Political Representation, and Legitimacy in the European Union

POLITICAL REPRESENTATION AND LEGITIMACY IN THE EUROPEAN UNION

This page intentionally left blank

POLITICAL REPRESENTATION

AND LEGITIMACY IN THE EUROPEAN UNION

Edited by

HERMANN SCHMITT and

JACQUES THOMASSEN

OXFORD UNIVERSITY PRESS

This book has been printed digitally and produced In a standard specification in order to ensure its continuing availability

OXFORD UNIVERSITY PRESS

Great Clarendon Street, Oxford OX2 6DP Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan South Korea Poland Portugal Singapore Switzerland Thailand Turkey Ukraine Vietnam

Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © the several contributors 1999 The moral rights of the author have been asserted Database right Oxford University Press (maker) Reprinted 2007 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, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover And you must impose this same condition on any acquirer ISBN 978-0-19-829661-4

Preface and Acknowledgements THIS book reports on the results of the European Election Study 1994. The European Election Studies (EES) is a long-term project that came into existence in anticipation of the first direct elections for the European Parliament in 1979, Ever since, more or less the same international group of scholars has been responsible for successive European Election Studies on the occasion of all but one election to the European Parliament. The empirical basis of each of these studies consisted of a survey among a random sample of the electorate in each of the member-states immediately after the election. These surveys were conducted in co-operation with the Eurobarometer. The objectives of this project always went beyond the study of electoral behaviour, as is aptly demonstrated by its last major publication, Choosing Europe, Edited by Cees van der Eijk and Mark Franklin, this book addresses two of the most pressing problems of the European Union—its legitimacy and the quality of its democratic system. They conclude that elections to the European Parliament fail as a link between the will of the European people and European public policy, mainly because both political parties and voters are prevented from playing the role expected of them in a properly-functioning representative democracy. The growing ambitions of the project rendered the original study design gradually less satisfactory. The critical reactions of mass publics in many countries to the Maastricht Treaty gave additional reason to wonder whether represented and representatives were still living in the same European world. As any true representation study is based on surveys of both voters and political elites, the 1994 European Election Study was designed to include not only a voter study, but also a survey of the candidates for the European Parliament and a survey of the elected members of the European Parliament, In order to get an accurate picture of the legitimacy and functioning of the system of political representation in the European Union, surveys of members of national parliaments were an indispensable additional part of the design. This volume will focus on the relationship between voters and their representatives, primarily at the European level, but at the national

vi

Preface and Acknowledgements

level as well. The relationship between the European Parliament and national parliaments is the subject of a companion volume published jointly with this volume, The European Parliament, the National Parliaments, and European Integration, edited by Richard S, Katz and Bernhard Wessels. A project like this is more than just another research project. It is an enterprise involving a great number of people over an extended period of time. The voters' study was prepared some five years ago by a core group of researchers consisting of Pilar del Castillo (Madrid), Roland Cayrol (Paris), Cees van der Eijk (Amsterdam), Mark Franklin (Houston), Renato Mannheimer (Genoa), Michael Marsh (Dublin), Karlheinz Reif (Brussels), and Colette Ysmal (Paris). This group was coordinated by Hermann Schmitt and supplemented by a number of country specialists; Mario Baccalhau (Lisbon), Michael Guilljam (Goteborg), Ilias Nikolakopuolos (Athens), Lieven De Winter (Louvain), and Torben Worre (Copenhagen). The voters' study was carried out in the then twelve member-countries of the European Union as part of a special post-electoral Eurobarometer survey (EB 41.1). This was made possible by the energetic support of one of the pioneers of research into European Parliament elections, and the then director of Eurobarometer surveys, Karlheinz Reif. The study was funded by the Deutsche Forschitngsgemeinschaft and the Nederlandse Organisatie voor Wetenschappelijk Onderzoek, Additional support was provided by the University of Mannheim (Mannheimer Zentrum fur Europaische Sozialforschung, MZES), which served as the co-ordination centre of the study and funded a pre-test survey conducted by ZUMA in the Mannheim area, among other things. The three components of the elite study were prepared by a research group involving Richard Katz (Baltimore, MD), Pippa Norris (Cambridge, MA), Jacques Thomassen (Enschede), and Bernhard Wessels (Berlin). The European Candidates Study was co-ordinated by Jacques Thomassen. Financial support (for co-ordination, printing of questionnaires, and data processing) was provided by the Faculty of Public Administration and Public Policy of the University of Twente. Without the indefatigable help of Peter Geurts and Henk van der Kolk the management of production, mailing, and processing of thousands of questionnaires in nine different languages would have been impossible. This part of the project would also have been impossible without the generous co-operation of national study directors in each country and

Preface and Acknowledgements

vii

the financial support of their home institutions for the translation of questionnaires and the actual fieldwork. They were Tom Bryder (Copenhagen), Cecile Chavel (Paris), Irene Delgado and Lourdes Lopez Nieto (Madrid), Louisa Gardella (Genoa), Peter Linen (Edinburgh), Michael Marsh (Dublin), Maria Jose Stock (Lisbon), and Lieven De Winter (Louvain). The study among the members of the European Parliament was coordinated by Bernhard Wessels and funded by the Deutsche Forschungsgemeinschaft. The study among the members of the national parliaments was co-ordinated by Bernhard Wessels and Jacques Thomassen. Funds for this part of the study were gathered together from several sources, including the universities of Mannheim and Twente and the Wissenschaftszentrum Berlin, But again, this part of the study would not have been possible without the great efforts of our colleagues in the different countries. They translated the questionnaires, did the fieldwork, and, in most cases, raised the necessary funds as well. We are most grateful for the support of Irene Delgado and Lourdes Lopez Nieto (Madrid), Louisa Gardella (Genoa), Soren Holmberg, Martin Brothen, and Peter Esaiasson (Goteborg), Manina Kakepati (Athens), Christina Leston Bandeira (Lisbon), Michael Marsh (Dublin), Bernhard Wessels and Achim Kielhorn (Berlin), Lieven De Winter and Patrick Dumont (Louvain), and Colette Ysmal (Paris). The Wissenschaftszentrum Berlin became the nerve centre of both parliamentary studies. Bernhard Wessels and his capable staff did a splendid job organizing the fieldwork, and processing and distributing the data. This project brought us to a good number of conferences in Europe and beyond where we discussed the design of the project and, later, chapter drafts for a special issue of the EJPR (vol. 32, no. 2) and for this and the companion volume. Conferences were hosted and funded by the CEVIPOF (Paris), the Feltrinelli Foundation (Milan), the Johns Hopkins University (Florence), the Center for European Studies of Harvard University, and the Universities of Madrid, Mannheim, San Sebastian, and Twente. Two conferences (in Doom and Enschede) were funded by the Nederlandse Organisatie voor Wetenschappelijk Gnderzoek. The Mannheimer Zentrwn fir Europäische Sozialforschung has been extremely generous, granting Jacques Thoniassen a visiting professorship during the winter of 1997-8.

viii

Preface and Acknowledgements

We had excellent research assistants in Michael Klein, Henk van der Kolk, Angelika Scheuer, Evi Scholz, and Erik Voeten. Gretchen Wiesehan provided the language and copy editing and, together with Christoph Geppert, help with the index; Stefan Solleder produced the camera-ready copy. It is beyond any doubt that without the capable support and unfailing engagement of all these people, this book would never have seen the light of day. In Dominic Byatt and Amanda Watkins of Oxford University Press we found understanding and helpful editors. They confronted us with an impossible deadline that we ourselves had asked for. This final dose of pressure was invaluable for getting the book finished. Last but not least, our thanks are due to the members of the European Parliament, the members of national parliaments, the candidates running in the 1994 European Parliament elections, and the thousands of European citizens who were willing to participate in our surveys. Looking back, finally, over more than half a decade of close cooperation with so many people from different cultural backgrounds, with different interests and capabilities (and, not to forget, with large egos), it is hard to think of a better way to find out how difficult a process European integration is. At the same time, if the heart-warming willingness to work together and the lasting friendships that have developed over the years are symptomatic for more general developments in Europe, there is every reason to believe in the prospects of an evercloser union. Hermann Schmitt Mannheim January 1999

Jacques Thomassen Enschede

Contents List of Figures

xi

List of Tables

xiii

List of Contributors

XV

Introduction I . Introduction: Political Representation and Legitimacy in the European Union Jacques Thomassen and Hermann Schmitt

3

Part I: Legitimacy 2. A Political Community? Angelika Scheuer 3,

The Scope of EU Government Lieven De Winter and Marc Swyngedovw

25 47

4. The Political Regime Pippct Norris

74

5. Policy Performance Michael Marsh

90

Part II: Representation 6. Distinctiveness and Cohesion of Parties Hermann Schmitt and Jacques Thomassen

111

x

Contents

1. The Cognitive Basis of Voting Wovter van der Brug and Cees van der Eijk 8,

9.

129

Policy Preferences and Party Choice Cees van der Eijk, Mark N. Franklin, and W outer van der Brug

161

Issue Congruence Jacques Thomassen and Hermann Schmitt

186

10. Whom to Represent? Role Orientations of Legislators in Europe Bernhard Wesseh

209

1 1 . Wishful Thinking Among European Parliamentarians Sören Holmberg

235

Conclusion 12. In Conclusion: Political Representation and Legitimacy in the European Union Jacques Thomassen and Hermann Schmitt

255

Appendix: The European Representation Study

269

ReferenAces

275

Index

297

List of Figures

1.1.

Radical view of representative democracy Intergovernmental model of European political representation 1,3. The federal model of European political representation 1.4. The study design of the European Representation Study 2.1. Concepts, indicators, and data sources 2.2, National or European identifications? 4.1. The development of support for EU principles 5.1. Relative benefit from EU membership: original six members 5.2. Relative benefit from EU membership; DK, IRL, UK 5.3. Relative benefit from EU membership: OR, E, P, A, S, SF 5.4. Predicted and actual policy satisfaction, 1983-96 6.1. Types of issues analysed 6.2. Left-right orientations of citizens and placements of elites 6.3. Unemployment concerns of citizens and emphasis of elites 6.4. Currency preferences of citizens, positions of party elites 7.1. Left—right: perceptions and 'true* positions 7,2. Borders; perceptions and 'true' positions 7.3. Currency: perceptions and 'true' positions 7.4. Employment programme: perceptions and 'true' positions 9,1. Voters* and MEPs' positions: left-right 9,2. Voters' and MEPs' positions: employment 9.3. Voters' and MEPs' positions: borders 9.4. Voters' and MEPs' positions: currency 9.5. Left-right congruence of voters and MEPs 9.6. Policy congruence of voters and MEPs: employment 9.7. Policy congruence of voters and MEPs: currency 9.8. Policy congruence of voters and MEPs: borders 9.9. Policy congruence of MEPs sad MNPs: currency 9.10. National congruence of MEPs and voters: currency 10.1, Role conceptions: representational foci 10.2. Electoral competition and constituency focus of MPs in EU

15

1.2.

18 18 21 31 35 78 96 97 98 105 117 119 120 121 146 147 148 150 193 193 194 194 200 200 201 201 203 205 214 225

xii

List of Figures

10.3. Membership experience, power, and European focus 1 1 . 1 . Currency: MEPs' positions compared to

voters' perceptions 11.2. Currency: MEPs* attitudes compared to mass public's 11.3. Members* attitudes and perceptions and voters' attitudes

232 239 240 243

List of Tables 2.1. 2.2. 2.3. 2.4. 2.5. 2.6. 3.1. 3.2, 3,3. 3.4. 4.1. 4.2. 4.3. 4.4. 4.5. 4.6. 5,1. 5.2. 6.1. 6.2. 6.3. 6.4. 7.1.

Feeling like a European citizen and thinking of oneself Pride in being European Mutual trust Trust in the people of the other EU member-countries Citizens' views about new members Political elites' views about new members Perceived and preferred levels of decision-making Perceptions of importance of policy sectors Most important policy sectors and preferred levels Logistic regression on preference for European level Dimensions of EU regime support Support for European principles Satisfaction with democracy Confidence in parliaments Confidence in government institutions Models of support for EU regime processes Regression estimates of EU policy satisfaction Regression estimates of EU policy satisfaction The group structure of the European Parliament Main party distinctiveness Overall distinctiveness Main party cohesion Respondents not locating parties on various issue dimensions 7.2. Perceptual agreement about party positions and competence 7.3. Regression: perceived positions and true positions 7.4. Correlations between positions on policy dimensions 7.A.1 , Overview of how voter and elite data are linked 8.1, Respondents without preferences for various issue areas 8.2. Respondents supporting party with highest vote probability Explained variance in party preferences 8.3. R2s of seven raultivariate models to predict party choice 8.4.

33 37 39 40 41 43 55 58 59 64 76 79 80 81 82 87 101 104 114 122 123 125

134 138 145 152 156 164 168 174 176

xiv

8.5. 9.1. 9.2. 9.3. 10.1. 10.2. 10.3. 10.4. 10.5. 10.6. 11.1. 11.2. 11.3.

A.I.

