approach concentrates on the potential beneﬁts of a common currency. Currency unions may then become more attractive for country pairs with some dissimilarity. For instance, given that the adoption of a new currency cannot be easily reversed, entry into a monetary union appears to be a serious and credible commitment device that is particularly useful for countries that lack the internal discipline for monetary policy. Alesina et al. (2003) argue that ‘‘the countries that stand to gain the most from giving up their currencies are those that have a history of high and volatile inﬂation.’’ Moreover, as Jeffrey Frankel and Andrew Rose (1998) have shown, the OCA criteria for whether a country should join a monetary union are largely endogenous; sharing a common currency increases trade integration and business cycle correlation. In summary, convergence before entry into a monetary union may be of less importance. Finally, it is possible that some other criteria may affect the success of monetary integration.5 Potential candidates include the Maastricht criteria that must be legally met by countries willing to join the EMU. As is well known, the economic rationale for these conditions, price stability, exchange rate stability, convergence of nominal long-term interest rates, and ﬁscal discipline, is weak. Frankel (2004, p. 15), for instance, notes: ‘‘The four Maastricht conditions, particularly the ﬁscal criterion, are not very closely based on international monetary theory.’’ Other potentially important macroeconomic variables can be borrowed from currency crises literature. In these studies, typically money growth and the current account, among others, feature prominently for identifying changes in the exchange rate regime. Given these general ambiguities, ﬁnding conditions for successful monetary integration appears to be ultimately an empirical issue. I now turn to this task. 4. Data Until recently, only a few economies used currencies other than their own.6 For more than half a century, no new currency union was created, and although some currency unions remained in existence, many more common currency links

5

I focus here exclusively on economic criteria. In recent years, there has also been a growing discussion on the potential relevance of non-economic criteria, such as public support for the adoption of another currency (R. Hausmann and A. Powell, 1999) or the existing degree of currency substitution (G. Calvo, 1999). W. Gruben et al. (2003) provide detailed implementation guidelines for the introduction of a monetary union. 6 Alesina and Barro (2002, p. 381) note that roughly 60 territories have for some time been members of a currency union. Most of these territories, however, are extremely small (e.g., Andorra, Liechtenstein, Monaco) so that the group of countries is much smaller economically (e.g., in terms of GDP) than the plain number of countries suggests. Rose and Charles Engel (2002) provide a list of member countries of monetary unions.

268

Volker Nitsch

were dissolved. Reuven Glick and Rose (2002), for instance, compile regime transitions for the period from 1948 through 1997, and ﬁnd that 130 of the 146 switches (for which they have data) were currency union exits. Nonetheless, some countries and territories have adopted a foreign currency in the post-war period; these experiences provide the basis for the empirical analysis in this paper. Table 1 lists the currency union entries.7 Given the small number of episodes and their considerable heterogeneity (covering different time periods and countries of various sizes and income levels), the choice of an appropriate benchmark is potentially important. I use as control group the member countries of the monetary union whose currency is being adopted. That is, I construct country pairs of joining or client countries, on the one hand, and the respective currency union member or anchor countries, on the other, thereby excluding all countries that are not members of a currency union and the member countries of other currency unions.8 This approach offers at least two advantages. First, the countries in the control group are geographically close and often share the same economic characteristics with the entering country so that the empirical results should not be affected by nonlinearities. Second, the approach allows exploring the degree of convergence that has been achieved by the country that adopts the foreign currency shortly before monetary union entry. In total then, my sample consists of seven currency union entries and 58 bilateral country pairs. If I additionally include the formation of the EMU (which appears to be reasonable), the sample increases to 17 entries and 68 country pairs.9 Owing to missing observations, however, the actual sample is often somewhat (but not much) smaller. The data are compiled from a number of sources. Since data for some countries were often not available or were of poor quality, I collected data from various standard ﬁles and cross-checked the information or ﬁlled in missing observations. The main source of data is the World Bank’s World Development Indicators, which provides information on several country-speciﬁc variables, including population, land area, and inﬂation, as well as on exports and imports, gross foreign direct investment, gross private capital inﬂows, the current account balance, the monetary aggregate M2, the overall budget deﬁcit, and the central government debt (all as shares of the GDP). Additional variables, such as real GDP growth, the nominal exchange rate, and exports and imports (in US dollars), were obtained from the IMF’s International Financial Statistics.

7

I do not include monetary uniﬁcations between territories that also form a political union (e.g., German uniﬁcation in 1990). 8 For a comparison of countries in currency unions to countries with sovereign monies, see Rose and Engel (2002). 9 Substituting the EMU average (minus the country in question) for Germany as pre-EMU anchor does not change any of the results.

Black Tie Required? How to Enter a Currency Union

Table 1.

Currency union entries in sample

Countries that have entered a currency union after 1945 (date of entry):

Existing currency union members/Anchor countries

Equatorial Guinea (August 27, 1984) Guinea-Bissau (May 2, 1997) Mali (June 1, 1984)

Benin Burkina Faso Cote d’Ivoire Niger Senegal Togo Cameroon Central African Republic Chad Congo, Rep. Gabon United States Panama

The Bahamas (February 1970) Bermuda (February 6, 1970) Ecuador (March 13, 2000) El Salvador (January 1, 2001) Greece (January 1, 2001)

Austria (January 1, 1999) Belgium (January 1, 1999) Finland (January 1, 1999) France (January 1, 1999) Ireland (January 1, 1999) Italy (January 1, 1999) Luxembourg (January 1, 1999) Netherlands (January 1, 1999) Portugal (January 1, 1999) Spain (January 1, 1999)

269

Austria Belgium Finland France Germany Ireland Italy Luxembourg Netherlands Portugal Spain Germany

Wherever possible, I also add information from the Penn World Table 6.1 and the United Nations’s statistical databases. The bilateral trade data are taken from the IMF’s Direction of Trade Statistics. 5. Empirical results I begin with a simple characterization of countries that have entered a currency union. In Table 2, I report for a number of variables means (and standard

Volker Nitsch

270

Table 2.

Characterizing currency union entrants CU entrant

Variable Population (thousands) Land area (sqkm) Real GDP per capita ($) Real GDP growth (%) Inﬂation (%) M2/GDP (%) Exports/GDP (%) Imports/GDP (%) Export growth (%) Import growth (%) Current account/GDP (%) FDI/GDP (%) Private capital ﬂows/GDP (%) Budget deﬁcit/GDP (%) Government debt/GDP (%)

Obs. 68 68 60 64 55 32 48 49 59 59 49 19 19 43 35

Mean

Existing CU member

Std. Dev.

7,280.7 10,873.2 274,759.5 432,069.7 7,112.4 8,522.8 3.80 3.17 13.25 20.98 29.96 27.43 24.01 19.40 34.22 12.45 0.09 16.37 11.01 24.26 6.36 8.02 1.91 2.79 15.82 30.98 5.02 137.05

4.33 104.15

Obs.