A.2. A3

List of Tables Regression coefficients: three most encompassing models Issue priorities of EU citizens Issue priorities compared: EU citizens, MEPs, and MNPs Requirements of the Responsible Party Model in the EU Distribution of representational foci among MEPs and MNPs Individual-level explanation of representational focus Competition, uncertainty, and representational focus Encompassiveness and special group focus of representation Inclusiveness of interest group systems and special group focus Country size, membership experience, and European focus Distance to voters: true, perceived, and misperceived Issue agreement and perceptual accuracy MEPs'/MNPs' perceptions of voter opinion: tested models EES94: fieldwork period, number of interviews, and return rate ECS94: number of interviews and return rates MEP Study and MNP Studies

179 190 191 197 216 219 224 228 229 231 246 247 248 270 271 273

List of Contributors LIEVEN DE WINTER, Professor at the Catholic University of Louvain and at the Catholic University of Brussels, Belgium. MAWCN. FRANKLIN, Reitemeyer Professor of Political Science, Trinity College, Hartford, Connecticut, USA. SOREN HOLMBERG, Professor of Political Science, Faculty of Social Sciences, Gdteborg University, Sweden, MICHAEL MARSH, Senior Lecturer, and Head of Department of Political Science, Trinity College, Ireland. PlPPA NORMS, Associate Dkector of the Joan Shorenstein Center on the Press, Politics and Public Policy and Lecturer at the Kennedy School of Government, Harvard University, USA. ANGELIKA SCHEUER, TMR Research Scholar, Faculty of Social and Behavioural Sciences, University of Amsterdam, The Netherlands. HERMANN SCHMITT, Senior Research Fellow at the Mannheimer Zentram fur Europalsche Sozialforschung, University of Mannheim, Germany. MARC SWYNGEDOUW, Professor of Political Science and Sociology, Catholic University of Leuven and Brussels, Belgium. JACQUES THOMASSEN, Professor of Political Science, Faculty of Public Administration and Public Policy, University of Twente, The Netherlands. VAN DEE BRUG, WOUTER, Assistant Professor, Faculty of Social and Behavioural Sciences, University of Amsterdam, The Netherlands. VAN DER EUK, CEES, Professor of Political Science, Faculty of Social and Behavioural Sciences, University of Amsterdam, The Netherlands. BERNHARD WESSELS, Senior Fellow at the Wissenschaftszentrum Berlin ffir Sozialforschung (WZB), Germany.

This page intentionally left blank

INTRODUCTION

This page intentionally left blank

1 Political Representation and Legitimacy in the European Union

JACQUES THOMASSEN AND HERMANN SCHMITT

INTRODUCTION

A major concern of the European Union and its predecessors has been the legitimacy and democratic quality of its institutions. In Monnet's vision one of the major tasks of the European project was '...to ensure that in their limited field the new institutions were thoroughly democratic' (as quoted in Holland 1993: 145). At the end of the twentieth century many Europeans might still share Monnet's ideals, but hardly anyone will have the illusion that the task defined by Monnet has been accomplished. Certainly, comparing the present Europe with the early days of the European communities it is hard to argue that the European project has not been a success. The common market is irreversible, and the inner borders of the Union hardly exist anymore. Binding European legislation is such that it is foreseeable-—if not already the case—that European legislation will overtake the national legislation of the member-states in importance. For this very reason it is more and more felt to be a problem that the democratization of the Union might not have kept pace with this progress. Whereas the supranational competencies of the EU have been growing, an effective system of political representation is said to be missing. Since the Single European Act opened the possibility for the

4

Introduction

Council of Ministers to decide by majority vote instead of unanimity, the national parliaments have lost part of their power to scrutinize and control the positions taken by (ministers of) their national governments in the European arena. This loss of power was not fully compensated by an extension of the power of the European Parliament, hence the famous democratic deficit. Moreover, it is argued that the system of political representation is inadequate not only because of the lack of a system of responsible government, but also because elections for the European Parliament fail for a number of reasons to link the views of the European people on political matters to the agenda of the European Parliament, let alone to the policy-making of the European Union (cf. van der Eijk and Franklin 1996). Despite these apparent democratic shortcomings it looked for a long time as if the European project could count on the permissive consensus of the European people. However, since the signing of the Maastricht Treaty and its aftermath, it is hard to maintain this assumption. One might wonder whether the governments and politicians responsible for the Maastricht Treaty were living in the same European world as the people they were supposed to represent. A possible lack of agreement between political elites and the mass public on the future of Europe might reveal not only an ineffective system of political representation but also a lack of legitimacy of the institutions of the EU. It is these questions we will address in this book. For the time being we will define political representation loosely as referring to the concept of representative democracy, including both the institutions of responsible government and the process of political representation, i.e. the political process by which the making of government policy is related to the wants, needs, and demands of the public. Legitimacy refers to the belief that the existing political order is right. Before we go into both concepts in greater detail, we will first try to explain what we mean by political representation and legitimacy in the European Union. POLITICAL REPRESENTATION AND LEGITIMACY IN

THE EUROPEAN UNION Before addressing these questions more directly, we need to clarify whether it makes sense at all to study the European Union or its prede-

Representation and Legitimacy in the EU

5

cessors as a political entity, over and above the member-states. This is only the case if the European Union is more than the intergovernmental co-operation between independent sovereign states. European integration used to be the undisputed terrain of scholars of international relations. According to one of the leading schools of thought in this field, the European Union is little, if anything, more than the co-operation between sovereign states. However, this might be a matter of perspective. The major competing theories on European integration all agree that they are less interested in the outcome than in the process of European integration, in explaining why the European Union developed as it did (Corbey 1993: 4). Because the 'Constitution' of the European Union is based on treaties between sovereign states, changes in the regime, the rules of the game, or in constitutional affairs can almost by definition only be based on the consent of the member-states. Therefore, it is hardly surprising that decision-making in the process of European integration has been dominated by the member-states. Our approach is different. Being interested in the legitimacy and the quality of political representation in the European Union we are less interested in the political process that led to the present Union than in the outcome of this process, i.e. the characteristics of the Union. Even though national governments are still by far the most important players on the European chessboard and although the European Union is still little more than the co-operation of sovereign states, one can argue that there is this 'something more'. Firstly, ever since the Single European Act came into force the member-states have given up some of their sovereignty by accepting a system of qualified majority voting in some circumstances in the Council of Ministers. This still does not mean that a sovereign European state has suddenly come into being, but that there is rather a pooled sovereignty in those circumstances where the national governments no longer have veto power. In those cases decision-making is removed from the scrutiny and control of the national parliaments. Once governments subscribed to qualified majority voting, national parliaments could not reasonably hold them responsible for being outvoted and accepting the majority decision (Dinan 1994: 289). Secondly, even though the structure and process of the European Union by no means approximate a realistic image of a modern state (Keohane and Hoffmann 1991: 12), 'the output of European governance is like that of a state, even a super state: an endless stream of laws

6

Introduction

in increasingly varied areas of public and private life. They are binding on governments and individuals as part of the law of the land. Indeed, they are a higher law of the land—supreme over conflicting state laws* (Weiler, Haltern, and Mayer 1995: 4). EU law has become a crucial, if not dominant, influence in policy areas as diverse as agriculture, environmental protection, and industrial policy (Pierson and Leibfried 1995: 6), By signing treaties of accession, each member-state accepts that in those areas where the Community has explicit treaty competencies, Community law takes precedence over all forms of national legislation. This legal expression of supranationalism indicates that the Community is 'endowed with independent authority, with its own sovereign rights and a legal order independent of the Member States to which both the Member States and their citizens are subject in matters for which the Community is competent' (European Documentation 1991:9). The Union has gradually assumed considerable authority in policy domains beyond those directly tied to the creation of a common market: The EU is actively engaged in policies of redistribution among both sectors (through the Common Agricultural Policy, or CAP) and regions (through the structural funds). On certain issues, such as gender equality, the Union has established important social regulations such as protections and benefits for women workers or regulations for workplace health and safety which have been set at a very high standard. The EU is also expanding the scope of its authority, although haltingly, in areas such as industrial relations and immigration. Finally and with less fanfare, the Union is becoming a formidable actor in the field of social policy by introducing 'market compatibility requirements* restricting certain social policy options of member-states that conflict with the construction of the single market (Pierson and Leibfried 1995: 3).

Thirdly, the institutional structure of the European Union is a complicated hybrid of intergovernmental and supranational institutions. The European Council and the Council of Ministers are purely intergovernmental, but the European Commission, the European Court of Justice, and the European Parliament are supranational institutions. Once in office the members of these institutions are not accountable to national governments or other national institutions. None of these institutions is devoid of power, even though the Council is the primary legislator of the EU. The European Parliament in particular would be a total anomaly in the institutional framework of the European Union if it

Representation and Legitimacy in the EU

7

could not be seen as part of an institutional structure that goes beyond the co-operation of sovereign states. The evidence that the European Union is this 'something more' than the co-operation between sovereign states is convincing enough. However, none of this is to argue that the European Union in reality is a supranational body rather than the product of intergovernmental cooperation. Once we leave the debate on the explanatory power of competing theories on the process of European integration and move to the characteristics of the Union, this is a senseless argument. A more productive approach for our purposes is the view on the European Union as a multi-tiered system of government (KOnig, Rieger, and Scbmift 1996; Leibfried and Pierson 1995; Weiler, Halter, and Mayer 1995). As Weiler argues, different theoretical approaches emphasize different realities. Irs some crucial spheres Union governance is international; in other spheres it is supranational; in yet others it is infranational. ...There are three approaches, but also three modes of governance. ... In the international mode the focus is on negotiation, intergovernmental bargaining and diplomacy. There is a relatively low level of institutionalization, and a premium on informal and unstructured interaction. Formal sovereign equality (including a formal veto) and the loose reflexes of international law prevail which, of course, should not be understood as leading to full equalization of power among the actors. The materia is often—though clearly not always—constitutional (in a non-technical sense). The modus operandi of the supranational mode is more structured, formal and rule bound. Bargaining and negotiation are far more akin to a domestic legislative process of coalition building, vote counting and rule manipulation. The materia is, frequently, primary legislation. Infranationalism is mostly about regulatory governance and management. There is a medium to low level of institutionalization, and informal networking between 'government' and corporate players abounds. The international mode is characterized typically by high actor visibility and medium to low process visibility. Supranationalism is characterized by medium (aspiring to high!) actor visibility and medium to low process visibility. Infranationalism has both low actor and process visibility (Weiler, Haltem, and Mayer 1995: 24-7).

From the perspective of democratic theory there is nothing wrong with intergovernmental decision-making, as long as both the political process and the political institutions are compatible with this level of decision-making. It would mean that positions taken by national governments in the negotiating process would be subject to the scrutiny and

8

Introduction

control of national parliaments and—indirectly—national electorates. However, as soon as the Union moves into its supranational mode, this is no longer satisfactory. Political scrutiny and political control cannot be effective at a lower level than decision-making. Moreover, the assumption of supranational decision-making is the only legitimization that we can think of for a European Parliament directly elected by a European electorate. Where the Union switches to its supranational mode, it becomes a relevant research question to what extent that mode is legitimate and functioning according to the criteria of normative democratic theory. This is the kind of question we will try to answer in this book. However, the very existence of different modes of governance implies that it would not be very fruitful to study the process of political representation and the legitimacy of the European Union exclusively at the level of the supranational mode without taking into account both the relationship between the national states and the EU and the processes at the national level. Hence, we have deliberately christened this book Political Representation and Legitimacy in the European Union, and not of the European Union. This reflects our belief that questions on political representation and legitimacy in the European Union refer as much to the relation between the national governments and the European Union and to the institutions and political processes at the national level as to the European level as such. Therefore, in this study we will focus on the following topics: - the legitimacy of the European level of government, including the legitimacy of the system of multi-tiered government (particularly the division of power between the European, national, and regional levels); - the democratic quality of the system of political representation at the European level. Part II of the book is devoted to the questions on legitimacy, and Part III to political representation. In this chapter we shall explore the concepts of legitimacy and political representation somewhat further and introduce the research questions to be taken up in the successive chapters of the book.

Representation and Legitimacy in the EU

9

LEGITIMACY Above we have defined legitimacy somewhat loosely as the idea that the existing political order is right. This definition immediately evokes two new questions: - What do we mean by the political order and which aspects of the political order are relevant for its legitimacy? - How do we assess whether a political order is legitimate? Dealing with the second question will bring us automatically back to the first. Basically there are two different ways to assess the legitimacy of a political system. The first way is to refer to normative theory: to what extent does a political system conform to certain normative criteria? The second is to determine to what extent the political system is right in the eyes of the relevant beholders, the members of a particular political system. In this book the emphasis will be on the second approach, and we will rely on the empirical evidence of survey data. We will concentrate on the extent to which both voters and political elites consider the European political system legitimate. However, in this first chapter we will start with the first approach, partly as a useful exercise in itself, but also in order to argue for which aspects of the political system people's attitudes are relevant from the perspective of legitimacy. The relevant aspects can be deduced from normative democratic theory. In order to do that, we go back to the very basic meaning of democracy: government by the people. As everyone knows from his or her first class in democratic theory, democracy stems from the Greek words demos and kratia. Thus the basic questions are: who ought to comprise the people, what does it mean for them 'to rule' and what should they rule on (Dahl 1989: 3). These three questions refer respectively to: - the demos, or the domain of the political system; - the scope of government, or the jurisdiction of the people: what is it that they should rule on; and - the institutions and processes of government.

10

Introduction

If these are the basic elements of a democratic system, we might ask what we can learn about them from democratic theory. The first two elements, the demos and the scope of government, are strongly interrelated. The demos should be understood 'in the political sense of a group of persons who, rightly considered, should govern themselves in a single democratic unit' (Dahl 1989: 3). According to Dahl, there is very little that democratic theory will tell us about the domain and scope of a democracy. Advocates of democracy characteristically presuppose that 'a people" already exists. Democratic principles depend on prior assumptions about the unit: that the unit within which it is to operate is itself legitimate and that the matters on which it is employed properly fall within the jurisdiction of that unit. In other words, whether the scope and domain of majority rule are appropriate in a particular unit depends on assumptions that the majority principle itself can do nothing to justify. The justification for the unit lies beyond the reach of the majority principle and, for that matter, mostly beyond the reach of democratic theory itself (Dahl 1989: 204).

However, (and this is most relevant for the discussion about the European Union), this is not necessarily the final word on the question of domain and scope. There is an obvious trade-off between the size and the scope of democracy. With respect to the domain, or the question of the size of the demos, there is the almost classic temptation to argue the smaller the better. If we accept the view that the greater the weight of each citizen is, the more democratic a system is, then ceteris paribus a larger system is bound to be less democratic than a smaller system. Thus, given a choice, a democrat should always prefer the smaller unit. There is a trade-off between the weight of the vote of each participant in the decision-making process and the weight of the decision itself. It is not a great triumph of democracy when people's votes weigh heavily in a decision while decisions that are really important to them are made elsewhere at a higher level. Taking the European Union as an illustration, people living in a small country like The Netherlands can have a democratic interest in decision-making at a higher level than their own country. If the Dutch have an interest in environmental policy directed at clean air and water, they face the fact that all democratic measures taken by the Dutch government will be of little use if other countries bordering on the main rivers of Europe continue to pollute them and if the air pollution is due at least as much to the industry in the Ruhr as to Dutch industry. In these circumstances having a vote

Representation and Legitimacy in the EU

II

with little weight in the decision-making at a level that can make a difference is more valuable than having a decisive vote in a situation where the demos cannot control the agenda. This must be taken into account when discussing a potential democratic deficit in the European Union; the diagnosis of such a deficit cannot be based only on decision-making rules. From a democratic point of view, decision-making at a level that is incapable of conducting an effective policy with respect to major problems in society might be just as unsatisfactory as political institutions and processes that do not folly conform to democratic standards. The question is not whether effective policy-making can be confined within national borders; it cannot. The real question is whether transnational decision-making can be democratic. Trying to find an optimal balance between the domain and scope of a democratic regime of the EU is beyond the purpose of this study. It would imply the application of an external criterion of the legitimacy of the scope of EU government. Instead, in order to assess the legitimacy of the political system, we will rely exclusively on the second perspective on legitimacy we distinguished: the eyes of the beholders. Still, the two perspectives might be less independent of each other than they seem. Even if we rely on the subjective judgements of the members of the political system as our criterion, we still need to decide for which aspects of the political system their judgements are relevant. From the scheme discussed above, we can deduce that the three most relevant aspects are the domain of the political system, the scope of government, and the institutions and processes of government. The problem of the domain of the European political system can be phrased in the question as to what degree there is a European demos in the hearts and minds of the European people. It touches upon the highly-debated issue of European citizenship. In this debate one can— again-—distinguish two approaches. The first is a formal one and refers to the concept of citizenship in the sense of a system of rights and duties at the European level, as distinct from the national level. However, it is generally recognized that perhaps the most fundamental requirement for a European polity that is more than the co-operation of sovereign states is a sense of collective European identity, a sense of a European community among the people of Europe. From a normative point of view one might argue that there can be no democratic Europe without people's sovereignty; however, one can hardly speak of a sovereign people if there is no sense of identity of a people as a political

12

Introduction

community (Kielmannsegg 1995). From an empirical perspective a sense of community has long been considered a basic element of legitimacy, which in turn is considered a condition for a stable political system (Almond and Verba 1963; Easton 1965), And in both perspectives, finally, it is the subjective feeling of identity that really matters, The extent to which such a European identity has developed is the subject of Chapter 2. The scope of government refers to the question which level in the multi-tiered European governmental system should be mainly responsible for different policy areas: the regional, the national, or the European level. And although it could be argued that the very nature of some problem areas demands a supranational approach (Sinnott 1995). from our perspective on legitimacy the relevant question is again to what extent the European people are willing to allocate such responsibilities to the European Union in those cases where that seems appropriate. This question is taken up in Chapter 3. The third basic element of a democratic system that we distinguished above referred to the institutions and processes of government. This is where most of the traditional critique of the democratic deficit of the European Union applies. This critique obviously involves a normative view of democratic institutions and processes: one needs criteria in order to make a judgement. With few exceptions such criteria are deduced from the idea of democracy defined as government by the people through elected representatives with a parliament as the central political institution (Andersen and Eliassen 1996: 5), The main focus of this criticism is twofold: first, it concentrates on the division of power between European political institutions and in particular on the fact that the European Parliament is not the central European political institution. A second line of criticism refers to the system of political representation in the European Union. To what extent the system and process of political representation meet democratic standards is a matter we will come back to below, as it is the object of Part II! of this book. The question as to what extent the roles of the European and national parliaments in relation to other institutions of the Union meet democratic standards is, of course, an important one. However, it is a question that we will not take up here but pursue in the companion volume of this book (Katz and Wessels 1999a), Our main interest in the institutions and policy processes (Chapter 4) of the Union is once

Representation and Legitimacy in the Elf

13

the extent to which they are considered legitimate in the eyes of the people. The reader might have noticed that the elements of democratic government distinguished above coincide with the different objects of support distinguished by David Easton (1965; see also Niedermayer and Westle 1995), According to Easton, political support is directed at one of the following three political objects: the political community (the group of people bound together by a political division of labour), the political regime (the values, norms, and structures of authority), or the political authorities (the actual incumbents in the roles of authority). He defines legitimacy beliefs as diffuse support for the political regime. Chapters 3 through 5 all refer to the legitimacy of the regime, but to different elements of it. Chapter 3 addresses the scope of government. The main research question is whether and in which areas EU citizens accept policy-making at the European level. This refers to the political regime of the European Union and more particularly to the legitimacy of the division of powers between the institutions of the different layers of government. Chapter 4 refers to the legitimacy of the institutions of the European Union, i.e. the institutions involved in policy-making and the process of decision-making at the European level. We will not consider support for the authorities as an aspect of legitimacy beliefs. It is one of the very characteristics of a democratic system that negative support for the incumbent leaders and their policies should lead to 'throwing the rascals out' rather than adopting a different political regime. In other words, negative support for the authorities and specific policies can be totally consistent with belief in the legitimacy of the regime. However, an empirical question that will be addressed in Chapters 4 and 5 is to what extent there is a spillover from negative support for specific policies and leaders towards the regime.

POLITICAL REPRESENTATION

In Part III of this book we will try to assess the quality of the system of political representation in the European Union, However, for two not unrelated reasons this is not an easy task. Firstly, in order to evaluate

14

Introduction

the quality of the system of political representation in the European Union, we again need criteria for quality. In principle these criteria can be deduced from normative theories of political representation, and, more generally, of representative democracy. However, it is not so easy to single out an undisputed set of criteria for the simple reason that there is not a single undisputed normative theory of political representation. Secondly, theories of political representation have been developed in the context of the nation-state, and it is still to be seen to what extent these apply to the European Union. At the same time this last complication is an interesting one because it will bring us back to some of the classic issues in the literature on political representation. The two classic questions in the literature refer to whom members of parliament should represent and how they should do this. Both questions go back to a debate on these issues in eighteenth-century England, in particular to Edmund Burke's famous speech to the electors of Bristol in 1774. In this speech Burke argued that it was the task of a member of parliament to defend national rather than local interests, and that he should follow his own mature judgement rather than the instructions of his constituents. In the literature these two questions of whom and how to represent became known as the questions of focus and style of political representation (Eulau et al. 1978). This distinction was the main subject of an endless stream of empirical research on the role conceptions of members of representative bodies. Elsewhere (Thomassen 1994: 239) it was argued that the scientific interest in the mandate-independence controversy is inversely proportional to its relevance in a modern representative democracy, in particular in the parliamentary democracies of Western Europe. Burke raised his voice in a discussion that was relevant in the transition from a pre-modem government based on the struggle between a parliament (consisting of representatives of the grievances of estates and communities) and the king, to a modern system of government based on the sovereignty of parliament. The more the parliament became the centre of power, the more responsible it became for the national interest, and the less welcome pure geographical interests became. This view became the general wisdom in Western Europe. Most of the constitutions of European countries after the French Revolution ordained that members of parliament should represent the general interest, and they were forbidden to take instructions from anybody. The liberal revolution of the nineteenth century was followed by what one might call the demo-

Representation and Legitimacy in the EU

15

cratic revolution, which resulted in the introduction of general suffrage. In the course of this development, a system of party government gradually developed which, although anything but undisputed, became the modern doctrine of political representation in the countries of Western Europe, In its democratically most radical form the philosophy behind it can be illustrated in the following simple diagram (Figure 1.1): if (the majority in) the parliament conforms to (the majority of) the electorate, and the government is formed and controlled by (the majority in) the parliament, government policy will be consistent with the will of the (majority of the) people.

Government

Parliament

People

FIG 1.1. Radical view of representative democracy This simple perspective is behind the Responsible Party Model of political representation. According to this model the will of the majority of the electorate will be reflected in government policy as the following requirements are met: 1. There must be different parties with different programmes. 2. The internal cohesion, or party discipline, of political parties must be sufficient to enable them to implement their policy programme.

16

Introduction

3. Voters are required to vote rationally, i.e. to vote for the party whose programme is closest to their own policy preferences. This implies not only that: a) voters do have policy preferences, but also that b) they are aware of the differences between the policy positions of different parties (Thomassen 1994: 251-52). As we said before, this model is anything but undisputed. From a normative view it is criticized as a populist theory of democracy, sacrificing essential democratic values such as the division of power, checks and balances, and the protection of minorities on the altar of majority government (Riker 1982). From an empirical perspective the model is often said to be totally unrealistic. Whether this is the case is a matter we have dealt with elsewhere (Thomassen 1994). The merit of the model is at least that it helps to study the role of different actors in the process of political representation in a systematic way, and that each of the requirements of the model can be used as a benchmark to evaluate the effectiveness of a particular aspect of the system of political representation. However, things start to become far more complicated when we try to evaluate the system of political representation in the European Union. The problem is what kind of a political entity the European Union is, or rather, what kind of political entity it ultimately should become. As we have seen before, the two extreme views on the future of the European Union are to see it: (1) as pure intergovernmental cooperation between sovereign states; or (2) as a federal entity that would take the form of a supranational sovereign European state. It is obvious which system of political representation at the European level goes logically with either ideal type. In the case of a purely intergovernmental system, there is no European people to be represented, but only European peoples. At the European level they would be represented by their governments who would behave as 'a congress of ambassadors from different and hostile interests, which interests each must maintain, as an agent and advocate, against other agents and advocates', to use the words with which Edmund Burke reprimanded such delegates. But in. the case of a purely intergovernmental Europe, he would have had little reason to do so, because there would not be a whole European nation whose interests could be represented. The relevant system of political representation would be at the national level, where the 'normal* system of party government would apply.

Representation and Legitimacy in the EU

17

If there were a supranational Europe, a European state, the same criteria could be applied as to any other national state, and with all its caveats, the Responsible Party Model could be used as a benchmark, although the sheer size of the Union would challenge our imagination even more than in the case of any of the member-states. However, the European Union is neither just a co-operation of sovereign states nor a supranational state. As a multi-tiered system of government it has different layers of government, each with its own system of political representation. Where it moves into its intergovernmental mode, the national systems of political representation apply (Figure 1.2). What is relevant here is to what extent there is an effective system of linkage at the national level with respect to issues of EU government. When EU government moves in its supranational mode, another system of political representation becomes relevant (Figure 1.3). However, it is immediately clear that European reality is far removed from this model. Firstly, it is clear that there is no such thing as a European government in any normal sense of a state government based upon and responsible to the majority in parliament. Moreover, it is unclear whether there is a European people whose sovereignty is embodied in the European Parliament. As a consequence, more than in the case of a sovereign democratic state it can be questioned whom members of the European Parliament should represent. However, the treaties on which the European political institutions are based do suggest that a system of political representation as expressed in the Responsible Party Model is the goal to which the Union should aspire. Therefore, we do believe it is useful and justified to apply the Responsible Party Model as a benchmark to evaluate the effectiveness of the system of political representation at the European level. Article 138a of the Treaty on European Union reads: Political parties at the European level are important as a factor for integration within the Union. They contribute to forming a European awareness and to expressing the political will of the citizens of the Union. It is hard to see how this article should be interpreted, other than from the perspective of a federal Europe. However, the very fact that this perspective is disputed makes it less obvious than in the national context which interests members of the European Parliament should represent: the interests of the European people and the general European interest, or the interests of the people of the country they come from (or—in a Europe of the regions—perhaps of the region they come

18

Introduction

European Council/ Council of Ministers National Government I . . . National Government 15 National Parliament 1... National Parliament 15 National Parties ... National Parties National Electorate 1 ... National Electorate 15

FIG 1.2. The intergovernmental model of European political representation

European Government

European Parliament

European Political Parties

European Electorate

FIG 1.3. The federal model of European political representation

Representation and Legitimacy in the EU

19

from). Therefore, the European Parliament is an extremely interesting arena for the kind of research that never led to much. As far as different views on the essence of the European Union—a supranational level of government or intergovernmental co-operation—clash within the European Parliament as anywhere else, these views might be expressed in different role orientations as to which interests MEPs should represent. The topics we want to deal with in Part III of the book follow logically from the argument so far. In Chapters 6, 7, and 8 we will see to what extent political parties, in particular European party groups, and voters meet the requirements of the Responsible Party Model. These chapters refer mostly to the quality of possible mechanisms of linkage. However, all these mechanisms are intended to bring about a linkage between the policy views of the electorate and the policy positions of the political patty they vote for, and indirectly between public opinion and public policy. Therefore, in Chapter 9 we will see to what extent members of the European Parliament, members of the national parliaments, and voters share similar attitudes on relevant issues. Although it would be a naive idea of representative democracy to expect total agreement on each and every issue between the policy views of the (majority of the) electorate and the party they vote for, a structural and permanent disagreement between political parties and their voters could hardly be seen as an indicator of a smoothly-functioning system of representative democracy. The Responsible Party Model is a member of the family of coercive models of political representation (Luttbeg 1974): political representatives are forced to take the policy preferences of the electorate into account at the penalty of being voted out of office. In a less rigid view of political representation the electorate plays a more passive role. According to the law of anticipated reactions political representatives will take the policy preferences of the electorate into account, either because they feel it is their duty to do so, or out of fear of being voted out of office, However, this can only be an effective mechanism of linkage if two conditions are met. First, they should consider it their role to take the views of the represented into account. This question of role orientations of members of parliament will be taken up in Chapter 10. A second condition is that political representatives have a correct view of voters' preferences. Chapter 11 will examine to what extent this is the case.