Mean

Std. Dev.

Test of equality (p-value)

68 33,439.9 59,842.7 68 874,895.1 2,115,092 60 10,258.3 10,936.2 64 3.69 6.90 55 4.30 4.99 32 31.00 21.26 48 35.41 23.79 49 38.85 26.30 59 16.90 63.99 59 12.11 85.59 49 6.52 20.32 19 1.40 1.06 19 13.16 11.16

0.00 0.03 0.00 0.91 0.00 0.61 0.02 0.30 0.07 0.06 0.96 0.37 0.66

43 35

0.00 0.00

1.14 68.99

2.35 28.25

Note: Maximum sample size ¼ 68.

deviations) for currency union entrants and existing currency union members; the last column provides the p-value of a t-test for equality. As shown, countries that have entered a currency union are typically signiﬁcantly smaller and poorer than existing currency union members; this result, which is robust to the exclusion of the US as an anchor country, conﬁrms previous ﬁndings about countries without sovereign monies (Zeljko Bogetic, 2000; Sebastian Edwards, 2001; Rose and Engel, 2002). Entering countries also have higher inﬂation. They are less open to international trade (despite their small size) and experience much lower growth rates in international trade.10 Fiscal conditions in entering countries are in signiﬁcantly worse shape, as measured by both the budget deﬁcit and the overall government debt (as percentages of the GDP). Somewhat surprisingly, output growth rates appear to be similar for entering countries and currency union members. In the next step, I analyze whether the differences in the macroeconomic conditions have become smaller over time; that is, I ask: Do countries converge before they join a monetary union? To analyze this issue, I estimate a simple time-series regression of the (absolute) pair-wise difference of the variable of interest on a linear time trend (with country-pair ﬁxed effects); a negative slope coefﬁcient would then imply that the countries had on average become

10

The ﬁnding of no signiﬁcant differences in ﬁnancial openness may be due to limited data availability.

Black Tie Required? How to Enter a Currency Union

271

more similar before they adopted the same currency. The results are reported in Table 3; I tabulate separate values for the three-year period before entry and the ﬁve years before entry. Reviewing the results, there is little evidence of convergence before joining a monetary union. For most variables, the differences are statistically unchanged in the run-up to the adoption of a common currency. In two cases, however, the coefﬁcient on the time trend variable is consistently negative and statistically signiﬁcant: the differences in export growth rates tend to decline over time (possibly reﬂecting some trade liberalization before currency union entry), and the differential in ﬁscal debt seems to become smaller (which may be mostly the result of the convergence process in Europe for which ﬁscal data are readily available). Interestingly, the inﬂation differential ﬁrst appears to get smaller when the ﬁve-year period is analyzed, but then gets on average larger in the three-year period before accession. This divergent pattern probably reﬂects the different motives for currency union entry; some countries abolished their national currency when they faced accelerating inﬂation. My main interest, however, is to identify possible determinants for the success of monetary integration. A natural starting point for this analysis is to Table 3.

Is there convergence before currency union entry? Five-year period before entry

Dependent variable:

Coeff.

Real GDP growth 0.05 Inﬂation 1.96 M2/GDP 0.27# Exports/GDP 0.22 Imports/GDP 0.38 Export growth 12.03 Import growth 0.73 Export duties 0.32 Import duties 0.90 Current account/GDP 0.37 FDI/GDP 0.08 Private capital ﬂows/GDP 1.72 Budget deﬁcit/GDP 0.33 Government debt/GDP 1.87 Bilateral exchange rate 6.98

Std. Err. Obs. 0.19 0.37 0.16 0.44 0.61 4.85 1.80 0.10 0.29 0.47 0.05 0.83 0.14 0.58 5.28

318 275 200 240 243 314 318 87 86 236 134 134 194 147 340

Three-year period before entry

R2

Coeff.

Std. Err.

Obs.

R2

0.31 0.72 0.95 0.85 0.69 0.28 0.27 0.67 0.87 0.50 0.85 0.97 0.29 0.99 0.01

0.17 2.94 0.23 0.98 0.68 9.35 7.08# 0.69 0.31 0.58 0.06 0.95 0.44 3.55 17.51

0.45 0.88 0.31 0.46 0.74 4.01 4.26 0.12 0.46 0.98 0.17 1.83 0.33 1.14 12.87

192 165 102 142 145 190 192 37 37 136 68 68 128 100 204

0.27 0.64 0.94 0.97 0.91 0.43 0.39 0.88 0.94 0.51 0.84 0.81 0.25 0.99 0.01

Notes: The table reports country-pair ﬁxed effects OLS estimates of f1 from DepVari;j;t ¼ f0 þ f1 Time trend þ where DepVari,j,t is the absolute difference in the variable of interest between currency union entrant i and existing member country j at time t. For bilateral exchange rate, DepVari,j,t is the absolute percentage change in the nominal bilateral exchange rate. Intercepts are not reported. signiﬁcance at the 1 percent level signiﬁcance at the 5 percent level # signiﬁcance at the 10 percent level.

272

Volker Nitsch

assess the sustainability of a currency union. In currency unions that have been dissolved, the pre-conditions for successful monetary integration were obviously not in place; the beneﬁts of sharing the same currency were smaller than the costs. All currency unions, however, that have been either created or extended in the post-war period are still in existence.11 Therefore, I deﬁne two alternative measures for the macroeconomic performance of a currency union. First, I compute the change in the GDP growth rate of the accession country. This measure, which is the average difference in real output growth for the ﬁve years before and the ﬁve years after entry, captures the direct impact of currency union entry on the accession country.12 As a second measure, I relate the performance of the entering country to the macroeconomic performance of the existing currency union members by computing the average pair-wise difference in GDP growth rates for the ﬁve years after currency union entry. Both performance measures are positively related, but far away from being close substitutes; the correlation is 0.59. To provide more intuition, Table 4 lists the ﬁve best and the ﬁve worstperforming currency union entrants, according to our performance measures. Three observations appear to be particularly noteworthy. First, there is considerable variation in the relative growth performance of countries that have entered a currency union. Second, there are some notable outliers, possibly ampliﬁed by poor quality of the data. Third, the two measures provide in general a quite consistent picture; they indicate that Ecuador and Mali have been the most successful entrants, and Guinea-Bissau is the worst performer. In the following, I analyze whether the level of (pre-entry) convergence had a measurable impact on the subsequent performance of a country that has entered a monetary union. In particular, I estimate equations of the form PERFORMi;t ¼ a þ b CONVij;t1 þ Sij;t , where PERFORMi,t is the macroeconomic performance measure of entering country i as explained above and CONVij,t1 is the measure of convergence before entry (with variables often suggested by theory); b is the coefﬁcient of interest to me.13 In a ﬁrst exercise, I explore whether the standard OCA criteria help to explain the differences in the relative growth performance of countries that have entered a monetary union. I begin with a simple measure of bilateral trade integration: the share of trade with the prospective partner country in the joining country’s total trade. In my sample, this measure ranges from 0 to 67 percent (for the pair Bahamas-US), with a mean value of 7.3 percent. As shown in the

11

Nitsch (2004) provides a characterization of currency union exits. The approach is close in style to Hausmann et al. (2004) method to identify growth accelerations. 13 In view of the small sample size, I usually do not include additional control variables. 12

Black Tie Required? How to Enter a Currency Union

Table 4.