20

Introduction RESEARCH DESIGN

This book is part of the publication programme of the European Representation Study 1994 and based on data sets generated by it. A European Election Study has been conducted on the occasion of all but one election since the first direct elections to the European Parliament in 1979. These studies have been the source of a great number of publications ever since. The last in a long row of publications is Choosing Europe, mainly written by Cees van der Eijk and Mark Franklin, and reporting on the European elections of 1989 and (to a lesser extent) of 1994. In many ways the present volume can be considered a sequel to that book. Choosing Europe was devoted to the problem of representative democracy and legitimacy in the European Union probably as much as this book. However, due to the study design of the 1989 European Election Study, which focused on a series of cross-sectional surveys of the electorate, it could not empirically address all the questions that one would like to answer in a study of political representation. Any true study of representation requires cross-level analysis, incorporating a survey not only among the electors but also among those up for election and ultimately among the elected. In order to widen and deepen the understanding of the process of political representation in the European Union, the European Representation Study was designed to overcome the analytical limitations of former studies (see Figure 1.4). As in former studies, a voters' survey was carried out in the then twelve member-countries of the European Union. This study was realized as the major component of a special post-electoral Eurobarometer survey (EB 41.1). In addition to the voters' study, an elite study was carried out. It included; (a) a mail survey among candidates standing for office in the European Parliament elections, which was realized just before the European elections of 1994; (b) a survey among members of the European Parliament. Relying greatly on the candidates' questionnaire, the survey among the members of the European Parliament was administered as face-to-face interviews; (c) a mail survey among members of the national parliaments of the member-states; the questionnaire of this study is almost identical to the one used for the members of the European Parliament.

European Election Study1994 European Representation Study European Candidates' Survey 1994 European Parliamentarians' Study

Members of the European JPariiamgnj Survey 1996 Members of National Parliaments Survey 1996/97

FiG 1.4. The study design of the European Representation Study

This page intentionally left blank

PART I LEGITIMACY

This page intentionally left blank

2 A Political Community?

ANGELIKA SCHEUER

INTRODUCTION

THIS chapter deals with one core question: whether and to what degree the citizens of the European Union have grown together into a political community. In modern history, political communities were formed by processes of nation-building. A national territory was integrated through the creation of common institutions, channels of communication, and symbols of unity. In the course of this integration process, the inhabitants of a territory grew together into a national fellowship. A demos materialized when individuals began to identify themselves as members of a national political community. For this, inhabitants did not need to be homogeneous in linguistic, religious, or ethnic terms—although such homogeneity certainly facilitated nationbuilding and the creation of a political community. The idea of a political community is intimately linked with the concept of citizenship. 'Citizenship refers to the rights which a state confers upon certain or all individuals in a territory over which it has control' (Kuknle 1993; 94), The creation of a citizenry, i.e. the codification of the rights and duties of individual citizens, is a core element of nation-building. This citizenry, at the same time, is the source of authority of any democratic state: the principle of democracy requires that

26

Legitimacy

powers and executive competencies of the state must originate in and be justified by the citizens, European integration started out as an international alliance of nation-states. It concerned first and foremost economic issues. Economic integration reached a peak with the realization of the Single European Market when member-states transferred important policy-making competencies to the European Community. The Maastricht Treaty codifying this transfer of competencies is actually said to have shifted the balance of European Union government from a formerly predominantly intergovernmental to a now mainly supranational mode (Katz and Wessels 1999b). In policy areas where intergovernmental decisionmaking was replaced by supranational decision-making, the position of the European Parliament as the representative body of EU citizens has been strengthened. The Maastricht Treaty also for the first time warrants European Union citizenship which grants citizens of member-countries the right to vote in local, regional, and European elections in any other membercountry, consular protection in third countries, and the right to address the EP or the EU ombudsman. Still, Union citizenship does not have the same quality as national citizenship. This is at least the underStanding of the German Bundesverfassungsgericht (Federal Constitutional Court), according to which the legal bond 'EU citizenship' uniting citizens of different EU member-states is less cohesive than that uniting a national citizenry (Bundesverfassungsgericht 1993: 437). The increasing role of supranational, as opposed to intergovernmental, decision-making and the establishment of a European citizenship might have promoted the development of a political community of the European Union. But, as Thomassen and Schmitt have rightly pointed out in the introduction to this volume, the growing together of a political community depends at least as much on people's selfperceptions and identifications as on the provision of rights of citizenship or on predominant modes of government. Therefore, the central question of this chapter can be reformulated as follows: do EU citizens identify themselves as such? Do they perceive their fellow EU citizens to be alike? Have European citizens developed a 'sense of community'?

A Political Community?

27

1, The Common European Heritage History has shown that the emergence of a sense of belonging and community and related attitudes such as perceptions of identity and solidarity takes a long time. Compared to the time that nation-states took to consolidate, the history of European integration is still rather short. Feelings of identity and solidarity can hardly have fully developed during these brief periods of history. But, of course, centuries of common European history elapsed before European integration began, Are there traditions in European history upon which perceptions of a political community could be based? Is there a European tradition of unity? Is there something like a historical European identity? The tradition of the Greek polls and the Roman empire influenced in similar ways the development of institutions in the legal system, the armed forces, and the administration of European nation-states. Later, family relationships of the nobility and aristocracy created alliances above and independent of national borders. The Catholic Church established Latin as a universal language, created a continent-wide network of monasteries, and founded the first universities. All over Europe, the same sequence of reference cultures came into force: first Greek and Roman, then (during the Renaissance) Italian, and German and Austrian during the Enlightenment and Romanticism. Likewise, Europeans used to refer to common cultural achievements (literature, music, architecture) and to common European symbols: Roman monuments, the victory over Islam, the Crusades, the French Revolution (Pfetsch 1997: 104-5). Last but not least, Europeans consider themselves to be a community of values and ideas. The idea of liberty, of democracy, of the modern nation-state, individualism, human rights, freedom of speech, rationality, the political republic, and the separation of Church and State—all these are considered genuinely 'European' (Mintzel 1997: 325-6). However, these characteristics have become central elements of a world-wide political culture and are no longer distinctly European, As a result, they probably cannot provide much cement for a European political community because the existence of distinctive properties is a central prerequisite for the evolution of any community and common identity (Lubbe 1994: 111).

28

Legitimacy

2. Sources of Conflicts and Diversity Europe is characterized not only by its common heritage. There is as much diversity and conflict as there are common roots. Three religious cleavages are at the basis of distinct socio-cultural areas on the European continent: the division between Latin and Orthodox Christianity, that between the Christian and the Islamic world, and finally the division between Catholics and Protestants. In addition, Europe exhibits a great variety of languages, which has become even more distinctive with the development of the nation-states in the nineteenth century. It is against this background that some think of Europe as a huge 'multicultural society* composed of a variety of religious, national, and regional cultures (Mintzel 1997; 332-6). European diversity of course also has political aspects. Starting with the breakup of the empire of Charlemagne, Europe's history has consisted of divisions and violent conflicts. Throughout, fellow Europeans have waged wars in changing coalitions on the European continent. Examples are the Hundred Years' War between France and England; the rivalry between France and Habsburg with Spain, The Netherlands, and Austria over European dominance; the bitter confessional wars between Catholics and Protestants preceding the Pax Westfalica; the Holy Alliance of Restoration against the revolutionaries in the nineteenth century; the Entente Cordiale against the European middle states in the First World War; the German-Italian axis against most o the rest of Europe in the Second World War; and the 'cold war' between Western democracies and the communist bloc after 1947 (Pfetsch 1997: 102-3).

3. An EU Political Community? Reviewing the history of Europe proves that the traditions of diversity, division, and conflict are at least as strong as the common cultural heritage. This history of diversity does not necessarily prevent the evolution of a European political community. However, the mere existence of nation-states based on a century of cultural and political autonomy constitutes an obvious obstacle. First of all, these nationstates are linguistic communities that guarantee the communicative competence of every citizen.1 European Union citizens, by contrast, are

A Political Community?

29

confronted with an immense linguistic variety. As a result of this apparent Babel, a European public has not yet really emerged, more or less segmented national publics are perpetuated, and there is very little communication covering the whole EU. Moreover, a genuine EU system of opinion formation and interest intermediation has not yet fully developed over and above the respective national (sub-)systems (but see Chapter 6 on distinctiveness and cohesion of EU parties). As a result, processes of legitimizing European Union government still seem to rely largely on the effectiveness of the respective national (subJsystems. This seems to suggest that the objective conditions for the development of an EU political community have been fulfilled only to a small degree, However, there are other factors that might have promoted the development of a European political community. Not least among them is the obvious economic success of the process of European integration. Also, the greater permeability of national borders after the agreement of Schengen as well as the ever-increasing contact frequency of European citizens as a result of progressing economic integration might have promoted perceptions of community and mutual solidarity among EU citizens. The prime purpose of this chapter is to determine the degree to which the European Union has developed into a political community. After this brief review of objective conditions, we will now turn to both a more subjective and empirical view. According to Easton (1965: 177), a political community exists when members show some readiness or ability to work together to resolve their political problems. That a European political community in such terms exists is unquestioned, but we are interested in knowing whether European citizens, during almost half a century of European integration, have developed a European 'sense of community'. The existence of a political community does not necessarily require that its members are aware of it—i.e., the prior existence of a sense of community. However, the more strongly such a sense of community is developed, the greater are the system's stress-reducing capabilities. This concept of sense of community was first introduced by Karl Deutsch. He defines it as 'a matter of mutual sympathy and loyalties; of "we-feeling", trust and mutual consideration; of partial identification in terms of self-images and interests; of mutually successful predictions of behaviour, and of co-operative action in accordance with it'

30

Legitimacy

(Deutsch et al. 1957: 36), Easton (1979) follows Deutsch in his conceptualization of the 'sense of social community'; in his view, cohesion emerges between people regardless of the type of political regime they live in. He therefore distinguishes this 'sense of social community' from a more specific 'sense of political community'. In his typology of political support, the latter represents the highest (i.e. the most basic and enduring) category of diffuse support for the political system. Our empirical investigation of the sense of a European community distinguishes two basic dimensions. Identification refers to the citizens themselves: do they consider themselves European citizens and are they proud to be European? We-feeling refers to fellow citizens: do European citizens consider their fellow Europeans to be as trustworthy as their countrymen? Which new members if any are they ready to accept in 'their' Union? Figure 2.1 illustrates this conceptualization and specifies the operationalization strategy pursued in this chapter. These notions of identification and we-feeling are compatible with modern theories of intergroup relations. Their starting point is the distinction between ingroups and outgroups. Minimal differentiation is sufficient to give rise to an ingroup/outgroup distinction. This is reinforced by overstating intragroup similarity and outgroup differences. Ingroup membership is an important factor in the formation of personal identity. Ingroup-outgroup relations are driven by social processes of categorization, comparison, competition, and conflict. As a result of these dynamics, perceptions of ingroups are biased toward homogeneity, and attitudes towards outgroups and their members are characterized by stereotyping and hostility. In this view, the evolution of a sense of community among EU citizens is the result of ingroup formation. Shortly after World War II, intergroup conflict between European societies was still extremely high. One of the central aims of the founding fathers of the European Union was to reduce conflict and overcome hostility between European societies by creating a new, superior ingroup which eventually would lead to the development of European identifications and we-feelings, This chapter aims at measuring the success of European ingroup formation after half a century of economic and political integration. Our analyses follow a dual strategy. Firstly, we apply a diachronic perspective and investigate the development of European identifica-

31

A Political Community? political community

subjective criteria

'sense of community'

we-feelirjg

identification

perceived citizenship

pride of being European

mutual trust

new members

citizens

citizens - elites EB & ERS

citizens EES & EB trend

citizens - elites EB & ERS

EES & EB trend

FIG 2.1. Concepts, indicators, and data sources tions and we-feelings over time. Secondly, we compare the orientations of citizens with those of political elites. Since European integration has been described as an elite-driven process supported by a 'permissive consensus' of the mass publics (Lindberg and Scheingold 1970). it is crucial to determine the distance between leaders and led, i.e. to determine how permissive the consensus is. The data used for our purpose are those analysed throughout this book (see Figure 1,4), i.e. the data of the European Representation Study plus selected Eurobarometer trends. The indicators are discussed one by one, in the sequence suggested by the analytical scheme above.