273

Performance measures

Growth differential between CU entrant and CU member (difference in average real GDP growth rate ﬁve years after CU entry) CU entrant

Anchor/CU member country

Five best-performing entrants Ireland Germany Ecuador Bermuda Mali Congo, Rep. Luxembourg Germany Greece Belgium Five worst-performing entrants Guinea-Bissau Equatorial Guinea The Bahamas Panama The Bahamas United States Guinea-Bissau Mali Guinea-Bissau Cote d’Ivoire/Chad

GDP growth entrant

GDP growth member

Growth differential

8.75 4.27 4.58 4.85 3.97

1.46 0.43 1.17 1.46 0.68

7.29 4.70 3.41 3.39 3.29

1.27 5.17 5.17 1.27 1.27

33.43 3.53 2.34 4.48 4.10

34.70 8.70 7.51 5.75 5.38

Change in average growth rate of CU entrant (difference between average real GDP growth rate ﬁve years after and ﬁve years before CU entry) GDP growth before entry GDP growth after entry Change in growth rate Five best-performing entrants Ecuador Mali Belgium Spain Portugal

0.17 0.54 1.93 1.95 2.13

4.27 4.58 3.86 3.50 3.55

4.10 4.04 1.93 1.55 1.42

Five worst-performing entrants Guinea-Bissau El Salvador Netherlands Equatorial Guinea Finland

4.36 3.91 3.45 2.56 3.38

1.27 1.97 2.55 1.99 2.93

5.63 1.94 0.90 0.58 0.45

ﬁrst column of Table 5, the effect on the relative growth performance is essentially zero. The estimated coefﬁcient on this trade variable is statistically insigniﬁcant so that there is basically no evidence for the OCA hypothesis that entering countries beneﬁt from particularly close trade ties with the anchor country. Moreover, the results remain practically unchanged when I additionally control for the entering country’s level of trade openness by using the share of bilateral trade in the joining country’s GDP. In summary, bilateral trade intensity prior to monetary uniﬁcation appears to be no important precondition for successful monetary integration.

Volker Nitsch

274

Table 5.

Do OCA criteria matter? Growth differential after entry

Share of bilateral trade 0.003 in entrant’s total trade (0.051) Share of bilateral trade 0.003 in entrant’s GDP (0.002) Correlation of demeaned GDP growth rates Correlation of detrended GDP growth rates Obs. R2

60 0.00

47 0.10

0.06 (0.06)

2.42 (2.18)

0.002 0.001 (0.002) (0.001) 0.49 (2.51)

0.95 (2.08) 3.48# (1.93)

64 0.02

0.03 (0.04)

58 0.12

2.42* (1.05) 58 0.07

58 0.22

2.70 (1.75) 47 0.10

47 0.24

Change in entrant’s GDP growth rate Share of bilateral trade in entrant’s total trade Share of bilateral trade in entrant’s GDP Correlation of demeaned GDP growth rates Correlation of detrended GDP growth rates Obs. R2

0.04 (0.07)

0.06 (0.07)

0.02 (0.05)

0.001 (0.003) 0.37 (1.61)

1.66 (1.70) 1.65 (1.37)

58 0.02

47 0.02

64 0.00

0.002 0.001 (0.003) (0.002) 2.03 (2.01)

58 0.02

1.48 (1.15) 58 0.04

58 0.06

1.87 (2.21) 47 0.04

47 0.07

The results are somewhat stronger for the measures of business cycle synchronization. Following the literature (Rose and Engel, 2002), I experiment with two different measures of business cycle similarity – the pair-wise correlation of deviations from the country’s mean output growth rate (before entry into monetary union) and the pair-wise correlation of deviations from the country’s trend output growth rate; the correlation coefﬁcient for the two alternative measures is 0.62. For both measures, the estimated coefﬁcients are consistently positive and, in some speciﬁcations, statistically signiﬁcant at conventional levels of conﬁdence. Moreover, the estimates slightly improve when (to check robustness) the outliers in the performance measures are dropped (results are not reported). Taken together, the empirical ﬁndings are not very encouraging for the OCA hypotheses.14 Based on the experience of currency union entries in the post-war

14

I have no data on labor market ﬂexibility.

Black Tie Required? How to Enter a Currency Union

275

period, I ﬁnd that trade intensity with the anchor/currency union partner country is largely irrelevant for the subsequent growth performance of the country that abandons its national currency while there is at best only weak evidence that the commonality of shocks improves macroeconomic performance. In the next step, I test the stabilization or nominal anchor hypothesis of currency union formation. Do countries with high inﬂation beneﬁt most strongly from the adoption of another (stable) currency? The estimation results, presented in Table 6, suggest the opposite: the relative growth performance of a currency union entrant clearly seems to improve with lower inﬂation, a smaller inﬂation differential, and less inﬂation variability. A closer examination of the underlying raw data, however, shows that the sample includes two countries with particularly high rates of inﬂation; Guinea-Bissau and Ecuador experienced inﬂation rates in excess of 45 percent before they entered a currency union. For all other countries in the sample, inﬂation is lower by almost an order of magnitude, as illustrated in Figure 1.15 Therefore, to explore the sensitivity of the estimates, I reestimated the regressions when the high-inﬂation countries are dropped. For this perturbation, as shown in the bottom half of Table 6, the estimated coefﬁcients generally lose signiﬁcance and sometimes even change sign. More generally then, I ﬁnd no convincing evidence that inﬂation rates affect the success of monetary integration. (I note in passing that the relative growth performance of the high-inﬂation countries varies sizably, with Ecuador being a highly successful dollarizer while Guinea-Bissau showed the by far worst macroeconomic performance of all currency union entrants [as noted above; see Table 4]. However, since Guinea-Bissau joined a multilateral monetary union with 13 member countries, this experience entered with a much larger weight in my regressions.) Given the generally negative results for convergence in the variables suggested by theory, perhaps other factors affect the outcome of monetary uniﬁcation. A ﬁrst set of potential candidates is the Maastricht criteria, which have to be met by countries that wish to join the EMU. Apart from price stability (for which I ﬁnd no measurable effect on the performance of the entering country), these conditions include exchange rate stability and ﬁscal prudence. Table 7 suggests that the evidence on these variables is mixed. On the one hand, exchange rate stability appears to be completely irrelevant for the success of monetary integration; the coefﬁcient on various measures of exchange rate volatility is, if anything, positive, suggesting that countries that face large exchange-rate ﬂuctuations tend to beneﬁt most strongly from currency union entry. As shown in Figure 2, Ecuador, one of the best performing entrants, experienced a large devaluation before currency union entry.

15

Figure 1 illustrates the striking pattern of inﬂation convergence across European countries before the decision on membership in the EMU was made and inﬂation divergence afterwards; see P. Honohan and P. Lane (2003) for a discussion.

Volker Nitsch

276

Table 6.

Do high-inflation countries benefit from currency union entry? Growth differential after entry

Full sample Inﬂation in entering country Inﬂation differential Inﬂation variability in entering country Difference in inﬂation variability

Change in entrant’s GDP growth rate

Coeff.

Obs.

R2

Coeff.

Obs.

R2

0.13* (0.06) 0.12** (0.04) 0.25** (0.04) 0.25** (0.04)

57

0.22

53

0.31

55

0.16

51

0.37

57

0.29

53

0.64

57

0.20

0.09 (0.07) 0.11 (0.07) 0.24** (0.04) 0.27** (0.05)

53

0.62

40

0.07

36

0.85

39

0.09

35

0.00

40

0.12

36

0.09

40

0.00

36

0.35

Drop high inflation countries (inflation >40%) Inﬂation in entering country 0.24 (0.31) Inﬂation differential 0.30 (0.30) Inﬂation variability in entering 0.60 country (0.58) Difference in inﬂation variability 0.04 (0.11)

0.62** (0.07) 0.002 (0.10) 0.46 (0.46) 0.36** (0.08)

Notes: In the second sample, the two high-inﬂation countries (inﬂation >40%) – Guinea-Bissau and Ecuador – are dropped.

On the other hand, the results seem to be more supportive for ﬁscal conditions. The coefﬁcients take on the expected sign and are often statistically different from zero: a positive budget balance and lower debt levels seem to be associated with improved macroeconomic performance. There are two caveats, however, to these results. First, the threshold levels that are deﬁned in the Treaty of Maastricht (i.e., that countries are required to keep the deﬁcit below 3 percent of GDP and to maintain a debt level below 60 percent of GDP) are of no measurable importance for monetary integration; the coefﬁcient on a dummy variable for compliance with these criteria is highly insigniﬁcant. Second, the results appear to be affected by an outlier; Guinea-Bissau entered the CFA franc zone in serious ﬁscal disorder with a budget deﬁcit of 12.2 percent of the GDP and a debt-to-GDP ratio of 319.1 percent. When Guinea-Bissau is dropped from the sample, the coefﬁcients generally lose signiﬁcance and often change sign (bottom panel of Table 7). Finally, I explore a diverse set of other variables that are occasionally found to be important for the sustainability of exchange rate regimes. Table 8 reports the results. For most of these variables, the estimated effect on the entering

Black Tie Required? How to Enter a Currency Union

Figure 1.

277

Inflation rates.

120 100 80 60 40 20 0 -20

GuineaBissau

Ecuador

-40 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5

0

5

10 15 20 25 30 35 40 45 50 55 60

5

10 15 20 25 30 35 40 45 50 55 60

30 20 10 0 -10 -20

Mali Bahamas El Salvador

EquatorialGuinea Bermuda Greece

-30 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5

0

8 6 4 2 0 -2

Austria Italy

Belgium Luxembourg

Finland Netherlands

France Portugal

Ireland Spain

-4 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5

0

5

10 15 20 25 30 35 40 45 50 55 60

country is close to zero, with one exception: the coefﬁcient on the import growth differential is signiﬁcantly negative. This ﬁnding, however, which suggests that large pair-wise differences in the import growth rate are typically associated with a worse macroeconomic performance of the entering country, is probably due to simultaneity.

Volker Nitsch

278

Table 7.

Do the Maastricht criteria matter for the success of monetary integration? Growth differential after entry

Full sample Exchange rate conditions Average yearly percentage change in bilateral exchange rate (2 years before entry) Standard deviation of bilateral exchange rate (2 years before entry) Dummy for yearly percentage change in bilateral exchange rate o15% Fiscal conditions Deﬁcit/GDP in entering country Deﬁcit/GDP (difference)

Change in entrant’s GDP growth rate

Coeff.

Obs.

R2

Coeff.

Obs.

R2

0.04* (0.02)

68

0.01

0.01 (0.02)

64

0.00

0.68# (0.35)

68

0.02

1.05* (0.37)

64

0.12

2.21 (3.12)

68

0.04

2.79 (2.64)

64

0.17

53

0.33

53

0.32

43

0.47

43

0.23

53

0.11

53

0.01

53

0.37

53

0.67

35

0.61

35

0.64

53

0.01

0.48# (0.25) 0.38 (0.23) 0.76 (3.04) 0.03** (0.005) 0.03** (0.005) 0.70 (2.20)

53

0.00

40

0.56

34

0.30

40

0.67

40

0.11

27

0.05

40

0.37

0.74** (0.16) 0.49** (0.08) Dummy for entrant’s deﬁcit/ 3.83 GDP ratio o3% (2.65) Debt/GDP in entering country 0.03** (0.004) Debt/GDP (difference) 0.03** (0.004) Dummy for entrant’s debt/ 1.29 GDP ratio o3% (2.21) Drop Guinea-Bissau, a country in extreme ﬁscal Fiscal conditions Deﬁcit/GDP in entering 0.10 country (0.12) Deﬁcit/GDP (difference) 0.23# (0.12) Dummy for entrant’s deﬁcit/ 0.24 GDP ratio o3% (0.53) Debt/GDP in entering country 0.01 (0.01) Debt/GDP (difference) 0.0004 (0.01) Dummy for entrant’s debt/ 1.21 GDP ratio o3% (0.78)

disorder 40

0.02

34

0.12

40

0.00

40

0.04

27

0.00

40

0.07

0.57** (0.08) 0.38** (0.10) 3.28** (0.65) 0.02 (0.02) 0.01 (0.01) 3.02 (1.15)

Black Tie Required? How to Enter a Currency Union

Figure 2.

279

Nominal exchange rates.