32

Legitimacy IDENTIFICATIONS

The aim of this section is to monitor the evolution of European identifications. First, mass perceptions of European citizenship are tracked over a period of thirteen years (1982 to 1995), Secondly, the pride of EU citizens and political elites in being European is compared.

1. Perceived Citizenship Eurobarometer provides us with two instruments for the analysing European identifications. The first was fielded eleven times between 1982-1992. People were asked whether, in addition to their national citizenship, they also consider themselves to be European citizens.2 The second instrument started a new time series in 1992 when the first was discontinued. In four surveys between 1992 and 1995, people were asked to think about their future political identification.3 Although the two measures are not strictly comparable, they still offer an impression of the direction of trends over the whole period. The trends using both indicators are displayed in Table 2.1. The figures report, country by country, proportions of respondents who 'never* think of themselves as European citizens (according to the first indicator), and who see themselves in the near future as "only national' (according to the second). The countries are ordered by their sequence of entry into the European Union. Additionally, average proportions for three country groups—the original six plus the northern and southem expansion countries—are displayed. This presentation of the data follows the expectation that duration of membership has a positive impact on identification levels: the longer a member, the higher the identification. We thus expect to find a pattern similar to the one identified for the development of general EU support (see for example Schmitt and Treiber-Reif 1990; Dalton and Eichenberg 1991, 1992; Eichenberg and Dalton 1993; Bosch and Newton 1995). This, however, is not borne out by the data. Rather, country group characteristics come to the fore: the northern expansion (adding the UK, Ireland, and Denmark to the Community) brought in Euro-distant publics, while the southern expansion (adding Spain, Portugal, and Greece) integrated

TABLE 2.1. Feeling like a European citizen 1982-1992 (per cent 'never') and thinking of oneself as European or national 1992-1995 (per cent 'only national*) 1982

1983

1985

1986

1987

1988

1989

1990

1991

1991

1992

1992

1993

1994

199S

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

54 39 17 41 16 57 55 64 72 36

49 41 25 46 27 48 45 69 74 45

49 32 29 37 27 47 46 68 70 39 37 39

43 31 29 36 30 42 46 58 68 39 39 34

54 45 41 47 34 62 57 59 67

50 36 40 39 44 59 45 58 62

45

41

36 40

42 39

45 40 47 46 41 61 55 59 68 36 38 35

46 41 53 43 42 60 50 67 71 41 47 45

35 32 50 37 27 53 46 59 69 39 31 31

44 40 51 43 36 60 53 64 69 36 37 35

42 46 59 41 35 57 48 62 68 39 38 33

37 31 41 24 27 42 48 52 54 38 38 34

32 32 43 26 36 40 50 48 59 41 41 38

29 22 29 25 17 33 48 38 49 46 41 34

33 28 34 22 17 34 48 40 53 52 44 42 52 47 56

Original six Northern expansion Southern expansion

35 70

41 38 71

34 68 39

33 65 38

Source: Eurobarometers; figures based on weighted data.

46 66 38

40 61 41

45 67 37

47 69 46

42 66 33

46 67 37

50 66 37

34 54 35

35 58 39

27 49 37

29 52 44

34

Legitimacy

distinctly pro-European publics. Contrary to our expectations, there is no structural difference in European identifications between the citizens of old and new member-countries. The duration argument also implies that, within national publics, European identifications should grow over time. This expectation is again not in accordance with empirical evidence. National trends do not in general increase in a linear fashion. Only a few countries are characterized by rnonotonic trends. Germany is the most striking case, but the trend is opposite to the expected direction. The proportion of Germans who never felt like European citizens grew from 17 per cent in 1982 to 59 per cent in 1992. Over the decade covered by these data, Germany moved from highest to lowest rank among the original six member-countries. This dramatic decline4 is tentatively explained by growing German self-esteem which was accelerated by reunification; public discontent about the German net-payer position (Zahlmeister Europas) may play an additional role. Declining European identifications are also characteristic for Luxembourg,5 while they are on the rise in Belgium.6 A rather stable predominance of national over European identifications is finally identified for respondents from Ireland and the British Isles. The general pattern, however, is not one of linear trends. Rather, we observe fluctuations that affect the different publics in more or less the same way. Tentative explanations for these ups and downs refer to the Single European Act (peak in 1986), northern concerns about the southern expansion (decline in 1988), and the growing discontent with the ratification process of the Maastricht Treaty (decline since 1991). Niedermayer (1997) actually attributes this decline at the beginning of the 1990s to two factors; first, the change in basic economic conditions (recession), social welfare (cut-backs), and security (dissolution of the communist bloc and war in former Yugoslavia) and, second, the increasing importance of EU policy-making for everyday life. Lilli (1998) maintains that the debate on European Monetary Union may have depressed European identifications. The second indicator suggests that the proportion of European identifiers increases again between 1993 and 1995 in almost all countries and in some even profoundly. This is probably the result of a relaxing of tensions produced by the Maastricht and EMU processes after decisions were made and measures implemented.

A Political Community?

35

Figure 2,2 illustrates the level of European identification in the different member-countries in 1995; the darker a country, the higher its proportion of European identifiers. Something of a centre-periphery pattern emerges. European identifiers are most numerous in the original six member-countries-—-in Italy above all, but also in the others. Spain, Portugal, and Ireland come next. The third and most distant group includes the UK and Greece as well as the new members Austria, Finland, and Sweden.

Source: Eurobarometer Flash #3? conducted in June 1995.

FiG 2.2, National or European identifications? (per cent 'only national')

2. Pride in Being European Pride in being European is another indicator of European identification. Here, we can compare voters' and elites' views. Based on the

36

Legitimacy

findings of earlier research (Lindberg and Scheingold 1970; Bardi 1989) we cannot but expect elites to be prouder of being European than the mass public. What is more interesting is how large the gap between voters and elites actually is. A particularly wide gap might signal an erosion of the 'permissive consensus'. A narrow gap, on the other hand, might suggest that the elites are no longer prepared to march ahead of the mass public, that their willingness and energy to advance the process of integration is lacking. Due to the lack of appropriate time-series data, we cannot test the dynamic aspects of our reasoning in a proper sense. Even worse, crosslevel comparisons between elite and mass data are again somewhat difficult because of differing question wordings and answering scales.7 However, we can distinguish on both levels the proud from the not proud and the indifferent. Table 2.2 indicates the proportion of 'proud' respondents among voters, members of national parliaments (MNPs), and members of the European Parliament (MEPs). For the elites, proportions are given both in absolute terms and relative to the voters' proportion (proportion of proud voters taken as percentage base). Among voters, we find roughly the same country pattern as with the identification measures discussed earlier: Italians and Luxembourgers are the proudest, while Germans, Finns, and Swedes are the least proud. Pride in being European is closely related to our earlier measures of European identification. Comparing levels of pride between voters and elites, the expected pattern predominates: elites generally are prouder Europeans than voters. But there are exceptions to the rule. The most striking exception is Denmark. Danish members of the European Parliament are found to be less proud of being European than the average Danish citizen. We lack information about Danish national representatives and cannot determine whether this is characteristic of the Danish political elite in general, or of the Danish MEPs in particular. MEP data from Denmark suggest that this is a result of selective electoral mobilization on the occasion of European Parliament elections, when anti-integrationist views are more prominent than usual and critics rather than supporters of European integration are successfully mobilized to participate. Another deviant situation is found in The Netherlands and Luxembourg: national representatives are less proud to be European than voters in general, while these countries* delegations to the European Par-

A Political Community?

37

TABLE 2.2. Pride in being European (percent 'proud') Voters

MNP

MEP

MNP

MEP

absolute

absolute

relative

relative

Denmark

49

-

France Italy Portugal Spain Belgium United Kingdom Ireland Austria Finland Sweden Greece Germany

65 80 64 66 60 51 64 46 41 37 47 42

72 86 76 92 66 . 81 -

Luxembourg Netherlands

70 45

n.a.

58 62 50 42

38 64 80 69 86 78 68 89 70 73 64 82 83 100 86

-

78

Ill 108 119 139 110 127 -

98 100 108 130 130 133 139 152 178 173 174 198

n.a.

123 148 71 93

143 191

Source: European Representation Study, weighted data. n.a.= not ascertained; - = not surveyed. liament are considerably prouder than both MNPs and voters. This indicates less of an anomaly in Luxembourg where most voters are European-minded. The situation is different in The Netherlands, where members of the Second Chamber apparently pull public opinion in a Euro-sceptical direction, while Dutch MEPs are among the proudest Europeans in the Union. In a large majority of member-countries, elites identify in (relatively) greater numbers with Europe than the mass public does. This trend is the more pronounced the more distant the respective public is.

38

Legitimacy WE-FEELING

Our operational definition of 'sense of community* distinguishes two dimensions: identifications and we-feelings. In this section, we turn to the second and investigate (1) whether EU citizens trust their fellow Europeans, and (2) where the geographical borderline between ingroup and outgroup is drawn.

1. Mutual Trust Mutual trust is a fundamental condition for the development of a sense of community. It is expected to grow with growing experiences of positive conduct of fellow citizens. The existence of a common enemy is another factor contributing to the development of a sense of community. For most of the post-war period, the communist threat was an external reference point that might have fostered perceptions of a common bond amongst the people of the European Union. Actually, since the collapse of the Soviet Empire, observers complain about the return of nationalism, and fears have grown that the community may fall apart without the Eastern threat. On the basis of these considerations, we expect to find two different developments in the data: first, a general increase in mutual trust between the member-countries of the European Union over the years and, second, a weakening of this common bond since 1990. Trust in people of various countries has been measured repeatedly in Eurobarometer surveys between 1970 and 1994 (including the EES '94).s Earlier work has shown that trust between EU member-countries is generally higher than between members and non-members, and that mutual trust between the EU member-countries has grown over time (Niedermayer 1995). We look at the data in a slightly different way. Taking trust in the people of one's own country as a baseline, we will establish, for every respondent, whether or not the people of various other countries are equally trustworthy as one's countrymen. Table 2.3 displays for every country how much its people were trusted by the people from other countries of me Union. In order to avoid composition effects, we re-

A Political Community?

39

strict our analysis to the EU9 countries which are surveyed at all points in time, It is evident that mutual trust between the people of EU membercountries substantially increased during the 1970s and 1980s. The breakup of the Soviet Empire seems to have interrupted this trend. In the first half of the 1990s, this decline was only partly made up, and mutual trust had not yet regained the high levels of the 1980s, Table TABLE 2.3. Mutual trust (figures are per cent)

Belgium France Germany Italy Luxembourg Netherlands Denmark Ireland Great Britain Greece Portugal Spain

1976

1980

1986

1990

1993

1994

62* 49 62 31 64 66 64 41 51 73

60 50 61 31 63 64

66 57 67 45 71 70 73 55 56 76 51 52

58 53 56 39 61 59

67 57 57 40 70 67 67 52 53 69

63 58 57 42 66 62

64 47 55 71 39 41

60 45 43 67 41 46

47

53

64 52 52 70 46 52

Source: Eurobarometers and European Election Study 1994; weighted data. * Read: in 1976, 62 % of all non-Belgian EC9 citizens trusted Belgians at least as much as people from their own country.

2.4 facilitates comparison by providing just two measures per country: (1) whether, in 1994, the people of a particular country were considered as trustworthy as one's own by a majority of citizens from the other member-countries, and (2) how this nation's balance shifted from 1974 to 1994. The people of all but two EU9 member-countries are trusted by a majority of fellow Europeans. Only Italians and Portuguese just miss the 50% threshold, However, looking at changes over time, it is striking that those who were least trusted at the beginning of the time series have gained most credit since, among them Italians (+10%) and

40

Legitimacy

Portuguese (+8%). The Germans, Dutch, and Greeks have lost some of the trust accumulated earlier. In the German case, this may reflect a sentiment of caution towards reunified Germany. Altogether, levels of trust in the people of other member-countries have adjusted and become more similar over time. TABLE 2.4. Trust in the people of the other EU member-countries (figures are percentage points)

Ireland Italy Spain France Portugal United Kingdom Belgium Luxembourg Denmark Greece Netherlands Germany

Majority In 1994 ?

Difference 1994-1976

yes no yes yes no yes yes yes yes yes yes yes

12 10 10 9 8 2 2 1 0 -3 .4

-5

Source: as in Table 2,3,

2. New Members We-feelings are expressed as appreciation of actual members of an ingroup. In the case of the European Union, however, there is a large number of would-be members beyond those who already take part. Over the years, the domain of the European Union has grown in various steps, and further expansion is planned following the dissolution of the Soviet Empire. Currently, the membership of Poland, the Czech Republic, Hungary, and Slovenia, as well as Estonia and Cyprus, seems imminent, while other Central and Eastern European countries, as well as Turkey, are waiting impatiently. Is there any limit beyond sheer practicability? Is there a mental map of the European Union in

A Political Community?