1200

Guinea-Bissau

Ecuador

1000 800 600 400 200 0 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5

0

5

10 15 20 25 30 35 40 45 50 55 60

130 120 110 100 90 80 70

Mali

Equatorial Guinea

Bahamas

Bermuda

El Salvador

Greece

60 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5

0

5

10 15 20 25 30 35 40 45 50 55 60

130 120 110 100 90

Austria Finland Ireland Luxembourg Portugal

80 70

Belgium France Italy Netherlands Spain

60 -60 -55 -50 -45 -40 -35 -30 -25 -20 -15 -10 -5

0

5

10 15 20 25 30 35 40 45 50 55 60

Note: The base of the index is 100 for the month prior to currency union entry.

Volker Nitsch

280

Table 8.

Do other criteria affect the success of monetary integration? Growth differential after entry

Log of real GDP per capita (difference) M2/GDP (difference) Exports/GDP (difference) Imports/GDP (difference) Exports growth (difference) Imports growth (difference) Current account/GDP (difference) FDI/GDP (difference) Private capital ﬂows/ GDP (difference)

Change in entrant’s GDP growth rate

Coeff.

Obs.

R2

Coeff.

Obs.

R2

0.12 (0.61) 0.06 (0.05) 0.06 (0.05) 0.10 (0.06) 0.06 (0.04) 0.06** (0.01) 0.19** (0.02) 0.005 (0.151) 0.04** (0.01)

60

0.00

60

0.00

32

0.01

32

0.04

48

0.10

48

0.02

49

0.29

49

0.02

64

0.22

62

0.15

64

0.63

62

0.19

49

0.44

49

0.00

19

0.00

19

0.54

19

0.50

0.31 (0.69) 0.08 (0.08) 0.02 (0.02) 0.02 (0.01) 0.03 (0.02) 0.02** (0.01) 0.003 (0.011) 0.55** (0.06) 0.02 (0.01)

19

0.08

6. Robustness To check the robustness of the (generally negative) results, I perform extensive sensitivity analyses. None of the perturbations, however, sizably affects the main ﬁndings.16 For instance, I have increased sample size by additionally including currency boards, thereby also covering moves to another (very) hard exchange rate regime. Table 9 lists the episodes in which countries have introduced a currency board arrangement in the sample period. Not surprisingly, the results are basically unchanged; the number of additional observations is rather small. Moreover, the estimates (unreported here to save space) are, if anything, even somewhat weaker for the beneﬁts of pre-entry convergence hypothesis since most of the (unilateral) moves to a currency board are exercises of (successful) stabilization. I have also experimented with extending the sample by making better use of the time-series dimension of the data, but without much success.

16

In future work, I also aim to apply another estimation approach that makes it possible to deal more explicitly with issues such as endogeneity. A promising alternative seems to be the two-stage approach suggested by the discussant.

Black Tie Required? How to Enter a Currency Union

Table 9.

281

Currency board entries

Countries that have adopted a currency board after 1945 (date of entry):

Anchor countries

Argentina (April 1, 1991) Bosnia (August 12, 1997) Bulgaria (July 1, 1997) Cayman Islands (May 1, 1972) Estonia (June 20, 1992) Hong Kong (October 17, 1983) Lithuania (April 1, 1994) Lithuania (February 2, 2002)

United States Germany Germany United States Germany United States United States Germany

Another potential issue is the large heterogeneity in the sample. In principle, it could be argued that convergence is more important for multilateral currency unions, in which all member countries have a say, than for unilateral dollarization. Seeking to avoid negative consequences for themselves, incumbents in multilateral unions typically have a much stronger incentive to impose entry criteria on candidates than anchor countries. However, if I control for this difference in the type of monetary union, the results remain largely unaffected (mainly because of non-convergence of entrants to the CFA franc zone). Finally, I substitute inﬂation for economic growth as performance measure. However, given the large positive correlation between the two measures (of 0.79 and 0.45 for the change in the entrants’ performance and the difference in performance to existing members/anchor countries, respectively), the effect on the results is again negligible. 7. Conclusions In this paper, I analyze post-war experiences of currency union formation to identify possible factors that affect the feasibility of sharing a common currency. This issue is of large relevance for EU accession countries which, having recently entered the Union, are now also expected to join the EMU in the future. When should they adopt the euro? The economic performance of countries that have entered a monetary union differs widely, both relative to the period before they abandoned their national currency and to the experience of the anchor country. Analyzing these various episodes, however, I am unable to ﬁnd evidence that there are useful pre-requisites for monetary integration. Conditions suggested by theory (OCA) and those implemented by policymakers (Treaty of Maastricht) both do not appear to affect the success of monetary uniﬁcation. This result generally supports the claim by Hausmann and Powell (1999) that there are almost no pre-conditions for monetary integration, and notes (p. 5) that ‘‘minimum technical requirements [for dollarization] are few’’. As a result, the empirical analysis suggests

282

Volker Nitsch

that there is little economic reason for an extensive period of convergence of future EMU entrants. Acknowledgements I thank for useful comments Helge Berger, George Bitros, Sarantis Kalyvitis, Thomas Moutos, E´va Katalin Polga´r, Andrew Rose, and Ju¨rgen von Hagen; seminar participants at the Deutsche Bundesbank and the Free University Berlin; and participants at the conference on ‘‘Globalization of Capital Markets’’ at the Free University Berlin, the CESifo conference on ‘‘Designing the New EU’’ in Delphi, and the APF/Deutsche Bundesbank conference on ‘‘Asymmetries in Trade and Currency Arrangements in the 21st Century’’. References Alesina, A. and R.J. Barro (2002), ‘‘Currency unions’’, Quarterly Journal of Economics, Vol. 117(May), pp. 409–436. Alesina, A., R.J. Barro, and S. Tenreyro (2003), ‘‘Optimal currency areas’’, in: M. Gertler and K.S. Rogoff, editors, NBER Macroeconomics Annual: 2002, Cambridge, MA: MIT Press. Bogetic, Z. (2000), ‘‘Ofﬁcial dollarization: current experiences and issues’’, Cato Journal, Vol. 20(Fall), pp. 179–213. Bryant, R.C., N.C. Garganas and G.S. Tavlas (2001), ‘‘Introduction’’, in: R.C. Bryant, N.C. Garganas and G.S. Tavlas, editors, Greece’s Economic Performance and Prospects, Athens: Bank of Greece and Brookings Institution. Calvo, G.A. (1999), On dollarization, Maryland: University of Maryland. Edwards, S. (2001), ‘‘Dollarization: myths and realities’’, Journal of Policy Modeling, Vol. 23(April), pp. 249–265. Eichengreen, B. (2002), ‘‘When to dollarize’’, Journal of Money, Credit and Banking, Vol. 34(February), pp. 1–24. Fischer, S. (2001), ‘‘Ecuador and the International Monetary Fund’’, in: A. Alesina and R.J. Barro, editors, Currency Unions, Stanford: Hoover Press. Frankel, J.A. (2004), Real convergence and euro adoption in central and eastern Europe: trade and business cycle correlations as endogenous criteria for joining EMU?, Cambridge, MA: Harvard University. Glick, R. and A.K. Rose (2002), ‘‘Does a currency union affect trade? The timeseries evidence’’, European Economic Review, Vol. 46(June), pp. 1125–1151. Gruben, W.C., M.A. Wynne and C.E.J.M. Zarazaga (2003), ‘‘Implementation guidelines for dollarization and monetary unions’’, in: E.L. Yeyati and F. Sturzenegger, editors, Dollarization, Cambridge, MA: MIT Press. Hausmann, R. and A. Powell (1999), Dollarization: issues of implementation, Washington, DC: Inter-American Development Bank. Hausmann, R., L. Pritchett, and D. Rodrik (2004), ‘‘Growth accelerations’’, NBER Working Paper, No. 10566.

Black Tie Required? How to Enter a Currency Union

283

HM Treasury (2003), Submissions on EMU from Leading Academics, London: HM Treasury. Honohan, P. and P.R. Lane (2003), ‘‘Divergent inﬂation rates in EMU’’, Economic Policy, Vol. 37(October), pp. 357–394. Kenen, P.B. (1969). ‘‘The theory of optimum currency areas: an eclectic view’’, in: R.A. Mundell and A.K. Swoboda, editors, Monetary Problems of the International Economy, Chicago: Chicago University Press. Luis, I and H. Jacome, (2004), ‘‘The late 1990s ﬁnancial crisis in Ecuador: institutional weaknesses, ﬁscal rigidities, and ﬁnancial dollarization at work’’, IMF Working Paper, No. 04/12. McKinnon, R.I. (1963), ‘‘Optimum currency areas’’, American Economic Review, Vol. 53, pp. 717–725. Mundell, R.A. (1961), ‘‘A theory of optimum currency areas’’, American Economic Review, Vol. 51, pp. 657–665. Nitsch, V. (2004), Have a break, have a y national currency: when do monetary unions fall apart?’’, CESifo Working Paper, No. 1113. Rose, A.K. and C. Engel (2002), ‘‘Currency unions and international integration’’, Journal of Money, Credit and Banking, Vol. 34(August), pp. 804–826. Wyplosz, C. (1997), ‘‘EMU: why and how it might happen’’, Journal of Economic Perspectives, Vol. 11(Fall), pp. 3–23.

This page intentionally left blank

Comment Sarantis Kalyvitis

The European Monetary Union (EMU) is probably the most impressive achievement of European uniﬁcation. This astounding performance suggests that there is much scope for investigating the reasoning behind monetary uniﬁcation and its impact in an empirical context. Volker Nitsch attempts to sort out empirically the main determinants of successful monetary uniﬁcation. This is an admirable and difﬁcult task, since such fundamental reforms are very rare. Hence, any inference about the effect of monetary uniﬁcation must be identiﬁed from the cross-country variation, which raises a number of statistical issues. A main challenge is that monetary uniﬁcation is not random: countries with different regimes also differ in many other respects. Hence, how can the effect of the uniﬁcation be identiﬁed separately from that of other observable and nonobservable policy determinants? Leaving this aside for the time being, the starting point of the argument posed by Nitsch is that according to the theory of optimum currency areas (OCAs), monetary uniﬁcations is associated with both costs and beneﬁts. This is a well-known fact since the original analysis of OCAs in the 1960s. The current paper attempts to measure empirically the impact of monetary uniﬁcation by assessing the pre- and post-accession growth impact in the few cases where monetary uniﬁcation has taken place. Given this idea, the author constructs some empirical models and argues that the monetary uniﬁcation effects can be inferred. The author estimates these models as linear regressions with the proposed empirical speciﬁcations given by: Post-union performance ¼ f(constant, pre-entry convergence) In this model, performance plays the role of an effectiveness index and is captured by the post-union growth rate of the entrant country or, alternatively, by the corresponding growth rate differential between the entrant and existing union members. To tackle the issue of the determinant of performance, the author relies on the standard theoretical background and uses variables given by CONTRIBUTIONS TO ECONOMIC ANALYSIS VOLUME 279 ISSN: 0573-8555 DOI:10.1016/S0573-8555(06)79017-4

r 2007 ELSEVIER B.V. ALL RIGHTS RESERVED

286

Sarantis Kalyvitis

the theory of OCAs: along this line, the bilateral trade integration and the business cycle synchronisation are used, with little success though. Next, the author tests whether other variables affect the success of uniﬁcation: here, the inﬂation rate, the inﬂation differential, inﬂation variability, ﬁscal measures, and other variables are used, with limited success. Four critical ingredients are necessary to reach the conclusions mentioned above, most of which are mentioned explicitly in the paper: the heterogeneity in the dataset, the absence of control variables, the linear speciﬁcation imposed, and the potential endogeneity between the variables at hand. Each of these links merits attention and will be brieﬂy discussed here. 1. Data heterogeneity As the author points out , he has to rely on a variety of data to perform this exercise. A quick look at Table 1 reveals that there are data on entrants such as Guinea-Bissau and Mali, and also on G-7 countries (France and Italy). Obviously, this creates a great deal of discussion for the assessment of empirical results, as these countries have entered monetary unions under totally different initial conditions (also pointed out by the large standard deviations in Table 2). A natural consequence is the extreme sensitivity to outliers, which renders the results in tables 6 and 7 hard to interpret. 2. Control variables In this sort of empirical exercise, one would obviously want to hold constant a number of variables likely to shape the performance of a country. Here, the growth literature provides some standard covariates. Control variables will include country characteristics, such as the initial level of per capita GDP to control for standard convergence arguments, openness to account for the degree of exposure to shocks from abroad, regional dummies that typically capture geographical and historical characteristics of the countries examined, population size to capture possible scale effects, as well as measures of ethnic conﬂict and political instability to capture the ethno-linguistic fractionalization of groups within a country, which is correlated with bad policies and low growth, and political turmoil to capture civil unrest, which is also believed to affect growth. Finally, colonial history (British, Spanish-Portuguese, or other colonial powers) is also likely to matter as a determinant of successful monetary uniﬁcation. 3. Linearity By the assumption of linearity, the impact of monetary uniﬁcation is fully captured by the coefﬁcient of the regressor via a constant coefﬁcient. This is a very strong assumption: linearity is usually taken as a convenient local