41

the heads of citizens and elites, which includes some farther membercandidates and excludes others? Table 2.5 identifies the mental map of European citizens. It is based on pooled results of three Eurobarometers conducted between 1991 and 1994.9 The country list presented to interviewees varies a bit from survey to survey; it usually includes non-EU countries of Western Europe, selected candidates from Central and Eastern Europe, plus Turkey and Morocco—two countries bordering the Mediterranean Sea, TABLE 2.5. Citizens' views about new members (figures are per cent) For

Against

Don't know/ no answer

Sweden Switzerland Austria Norway Finland Iceland Malta Hungary Cyprus Poland

80 79 78 78 77 71 58 55 53 51

7 9 9 9 9 12 21 25 27 28

13 11 13 14 15 17 21 20 20 20

Czech Republic Russia Bulgaria Slovakia Romania Slovenia Turkey

49 47 46 45 45 44 39

29 34 32 32 34 33 42

22 19 22 23 21 24 20

Source: Eurobarometers; pooled results from EB 37,39 and 42; weighted data,

A definitive part of the citizens' mental map of the future Union is the countries that were admitted shortly after the fieldwork of the surveys analysed here (Sweden, Austria, and Finland) or had negotiated membership but finally declined (Switzerland, Norway, Iceland). They are followed by Malta and Cyprus. A majority of EU citizens also include

42

Legitimacy

two Central European countries: Hungary and Poland. The Czech Republic figures just below the 50% mark. Other Central and Eastern European countries, in a somewhat disorderly sequence, follow suit: Russia is ahead of Bulgaria, Slovakia, Romania, and Slovenia. The end of the list is occupied by Turkey, which is the least welcome applicant country, The mental map of the mass public for a future European Union thus includes, naturally, all of Western Europe plus the better-known of the neighbouring Central European countries such as Poland and Hungary. The message is less clear regarding Eastern Europe: these countries did not—probably not yet—-enjoy the support of a majority of EU citizens. The same goes for Turkey, which is even less popular. Table 2,6 presents the views of European political elites—-members of the European Parliament and of national parliaments-—on new members.10 The country list presented to them is again a bit different from the one presented to EU citizens, but this does not matter much. The result of our analyses is very clear: the mental maps of citizens and elites are virtually the same. The one difference might be that elites include Central European countries in a slightly more systematic fashion. Majorities of political elites in both the European Parliament and national parliaments support their admission to the Union within the next decade. Such majorities do not exist for the Eastern European countries Romania and Bulgaria (others such as Ukraine, Belarus, and Russia were not included in the survey). Finally, majorities of political elites disapprove of the potential Union membership of non-European countries such as Turkey and Morocco. National and European parliamentarians share much the same views. If there is a systematic difference between the two, it is that national representatives seem to be somewhat more liberal in admitting even more distant candidate countries. This is most clear in the case of Turkey, where only one in four national representatives is definitely against membership, while one in two MEPs opts for 'never' accepting Turkey as a Union member.

43

A Political Community?

TABLE 2.6. Political elites* views on new members (figures are per cent) Admit soon**

Admit later***

Admit never

No answer

(83) 86

(7) 6

(3) 1

(7) 7

(78) 79

(12)12

(3) 2

(7) 7

(14)13

(1) 1

(8) 8

(77) 80

(15)13

(1) 0

(7) 7

(68)72

(24)21

(2) 1

(6) 6

(67) 68

(21)21

70

22

(4) 3

(8) 8

(55) 59

(34)31

(2) 1

(9) 9

(52) 54

(36)35

(3) 2

(9) 9

MEP(15)11 MNP 11 MEP(15)11 MNP 11

(38)41

(48)45

(5) 5

(9) 9

(34)37

(50) 47

(6) 5

(10)11

MEP(15)11 MNP 11 MEP(15)11 MNP 1 1

Country

Elite Group*

Malta

MEP(15)11 MNP 11 MEP(15)11 MNP 11 MEP(15)11 MNP 11 MEP(15)11 MNP 11 MEP(15)I1 MNP 11 MEP(15)11 MNP 11 MEP(15)11 MNP 11 MEP(15)H MNP 11

Cyprus Czech Republic Hungary Poland Slovenia Slovakia Baltic States Romania Bulgaria Turkey Morocco

70

62 (77) 78

86 86 80

70 64

52

48

19 24 8 8

14

23 26 40

4 7 1 1 1

2 2 3 2

42

3

(13)16

(27)25

(53)51

37

32

24

(7)17

(15)21

(68) 65

17

21

54

7 7 5 5 5

6 5 7

6 7

(7) 8 7

(10) 7 8

Source: European Parliamentarians* Study (national samples weighted according to countries' population size). * MEP (15) U = (Members of the European Parliament from ail 15 member-countries) MEPs from the 11 member-countries for which MNP data are available; MNP 11 = members of the 11 national parliaments surveyed; "within 5 or within 10 years; *** more than 10 years.

44

Legitimacy SUMMARY

Over the past centuries, the common cultural and political roots of the people of Europe could seldom prevent long-standing hostilities from violent eruption. It was only after World War II that political elites started to initiate the process of European integration which deliberately aimed at creating a common framework of social and political identifications. The political-institutional success of these efforts is obvious. But how about its social basis? Have the people of Europe grown together into a political community; is there a 'sense of community' among EU citizens? The general answer is yes. Roughly half a century after World War II, a majority of EU citizens identifies with the new political community of the European Union. Lacking pertinent and comparable survey information for most of this 50-year period, we cannot determine when and how these identifications came into being. However, based on our findings from the analysis of available data, we must assume that they have been growing slowly. At least between 1982 and 1992, there was not much of a secular change in European identifications; seasonal effects prevailed. A map of European Union identifications in the mid-1990s reveals a centre-periphery picture. The highest level of identifications exists in the six original member-countries; the farther away from this core of the Union one gets in geographical and/or temporal terms, the weaker identifications become. Elites generally identify more strongly than ordinary citizens. To the degree that the process of European integration is elite-driven, this points in the direction of a further increase of European identifications among the mass public. A few country-specific findings deserve mentioning in the summary: in Germany identifications with Europe have been shrinking considerably since the 1980s; a stable majority of British citizens find it hard to identify with Europe; and Danish members of the European Parliament and members of the Dutch national parliament identify less with Europe than their countrymen on average, Majorities of EU citizens trust the people of other member-countries as much as they trust their own countrymen." Trust in the people of Southern Europe and Ireland increased substantially over the last two decades, while reunified Germany and The Netherlands (but also

/I Political Community?

45

Greece) are considered somewhat less trustworthy today than in the past. Not only do majorities of citizens identify with the Union and trust their fellow Europeans, they also have a rather clear-cut mental map of where the Union ends, a shared understanding of who belongs to it and who does not. EU citizens and elites broadly agree that the European Union might include all people of Western Europe, Some Central European candidates are welcome, while the countries of Eastern Europe and in particular non-European countries are not.

NOTES 1

This is not to say that states must be linguistically homogeneous; Belgium and Switzerland are obvious examples of nation-states that are not. How ever, in order to meet democratic requirements, every citizen must be able to communicate with state authorities in his or her own language (BVG 1993: 438). This implies in places that there is more than one official language (such as three in Belgium and four in Switzerland). 2 'Do you ever think of yourself not only as a [nationality] citizen but also as a citizen of Europe? (I) often, (2) sometimes, (3) never.' 3 In the near future do you see yourself as (1) [nationality] only, (2) [nationality] and European, (3) European and [nationality] or (4) European only?' 4 Regressing identification levels on time results in r = .961. 5 Regressing identification levels on time results in r = .663. 6 Regressing identification levels on time results in r = -.679. 7 The question wording in the Voter Study was; 'European Union Member States are "European Citizens". Are you personally proud or not to be a "European Citizen"? Would you say that you are (I) very proud, (2) fairly proud, (3) not very proud or (4) not at all proud?' Elites were confronted with the statement 'I feel proud to be a European' and asked to express their agreement or disagreement on a seven-point scale running from (1) agree strongly to (7) disagree strongly. 8 The following question was asked: 'I would like to ask you a question about how much trust you have in people from various countries. For each, please tell me whether you have (1) a lot of trust, (2) some trust, (3) not very much trust, or (4) no trust at all.' 9 People were asked: Tor each of the following countries, are you in favour or not of [country] becoming part of the European Union in the near future?' 10 Parliamentarians were asked: 'For each of the following countries would you indicate when it should become a member of the European Union or should it never become a member?"

46

Legitimacy

11 With the exception of Italy and Portugal. Note that Italians, who identify most strongly with the Union, are least trusted by their fellow Europeans.

3 The Scope of EU Government

LIEVEN DE WINTER AND MARC SWYNGEDOUW

INTRODUCTION

THE question of which level of government or decision-making is or should be the most appropriate to deal with different policy areas is multifaceted and has been posed by political philosophers for centuries,1 As Thomassen and Schmitt argue in the introductory chapter of this book, this question refers to one of the aspects of the legitimacy of the political system. In this chapter we will not try to argue the appropriate level of government from normative political theory. In order to determine how legitimate the division of power between different levels of government in the European Union is, we will rely largely on the evaluation of the citizens. In the framework of EU governance, the question of the appropriate level of government has been dormant for a long time. However, the subsidiarity dispute has turned it into a hot issue in the larger debate on a democratic deficit within the EU. This chapter will first examine to what extent and for which policies European citizens allocate decision-making responsibilities to the European Union, the national state, and regional authorities. Second, we will test a set of hypotheses concerning the structural and attitudinai characteristics that affect differences in preferences for government levels. Third, we will examine to what extent citizens' preferences are congruent with those of their elected representatives who can (co-) de-

48

Legitimacy

clde on potential future shifts of competencies between the national states and the EU,

THE LEVEL OF G O V E R N M E N T IN POLITICAL THEORY AND THE EU-LEGITIMACY DEBATE I. Methods for Attributing Levels of Government The grounds on which certain competencies can be attributed to different levels of government are threefold (Sinnott \ 995), The first is the very nature of issues. Some issues are intrinsically international, as they penetrate or transcend national borders, and therefore cannot effectively be dealt with at the national level. Others require a largerthan-national scale in order to mobilize the resources necessary to solve the problem. Likewise, sorne problems (like improving traffic security at an accident-ridden crossing in a particular neighbourhood) are so narrow that only local government can effectively deal with them. These arguments fall under what we will call endogenous attribution of levels of governance. Second, the level of governance can be attributed in an exogenous way when a given level of governance claims that a particular issue or policy sector lies within its legallydefined sphere of competence. Last but not least, issues can be attributed to a given level of government because the citizens, the media, and political elites believe that this problem ought to be dealt with at that level (normative attribution of level of governance). Of course there is a logical connection between the three different grounds of attributing competencies to a particular level of government. The subsidiarity principle is based on the assumption that the nature of problems should define which level of government should be allocated a primary responsibility. Also, it is an important empirical question as to what extent people's convictions about the appropriate level of government in a particular policy area depend on the nature of the problems concerned. This is one of the major research questions of this chapter.

The Scope ofEU Government

49

2. The Level-of-Government Discussion in the EU Limiting ourselves to the three levels of government which are relevant to the EU governance debate (i.e. the regional, national, and European levels), the question of the appropriate level of government has been posed in three types of competence conflicts. The oldest debate originates in federal states and has to do with which competencies should be allocated to the federal level. In formerly unitary states that are currently undergoing a process of federalization, regionalization, or devolution (Spain, Belgium, France, the UK, Italy) this question figures high on the political agenda, opposing maximalist demands voiced by ethno-regionalist parties and the desire of the state-wide parties and institutional actors to minimize this transfer of competencies to the sub-state level (Keating 1995; De Winter and Ttlrsan 1998). Also, in consolidated federal states such as Germany and the US, the question is nourished by the increasing dependence of states on the federal government for financial aid, which has enabled the federal government to influence policies which are nominally within the control of the state government. Yet, even in stable non-federal countries, the question about the desirable degree of autonomy of the sub-national level is recurrently posed in many democracies (Hesse and Sharpe 1991).2 Second, there is the debate about international governance, in particular EC/EU governance, vis-a-vis the sovereignty of the national states. International organizations historically emerged and expanded in number and scope precisely due to the fact that more and more issues inherently transcend national borders, and therefore can only be dealt with through some form of permanent international co-operation. In the European context, this question has focused on policy sectors in which European institutions can operate more effectively than member-states separately. Here several theories are competing. Neofunctionalist integration theory states that internationalized governance of inherently supranational issues is to be achieved by internationalizing the less political sectors (such as transport and communication, economy, finance, and cultural exchange), and that the benefits of internationalization in these sectors should entice national government to slowly expand international co-operation into sectors of'high polities', such as foreign affairs, defence, and judiciary and police, that are all at

50

Legitimacy

the heart of the sovereignty of the nation-state (Haas and Schmitter 1964)3 This philosophy, which has implicitly or explicitly guided the strategy of the federalist founding fathers of the European Community, must to some extent be tempered by the Maastricht Treaty, which puts a strong, albeit only symbolic, constraint on this progressive and deterministic vision. The Maastricht Treaty's adoption of the principle of subsidiarity aimed at explicitly installing a set of rules covering the division of competencies between member-states and European institutions. Art, 3 states that: In areas which do not fall within its exclusive competence, the Community shall take action, in accordance with the principle of subsidiarity, only if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States and can, therefore, by reason of scale or effects of the proposed action, be better achieved by the Community.