Comment

287

approximation of a more general model. But, as mentioned earlier, the comparison takes place here between very different groups of countries. 4. Conditional Independence Given the non-random occurrence of monetary uniﬁcation, conditional independence is another strong assumption, as historical variables determining the current regime could have also inﬂuenced policy outcomes. This is not a problem if all the common historical determinants of policy outcomes and monetary regime appear in the empirical relationship and the model is linear. (This is another reason why, when estimating this type of models by Ordinary Least Squares (OLS), the standard set of control variables should always be added to the regressors.) But if there are some omitted determinants of policy outcomes, which are correlated with the process of monetary uniﬁcation, conditional independence is violated and the OLS estimates are biased. 5. A suggested remedy: the average treatment effect Another setup to assess the impact of monetary uniﬁcation can be given by considering the following model, where the process of monetary uniﬁcation (say Xi) in country i takes two values1: ( 1; if GðZi Þ þ ei 40 Xi ¼ 0; otherwise where Zi is a vector of observable variables and ei is an error term. Now, if there is an impact of monetary uniﬁcation on the performance of an economic variable Yi in country i, one can write Y i ¼ F ðX i ; Wi Þ þ ui where Wi is a vector of control variables (which can be overlapping with Zi) and ui is an error term. Note that the set of observable variables can be determined by the list given above. Here, the aim is to estimate the average effect of a monetary uniﬁcation reform, say a shift from Xi ¼ 0 to Xi ¼ 1, on the country performance. This is a well-known approach from the labour economics literature and is called the average treatment effect. An advantage of this approach is that, in general, one can increase the number of observations considerably, as countries that have not gone through a monetary uniﬁcation process can serve as the control group, thus increasing the sample size considerably.

1

See also Persson and Tabellini (2004).

288

Sarantis Kalyvitis

More important, in the above context, one can relax the assumptions of linearity and conditional independence. There are two alternative ways to tackle with linearity (and indirectly with data heterogeneity) in this model: (i) by deﬁning a number of interaction terms and estimate via OLS or another method – but this poses the problem of a very large number of possible combinations, and (ii) by relying on the use non-parametric methods, which give more weight to comparisons of similar countries and thus reduce the effects of potential nonlinearities. Regarding conditional independence, an Instrumental Variables estimator can be used, where the two equations should be viewed as a two-stage approach and estimated explicitly. Again, historical variables can be considered that have some relationship with monetary uniﬁcation, but are uncorrelated to the unobservable determinants of policy outcomes ui. A straightforward extension within the above context, which would highlight more aspects of the monetary uniﬁcation impact, might be the investigation of similar effects on more variables. These could be separated into real variables (such as the per capita growth rate, real wages, measures of ﬁscal stance, e.g., domestic credit to public sector as percentage of total credit ) and nominal variables (such as inﬂation and inﬂation variability ). It would be interesting to see whether all these variables react similarly to a monetary uniﬁcation, or if there are marked differences in their behaviour.2 To sum up, the paper by Nitsch opens a debate that will gain importance in the years to come. This is particularly important from a European point of view, as a number of countries are expected to join the EMU, and there is keen interest among both academics and policymakers in the likely outcome of this major regime change. References Edwards, S. (1989), Real Exchange Rates, Devaluation, and Adjustment, Cambridge, MA: MIT Press. Persson, T. and G. Tabellini (2004), ‘‘Constitutional rules and ﬁscal policy outcomes’’, American Economic Review, Vol. 94(1), pp. 25–45.

2

Edwards (1989) is an example of an attempt to measure the impact of a policy measure (devaluation) on a number of real and nominal variables.

Subject Index adjustment mechanism 4, 232 Amsterdam Treaty 5, 36, 41, 61, 71, 72, 101n1, 167 Balassa–Samuelson effect 22, 236 capital mobility 66, 67, 69, 100, 104, 115 common agricultural policy (CAP) 14, 15, 40, 51, 74, 103, 113, 114, 116, 149 competition policy 75, 76, 100, 133, 201n14 convergence 263–266, 268, 271, 272, 277, 281 cooperation intergovernmental 16, 132, 133, 137, 161, 162 currency union 263–282 delayed integration 13, 108, 110, 116, 122–127 economic and ﬁnancial committee (EFC) 170, 171, 195, 206–210, 212, 215, 223–225 enhanced cooperation agreements 10, 60, 67, 69, 85, 87 euro area 4, 16–22, 24, 25, 101, 130, 169, 182, 183, 201, 231–235, 237, 239–242, 244, 265 candidates 231 exchange rate 241, 244 inﬂation differentials 231, 232, 235, 237, 240, 241, 244 inﬂation persistence 233 inﬂation rates 22, 231

European Central Bank (ECB) 3, 4, 17, 40, 64, 117, 168, 186, 204, 232 functions 17, 211 European Constitution or Constitutional Treaty and economic issues 6, 7, 74 and democratic decision making 7, 34 and enhanced cooperation agreements 10, 60, 66, 69, 85, 87 and parliamentary democracy 47, 48, 50, 55, 135 and presidential democracy 47–50, 55, 135 and public services 7 and qualiﬁed majority voting 37–39, 46, 51 and social policy 115 and the budget 35, 39 and the budgetary procedure 129–131, 134–137, 143–150, 152, 155 and the Court of Justice 38, 40, 45, 60, 185 and the European Commission 38 and the European Council 6, 33, 37–39, 42–47, 59 and the European Parliament 34, 36, 44 and the future of Europe 25, 33 and the president of the Commission 36, 42–45, 47 and the president of the Council 37, 38, 42–44 and the separation of powers 48 as a ‘‘veil of ignorance’’ 8, 54 as an incomplete contract 8, 33, 64 parts of 3–40 ratiﬁcation 3, 4, 38, 41, 47 European Council 33–35, 37–39, 44–47, 133, 138, 144, 147, 148, 170, 171

290

Subject Index

European Economic and Monetary Union (EMU) 16, 100, 167, 191, 224, 231, 232, 263–265, 267, 268, 285 European integration 7–10, 16, 25, 33, 41, 42, 51, 73, 100 of factor and product markets 99, 115 in supervising ﬁnancial markets 195, 196, 201, 206 European Parliament 5, 6, 18, 34, 37–40, 42–45, 47, 50, 57, 62, 74, 129, 133, 135, 137, 140, 141, 143–150, 152–154, 182, 184, 186, 187, 206 European Union (EU) accountability 55, 133, 135, 164 achievements 3 and allocation of powers 8, 35, 75, 88, 153 and its democratic deﬁcit 11 as a functional organization 7 as a political community 7 budget 9, 14–16, 113, 126, 129, 130, 135, 140–144, 147 budgetary procedure 4, 15, 129–131, 135–137, 142, 145, 148, 151, 152, 157 centralization and decentralization of functions 11, 12, 64, 66–71, 77–81, 85, 88 Convention for the future of Europe or Convention 5, 33–35 economic policy 17, 81, 96, 167 efﬁciency of decision making 8, 10, 25, 51, 59, 63, 103, 132, 134, 140 ﬁnancial perspective (FP) 143–149, 152, 154, 162 governance 6, 59, 65 institutional design 8, 16, 58n2 institutional reform 100, 130, 134, 153 legitimacy 5, 130–135, 137, 153, 161–163 security markets 19 single market 21, 42, 100–102, 109, 133 supranational governance 16, 132, 133, 137, 161, 162, 164 exchange rates changes 232, 237, 240, 251 regimes 267, 276, 280 stability 24, 267, 275 externalities 10–12, 16, 17, 63, 72, 79, 99, 109, 114, 125, 162, 192, 197