Hence, only those policy sectors that cannot be dealt with effectively at a lower level of government are eligible for Europeanization. Still, this principle is so vague that it serves predominantly political objectives, i.e. putting the Euro-sceptics at ease, rather than serving as a practical guideline for stipulating which policy sectors should remain national or not.4 Nor can it be considered 'justiciable' (Schilling 1995). Finally, the most recent debate opposes the policy competences of the European Union vis-a-vis the regional level. The principle of subsidiarity also implicitly recognizes the potential role of regional and local government. The creation of the Committee of the Regions in 1994 illuminates the fact that also the regional level is willing to contest the irreversible trend towards gradually increasing the decisionmaking prominence of European institutions, implicitly and explicitly referring in its opinions to the principle of subsidiarity or 'proximity' (du Granrut 1994; Hooghe and Marks 1996; Dony 1997; de Broycker 1997; Vandersanden 1997). 3, Level of Government, Legitimacy, and Political Representation The question as to which issues or policy sectors are subject to endogenous internationalization is not easy to answer. There is no consensus on the range of problems that, due to the nature of the issue,

The Scope of EU Government

51

belong to the remit of local, regional, national, European, or international governance.5 The degree of endogenous internationalization of a particular issue can vary between countries due to structural constraints. First, certain countries can be considered the cause of a problem suffered by others, as in the case of a country producing heavy air pollution surrounded by cleaner neighbours. For the latter, the problem of air pollution is endogenously international; for the former it is in the first place a problem that can and should be solved by the national government.6 Second, in some countries, a particular level of governance (national, regional, or local) can be better equipped to deal effectively with an issue. This can be due to economy-of-scale effects, depending on the size of the territory covered and material resources and expertise different levels of governance have at their disposal.7 Third, the degree of interdependency of economic and social systems may vary considerably: in a closed, autarchic economy, the need for international collaboration and governance may be less strong than in open economies where prices, wages, interest rates, and so on are highly dependent on the policies and economic fortunes of the main trading partners. To conclude, establishing the endogenous European character of a given policy problem is a hazardous exercise, as this character may objectively vary considerably from one member-state to another. Contrasting citizen opinion on the division of power within the EU with opinions of elites enlarges the problem of legitimacy to one of political representation. Through the ratification of European treaties, the transcription of EU directives into national legislation, and constitutional reform, national MPs control the transfer of competencies between the national and European levels, and between the regional and the national. Hence, decisions on level of government can be considered a policy sector on which the represented and their representatives can diverge in opinion, just like any other policy sector.* The more elites and citizens are unaware of each others' preferences on this issue, the greater the probability that elites will follow their own vision in deciding on level attribution, rather than the preferences of the public in the sphere of institutional design.

52

Legitimacy

4. Research Strategy The first part of our empirical analyses focuses on the question as to what extent certain policy problems are currently dealt with at the regional, national, or European level, and more importantly, at what level they ought to be decided on. We will elaborate the type of problems for which each level of government is considered most legitimate. The second part of our analyses will focus on explaining variations in these preferences. One can expect potential differences due to the endogenous nature of the problem, as well as citizens' more general attitudes towards different levels of governance in the EU and structural variables. Finally, the analysis will show whether the represented and their representatives hold similar preferences with regard to the division of competencies between the three levels.

KNOWLEDGE OF DECISION-MAKING LEVELS AMONGST EUROPEAN CITIZENS

A first matter to be clarified is whether European citizens have any opinion at all about the question of division of labour between different governmental levels. If few citizens do, then one can hardly speak of a democratic deficit or inadequate representation of citizens* opinions on this issue, as there is no opinion to represent. This question can be answered by looking at the non-response to three questions asked in the voters' study of the 1994 European Election Study: the first asked respondents to identify the three most important problems their society is currently confronted with (choosing from a list of seven common problems included in all countries and four country-specific ones); the second asked at what level this question is currently being dealt with—the regional, national, or European level; and the third asked at which level respondents believed each of these problems should best be dealt with.9 Concerning the saliency question, only 1% could not identify one of the 11 problems that were presented by the interviewer as the most important; another 1% could not indicate a second-most important problem, and only 2% failed to identify a third-most important one. Hence,

The Scope ofEU Government

53

we can safely conclude that most European citizens do perceive the prominence of different issues in their country.'0 The non-response on the second question is somewhat higher: 13% could not indicate at what level of government the problems they perceived as most important were dealt with; for the second- and thirdmost important problems these figures are about the same (13% and 15% respectively). However, they have a slightly better idea regarding the level at which they believe these issues ought to be dealt with: I \% could not mention a decision-level preference for the most important problem, 12% for the second, and 15% for the third. In other words, about seven out of eight citizens do have a view about the current and desired level of decision-making. This comes somewhat as a surprise given the generally low degree of political interest repeatedly found among European publics.11 In spite of the general low interest of European citizens in national and European politics, a very large majority still manage to express their political agenda, the level at which the most important policy problems are currently dealt with (rightly or wrongly), and the level at which they would prefer to see them being dealt with. Hence, with regard to these topics, we are dealing with a public that expresses its opinions well. As expected, their elected representatives in the national and European parliaments display even higher degrees of information and opinion. Hence, the degree of correspondence between citizen opinions and those of the political elites becomes a relevant indicator for the quality of political representation in the domain of preferences for the level of government.

PERCEIVED AND PREFERRED POLICY LEVELS FOR THE MOST IMPORTANT PROBLEMS 1. General Trends A first question to address is the level at which European citizens perceive the most important policy problems being dealt with. Table 3.1 indicates that two-thirds of the most important problems mentioned by the respondents are believed to be currently decided at the national level. The other remaining problems are about equally perceived as

54

Legitimacy

being decided at the regional (16%) and European level (18%). Hence, in the eyes of the citizens, the nation-state is still the predominant level of government in the 'real world'. However, in normative terms one finds that nearly four out of ten (39%) would like to see the important problems allocated to the European level of government, which is more than twice as much as respondents perceive are currently being dealt with at that level (18%). Still, the end of national governments as decision-making agencies is not yet near: for nearly half (48%) of the mentioned problems, the national level is still preferred as most appropriate. It is also important to note that Europeans generally do not want to grant a more important decision-making role to the regional level: respondents prefer regional government to deal with only 14% of the important problems, which is even less than the already-low proportion of problems that are perceived as currently being dealt with at that level. In general terms, European citizens want to grant much broader competencies to the EU institutions than what they perceive to be the current case, at least in the sectors of the three most important problems. They also prefer a smaller role for the national governments, although the latter level is still perceived as the most relevant. These findings contradict the all-too-popular visions of a growing Eurofatigue and Euro-scepticism amongst the European publics. The data suggest that citizens would in fact welcome more transfers of competencies to 'Brussels' for a number of important policy problems. Now, does this surprisingly Euro-positive attitude of the general public correspond to the opinions of their representatives at the national and European decision-making levels? The comparison between elites* and citizens' attitudes in our analysis is somewhat complicated by the fact that surveys among national MPs could be conducted only in ten out of twelve states that were members of the EC at the time of the 1994 election. Hence, strict comparisons between the three surveys can only be based on data sets from which, in the case of the voters and the MEP study, the Danish and British samples have been removed (indicated as V10' data sets in the tables).12

The Scope ofEU Government

55

TABLE 3.1. Perceived and preferred levels of decision-making for the three most important issues (percentages based on responses) Mass public Perc/12 Pref/12

MNPs" Pref/10

Pref/10

Pref/15

MEPs Pref/12 Pref/10

Governmental level: Regional National European

16 66 18

14 47 39

12 45 42

7 48 44

3 44 53

3 44 53

3 43 54

Source: European Representation Study, Figures are based on weighted data which adjust the samples to the size of the nationai population each data set represents. The answers to the first, second, and third problem and the decision-making levels are aggregated using the multiple response technique, as this makes the percentages add up to 100% (rather than to 300%). Percentages are thus based on responses, not on respondents,

Let us first compare the opinions of citizens and their representatives in the European Parliament. Table 3.1 shows only minor differences with regard to the preferences of European citizens and MEPs for the nation-state as decision-making level. As among the general public, nearly one out of two MEPs (44%) prefer the national state as the political level for solving their societies' most important problems. However, differences are more pronounced in the other categories. MEPs want Europe to become competent for many more issues than the public at large. In their eyes, more than half of the important problems require a European solution (53%, or a difference of 14%). At the same time, MEPs consider the region even less appropriate for solving important problems than the citizens, and leave only 3% of the problems to be dealt with at this level. When we bring members of national parliaments into the picture, and therefore omit the Danish and British samples from the citizen and MEP data sets, we notice that national MPs give a priority to the national level (48%), which is slightly higher than the citizens but clearly more than MEPs (45% and 43%). National MPs are marginally more European-minded than the national citizenry (43% vs. 42%), but far less than the MEPs (54%). With regard to the regional level, they keep the mean between citizens' and MEPs' opinions (7% vs. 12% and 3% respectively). Hence, the differences between the national and Euro-

56

Legitimacy

pean MPs seem to be mostly linked to the tendency of elites to favour their own level of decision-making. In fact, national MPs prefer mostly the national level, and MEPs the European level,14 National and European parliamentary elites as well as European citizens are generally in favour of significantly increasing the competence of the European level in a number of important policy areas. This desire is stronger for the elites than for the public, but differences are not so important as to pose a major problem of incongruence. As far as the data suggest a problem of legitimacy, it is of an unexpected kind: the current division of competencies between the national and European levels (as perceived by the public) is considered to be biased in favour of the national level. While citizens believe that few important matters are decided at the European level, they want this level to have a bigger say in these sectors. The most interesting question of course is on which grounds these perceptions are based: is it the fact that the problems cannot be solved by member-states for reasons of scale or the cross-border effects of the policy actions required? Is it because national governments have lost touch with public opinion? In order to get a better understanding of people's preferences for different levels of government, we will look into the possible causes of these preferences. In the next section, we will see to what extent these preferences depend on the kind of the problems to be solved. Thereafter, we will elaborate and test a number of hypotheses with respect to other possible determinants of these preferences.

DIFFERENCES BETWEEN POLICY SECTORS

The survey asked European citizens to identify the three most important problems from a list of eleven issues. Seven issues were identical across the countries under study and four varied in a nationally-specific way. We have regrouped common and nationally-specific issues on the basis of their content into eleven categories: employment and economy, public finance, European integration, immigration, environment, law and order, welfare-state issues, centre—periphery eoofliets, international conflicts, improving democracy, and education (see Appendix).

The Scope o/EU Government

57

The importance of these eleven policy sectors is presented in Table 3.2. Not surprisingly, 'employment and economy' are overwhelmingly considered the main problem of contemporary European societies. Two sets of problems are mentioned by nearly half of all respondents: public finance and law and order issues. Welfare is mentioned by a third of the respondents, environment by one out of four, immigration by one out of five, and one out of six mentions one of a set of centre—periphery problems. One out often respondents mentions issues belonging to two sets of problems: furthering European integration and a series of measures aimed at improving the quality of democratic functioning. Least mentioned are educational matters and international conflicts. Mass and elite opinions on issue saliency are very much alike; in Chapter 9 we will come back to this in greater detail. Table 3.1 indicated that most European citizens believe that most important problems are still decided at the national level. Table 3.3 shows that this is particularly true for the following policy sectors: welfare and public finance and economy, improving democracy, and education. The least 'nationalized' sectors are European integration and international conflicts (as their endogenous international nature is evident), but also environment (the cross-border nature of which is equally evident, as well as increasing EU norm-setting). The latter three problem sectors are seen as being already rather strongly Europeanized, while welfare, public finance, education, and centre-periphery problems are perceived to be the least so. The low score of public finance might be due to the fact that this category includes fair taxation, budgetary orthodoxy, and stable prices, which, in spite of the Maastricht convergence norms, are still objectives to be achieved by national government action only. Finally, the regional level is perceived as currently being most relevant for centre-periphery conflicts (as it basically concerns conflicts between regions and the national state, with those considering this conflict important situated most of the time in the regionalists' camp), and least for improving democracy (as it often deals with changing or creating constitutional and legal provisions, in which neither Europe nor the regions are competent). We know that a considerable group of European citizens and elites want more competencies for the European level of decision-making, mainly by transferring them from the nation-state. Table 3.2 indicates that the problem sectors that are already perceived as being most Europeanized are at the same time those for which further Europeanization

58

Legitimacy

is demanded most: European integration, international conflicts, and the environment. In absolute terms, more than half of European citizens prefer these matters to be dealt with mainly at the European level, about four out of ten equally want this for education, employment and economy, immigration, law and order, and improving democracy. Finally, only about one out of three want public finance, welfare, and centre-periphery problems to be decided at the European level. TABLE 3,2, Perceptions of importance of policy sectors (percentages based on responses) Citizens Perc/12 Perc/10 Employ ,/econ. Public finance Europe Immigration Environment Criminality Welfare Centre—periph. International Democracy Education Total

29 15 3 7 8 15 11

30 14 3 7 9 15 11

5 1

5

3 2 100

2 4 2 100

MNPs Perc/10

32 11 8 6 7 11 11 9 2 2 3 100

MEPs Perc/12 Perc/10

30 13 14 4 10 8 9 5 2 3 2

100

31 11 14 5 8 9 11 5 2 3

3 100

Source: European Representation Study. Perc/12: perceptions calculated on the 12 EC countries; Perc/10: perceptions calculated on EC countries minus British aid Danish respondents.