ﬁnancial supervision 16, 18, 201, 202, 205 ﬁscal discipline 17, 98, 101, 134, 167, 168, 267 ﬁscal federalism 9, 47, 58, 75, 76, 80, 85, 89, 132, 164 ﬁscal policy 16, 24, 25, 99, 102, 104, 106, 167–169, 172, 174, 175, 181–183, 187, 189–192, 259 ﬁscal transfers 9, 105, 266 ﬁscal year 145, 150, 151 inﬂation adjustment 54, 256 differences 4, 21, 22, 236, 237, 239, 260, 286 gap 236 persistence 22, 232, 233, 244, 245, 251, 255, 256, 260 targeting 256 variability 275, 276, 286, 288 labor market policy 92, 102, 105, 107, 117 Lamfalussy model of ﬁnancial supervision 19, 206, 209, 210, 225, 226 lobbying 50, 76–81, 85, 89 Maastricht Treaty 5, 41, 167, 172, 173, 184, 264, 265, 276, 281 criteria 22–24, 264, 267, 275 migration 12, 13, 94, 96, 97, 105–111, 115, 122, 124–127 monetary policy 20–22, 58, 100, 101, 168, 177, 210–212, 223, 232–234, 244, 245, 251, 255–257, 267 Monetary Union 4, 60, 64, 73, 167, 175, 183, 256, 263–268, 271, 272, 274, 275, 281 Nice Treaty 5, 6, 10, 11, 34, 35, 41, 59–61, 63, 66, 67, 71–74 optimum currency area (OCA) 21–24, 232, 266, 267, 272–274, 285 output gap 241–243, 245, 248, 251, 256, 260 policy coordination 9, 10, 25, 198, 169 policy harmonization 11, 13, 65–68, 70–72

Subject Index

291

Qualiﬁed majority voting (QMV) 6, 10n7, 46, 51, 61, 133, 145–147, 154, 162

systemic risk 18, 196–198, 200, 202, 209, 211, 214, 215, 223–227

social insurance 14, 102, 111, 112, 122, 124, 142 social policy 12, 91, 92, 96, 100, 101, 104–107, 109, 114–116, 121–126 stability and growth pact (SGP) 16–18, 25, 167, 168, 170–174, 177, 182, 188, 191–193 structural and cohesion funds (SCFs) 14, 139 subsidiarity principle 35, 36, 122, 123, 181 sustainability council 18, 167–191, 193

trade and convergence 21 and growth 24 and integration 266, 267, 272 transfers 8, 11, 67, 70–74, 77, 87, 102, 103, 105, 106, 108, 112, 113, 122, 125, 126, 141, 143, 180, 181, 266 welfare policy 12, 92–96, 126 externalities 109, 114

This page intentionally left blank

Cognitive Economics: New Trends, Volume 280 (Contributions to Economic Analysis) (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Fighting Corruption in the Public Sector, Volume 284 (Contributions to Economic Analysis) (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Designing the New European Union, Volume 279 (Contributions to Economic Analysis) (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Cognitive Economics: New Trends, Volume 280 (Contributions to Economic Analysis) (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Fighting Corruption in the Public Sector, Volume 284 (Contributions to Economic Analysis) (Contributions to Economic Analysis) (Contributions to Economic Analysis)

New Directions in Macromodelling, Volume 269 (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Explaining Growth in the Middle East, Volume 278 (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Tourism, Trade and National Welfare, Volume 265 (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Urban Dynamics and Growth: Advances in Urban Economics, Volume 266 (Contributions to Economic Analysis) (Contributions to Economic Analysis)

The Single Server Queue (Contributions to Economic Analysis)

Contributions to non-standard analysis

European Union and the Euro Revolution, Volume 283 (Contributions to Economic Analysis)

Introduction to Economic Analysis

Introduction to Economic Analysis

Unfunded Pension Systems: Ageing and Migration, Volume 264 (Contributions to Economic Analysis)

Advances in Psychology Volume 52 - Human Mental Workload (Contributions to Economic Analysis)

Regionalisation, Growth, and Economic Integration (Contributions to Economics)

Using Economic Indicators to Improve Investment Analysis

Using Economic Indicators to Improve Investment Analysis

Using economic indicators to improve investment analysis

Mondragon: an economic analysis

Development of Economic Analysis

Development of Economic Analysis

Analysis of Economic Data

Mondragon: An Economic Analysis

Engineering economic analysis

Engineering economic analysis

Development of Economic Analysis

History of Economic Analysis

The Decision to Patent (Contributions to Economics)

Contributions To All Year Round

Contributions to Simulation Speed-up

Selected Contributions in Data Analysis and Classification

Designing the New European Union, Volume 279 (Contributions to Economic Analysis) (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Cognitive Economics: New Trends, Volume 280 (Contributions to Economic Analysis) (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Fighting Corruption in the Public Sector, Volume 284 (Contributions to Economic Analysis) (Contributions to Economic Analysis) (Contributions to Economic Analysis)

New Directions in Macromodelling, Volume 269 (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Explaining Growth in the Middle East, Volume 278 (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Tourism, Trade and National Welfare, Volume 265 (Contributions to Economic Analysis) (Contributions to Economic Analysis)

Urban Dynamics and Growth: Advances in Urban Economics, Volume 266 (Contributions to Economic Analysis) (Contributions to Economic Analysis)

The Single Server Queue (Contributions to Economic Analysis)

Contributions to non-standard analysis

European Union and the Euro Revolution, Volume 283 (Contributions to Economic Analysis)

Introduction to Economic Analysis

Introduction to Economic Analysis

Unfunded Pension Systems: Ageing and Migration, Volume 264 (Contributions to Economic Analysis)

Advances in Psychology Volume 52 - Human Mental Workload (Contributions to Economic Analysis)

Regionalisation, Growth, and Economic Integration (Contributions to Economics)

Using Economic Indicators to Improve Investment Analysis

Using Economic Indicators to Improve Investment Analysis

Using economic indicators to improve investment analysis

Mondragon: an economic analysis

Development of Economic Analysis

Development of Economic Analysis

Analysis of Economic Data

Mondragon: An Economic Analysis

Engineering economic analysis

Engineering economic analysis

Development of Economic Analysis

History of Economic Analysis

The Decision to Patent (Contributions to Economics)

Contributions To All Year Round

Contributions to Simulation Speed-up

Selected Contributions in Data Analysis and Classification

Recommend Documents