Decision-making levels which national MPs prefer are quite comparable with those of the public, at least as far as the European level is concerned (correlations between proportions of citizens and MNPs who prefer to allocate decision-making competencies for the eleven different policy domains to the European level are .83). Correlations are a bit lower as far as allocations to the national and regional level are concerned (.69 and .68 respectively). The congruence between MEPs and citizens about policy-making allocations to the European level is quite high again (r = .82); but less impressive with respect to the national level (r = .63). The congruence between MEPs and voters as measured by the correlations between proportions who prefer this level of

TABLE 3,3, Most important policy sectors and preferred decision-making levels (percentages based on responses) Regional

Level Sample Policies Employment/economy Public finance Europe Immigration Environment Criminality Welfare Centre-periphery International Democracy Education

National

MNP

Citizens

MEP

MNP

Citizens

European

MEP

MNP

Citizens

MEP

fete! Pref/ Pref/ Pret? Pref/ Pref/ Perc/ Pref/ Pref/ Pref/ Pref/ Pret7 Perc/ Pref/ Prrf7 Pref/ Preff Pref/ 12 12 10 12 10 12 10 10 12 10 10 12 SO 10 12 10 12 12 16 15 14 16 18 16 15 22 14 12 S5

14 15

12 13 12 13 16 15 7 8 20

Source: European Representation Study.

12 13 12 13 11 li 14 14 7 8 20

7 7 7 7 8 6 6 7 6 8 14

3 4

2 3 3 2 2 4 0 0 6

3 3 2 4 3 2 2 5 0 1 6

67 71 57 65 59 65 71 63 59 70 70

48 52 35 46 37 48 52 52 35 53 38

46 49 32 45 35 46 50 49 35 54 38

49 60 37 42 35 47 53 58 36 47 54

45 55 33 45 28 43 57 54 37 41

54

45

50 34 44 25 40 57 53 37 37 54

17 15 30 19 22 19 15 16 27 18 15

39 33 54 40 5! 38 32 33 58 39 42

42 38 56 43 54 43 36 37 58 38 42

44 33 57 52 57 48 42 35 58 45 32

53 41 66 52 69 55 40 42

64 59 40

53 47 64 52 72 58 40 42 64 61 40

60

Legitimacy

government is highest for the regional level (r = ,88) but the absolute differences remain very large, as MEPs generally attach very little importance to this level. Summing up, there is a reasonably high consensus between the opinion of national and European MPs and the citizens as far as the importance of the European level for solving the most important problems is concerned, but this consensus is still considerably lower than the one about the saliency of issues. We conclude that the current policy-making competencies of the European Union are seriously at odds with citizens' preferences, This discrepancy between perceived and preferred EU policy competencies is likely to shrink. The Maastricht convergence norms have reduced national sovereignty in setting policy objectives concerning budgetary deficits and inflation, the main topics included in our policy domain 'public finance'. Add to this the search for harmonization of fiscal systems. On the other hand, the European Union is still a giant on clay feet with regard to a common foreign policy, the sector for which Europeanization is wanted most. Also in environmental matters, the regulating role of Europe is still rather limited in comparison with other sector policies, especially with regard to implementation of EC directives (Nugent 1994: 287-287). Still, there is not a single sector for which European citizens want less European government. In other words, in all sectors mentioned above, European citizens would generally welcome steps towards further Europeanization of at least some (major) competencies.

I N D I V I D U A L AND STRUCTURAL DETERMINANTS OF LEVEL-OF-GOVERNMENT PREFERENCES

In this section we shall elaborate and test a number of hypotheses on possible determinants of preferences for a particular level of government. The discussion of the subsidiarity principle pointed to the importance of the endogenous nature of policy problems for allocating competencies to different levels of government. Respondents who mentioned as important those issues that can only be solved at the national or sub-national level will tend to prefer these levels as the most adequate, while issues with cross-border features or international mul-

The Scope ofEU Government

61

tilateral conflicts will incite more preferences for the European level15 Inspired by the criteria defined by the European Council for according competencies to the European Union following the principle of subsidiarity,16 we classify the problems mentioned by the respondents in four categories on the basis of their endogenous attribution: 1. genuine national matters;" 2. problems that are basically situated at the national level but whose causes or solutions are partially related to similar problems and solutions in other countries (and therefore may distort competition, weaken social and economic cohesion, or restrict trade).18 Amongst the problems mentioned in the survey, this category predominantly refers to issues regarding the labour market, social security, fiscal policy, and public expenditures (as all these influence a country's competitive position); 3. problems with important cross-border aspects as well as problems with international but basically bilateral dimensions.19 Some of these issues may have genuine (but not predominant) international dimensions;20 4. problems with predominantly genuine international dimensions. Sometimes this dimension is not due to the cross-border nature of the problem, but to long-lasting international co-operation in the policy sector such as the Common Agricultural Policy.2' One cannot expect legitimacy of EU governance to be determined only by the endogenous nature of policy problems. Citizens' consent to the rules of the game of this regime, including the division of labour between different levels of governance in the EU, is also expected to be important. This consent will be facilitated on the one hand by a general positive evaluation of EU integration processes and, on the other hand, by general and diffuse disaffection with national or regional decisionmaking bodies (Easton 1965). In fact, since the supposed ebbing away of the permissive consensus on European integration (Lindberg and Scheingold 1970: 41-2; Niedermayer 1995a), the legitimacy of the European Union as an increasingly important decision-making body has become a 'variable', fluctuating in time and space, between but also within member-states.22 Therefore, we expect that the preference of European citizens for a particular level of government will be related to the general support for the respective regime level, and its democratic legitimacy (Dahl 1989: 109). Generalized support for the

62

Legitimacy

European regime will be measured by the following variables (for a similar procedure, see Schmitt and Scheuer 1996): - the degree of agreement with European integration; - satisfaction with democracy in the European Union; - the degree of sorrow one would feel if the EU were dismantled. Support for the national regime is measured by the degree of satisfaction with democracy in one's own country.2j As argued in the introductory chapter of this volume, there could be a spillover of support for specific policies to the legitimacy of different levels of decision-making. Hence we expect citizen preferences for levels of government to be affected by their evaluation of the performance of EU institutions with respect to their own personal well-being and that of their country. In addition the incapacity of national or regional governments to produce satisfactory policy outputs is expected to have an impact on the appreciation for the EU. This 'specific' support for the outputs of the EU will be operationalized by the following variables: - the perception that one's country has benefited from EU membership;24 - the degree of confidence in EU decisions; - the strength of (dis-)agreement with EU membership of one's country. Third, we can derive from Niedermayer and Sinnott (1995) a number of individual attitudinal and socio-demographic characteristics that are expected to have an impact on preferences for different government levels; - the degree of political interest in EU and national politics: those that display a relatively high degree of interest in EU politics will more easily prefer EU decision-making, as only political levels of which one has a minimal degree of understanding can be expected to be considered appropriate as a level of government; - 'state-wide' nationalism: people with strong national feelings will tend to oppose more the delegation of powers to other levels of decision-making;25 - left-right attitudes: one can expect that left-orientated people, given the association between internationalism and socialism, will give more preference to the European level, while right-wing respondents

The Scope ofEU Government

63

will prefer the national level, given the general association between national conscience and a conservative outlook (Gabriel 1994; 112; Huber and Inglehart 1995: 84); - education: respondents with higher levels of education may tend to grasp more easily the international dimension and interdependency of problems; - age: we expect young people to give more preference to the European level given their more open attitude to the world than older generations (Reimer 1992; Gabriel 1994: 112). Finally, Niedermayer and Sinnott (1995) deal with a number of structural features that can be expected to affect preferences for government levels: - the size of a country: the smaller the country, the less appropriate the level of the national state objectively is and will be perceived; - the length of EU membership (Niedermayer 1995a): the transfer of competencies from the national level to the European Union is a unique but also painful learning process. It has created a democratic deficit that the expansion of the role of the European and national parliaments has not managed to bridge. This deficit is more likely to be tolerated when the transfers are gradual in time and extent. This has been the case for the citizens of the six founding members of the European Community for Coal and Steel as they have been socialized for nearly half a century into the process of dismantling national sovereignty 'slice by slice'. The democratic deficit costs are higher for newcomers who in a short span of time are asked to give up large parts of competence hitherto falling under national sovereignty; - the openness of a country's economy:26 the more open an economy is, the more people will be aware of the necessity of international cooperation, for instance in order to better face the challenge of internationalizing economic decision-making in the private sector; - the objective cost/benefits ratio of membership: citizens in countries that receive more from the EU budget than they contribute to it may support the European decision-making level more than those that are net payers.21 Obviously, many of the variables mentioned above may be interrelated. The general positive attitudes towards European integration may be influenced by perception of output benefits, and all attitudes may be affected by the structural variables. In order to estimate the net direct

64

Legitimacy

effect of these variables on the dependent variable, we have analysed all the potential determinants mentioned above in a multinomial logistic regression. The variables that the analyses retained as significant are presented in Table 3.4, As the BIC'k indicates, the fit of the global model result is superior to the base line model including only the intercept. As one can see from the probabilities, all variables included in the model are highly significant. The Pseudo R2 indicates that the model explains 42% of the initial log likelihood, while the Chi2/df can be interpreted as a measure of relative importance of the retained variables. TABLE 3.4. Logistic regression on preference for European government level (maximum likelihood analysis of variance) Variables

df

X2

Prob-

Intercept Duration EU membership Endogenous nature of problem Population size National vs. European feeling Sorrow if EU dissolved Education EU-budget benefits Left-right Pro-EU integration Confidence EU decisions Satisfaction national democracy

3 9 9 3 3

30.86 212,23 181.69 212,23 46.73 56.35 21.84 17.01 16,74 44.40 41.94 22.67

0,0000 0,0000 0.0000 0.0000 0.0000 0.0000 0.0001 0,0007 0.0008 0.0000 0.0000 0.0070

X2/df

23.58 20.19 19,80 15.57 6 8,38 3 7.28 5.67 3 3 5.58 9 4,99 9 4.66 2.52 9 29 Source: European Election Study 1994. Pseudo R2 M(:F= 0,42; BlC'k = -12,241,6,

Duration of membership, population size, endogenous nature of the problem, and national orientations are the most important variables for explaining the degree of support for problem-solving at the European level. These variables are followed by the feeling of sorrow if the EU were to be dissolved, education, EC budget benefits, left—right attitudes, and general support for European integration. Still significant but least important is satisfaction with national democracy. Table 3.4 indicates that structural as well as individual characteristics contribute to the explanation of the variation in the dependent variable.30 Let us now examine the general direction of the relationship of the retained variables with the dependent variable (in order of importance

The Scope ofEU Government

65

of the independent ones). There is a non-linear relationship between length of membership and support for European decision-making: citizens of countries that entered in the 1980s (Greece, Spain, and Portugal) are most positive about the EU as decision-making level, followed by the population of the founding member-states. The 1970s entrants (UK, Ireland, and Denmark) are the least positive. This tendency confirms the findings of Scheuer in the preceding chapter, and those of Niedermayer (1995a: 59-67) on the difference in general indicators of support between member-states in terms of their entry to the EC. The positive attitudes of Greece, Spain, and Portugal are attributed to the beneficial effect of EC membership on the democratization process (Niedermayer 1995a; 66). Still, we have to note that the linear relationship as expected in the hypothesis does not hold. A quite linear relationship emerges between the dependent variable and the endogenous nature of the problem in the sense that, if the nature of the problem is exclusively national, problem-solving is preferred at the national level, while for genuine international problems the European level is strongly preferred. This finding suggests that in order to gain legitimacy, authorities of the different government levels should make more of an effort to explain to the public the inherently European/national/regional nature of problems. Population size is slightly negatively but non-linearly related to pronational-level preferences. Smaller and medium-sized countries have a more pro-European attitude than the larger ones. National orientations are more linearly related; those with an exclusively national orientation are also least in favour of the EU decisionmaking level. The sorrow one would feel if the EU were dissolved is also rather linearly related to the level-of-government preference: those who would be the most relieved also strongly prefer the national level, and those who would be most sorry prefer the European level. A higher level of education stimulates a preference for decisionmaking at the European level: those who left school early favour national government as the appropriate decision level more than those who studied longest. The effect of net contributions to the EC budget is again non-linear and suggests that citizens of big net contributors to the budget give more preference to the national level. Right-wing attitudes tend to produce a preference for national government competence while a more left-wing attitude goes together with

66

Legitimacy

a European preference, and centrist attitudes reflect the overall population distribution. People who oppose European integration and have little confidence in European decision-making most strongly prefer the national level. Finally, satisfaction with national democracy does not confirm our hypothesis: those that are least satisfied with democracy in their own country do not prefer the European level as an alternative for bad government in their own country more than the most satisfied. On the contrary, they still prefer the national level. This could indicate that this attitude reflects a generalized protest or distrust of politics, and therefore this feeling may even be stronger for political bodies that suffer from a democratic deficit like the EU,3! Note however that this factor exerts the smallest impact of all the variables retained. To conclude, apart from the last variable, the findings all confirm the direction of the presumed relationship between dependent and hidependent variables formulated in the hypotheses above. Note however that a few factors deemed relevant were not retained by the model: among them are satisfaction with democracy in the EU, the perception that one's country has benefited from EU membership, degree of political interest in (EU) politics, age, and the openness of a country's economy.

CONCLUSION

The question as to which decision-making level is most appropriate to deal with different policy problems is a central aspect of the legitimacy of a political system. With regard to the level of governance in the EU, the discussion is structured around the meaning and application of the principle of subsidiarity, permitting EU action when member-states or regions cannot sufficiently act to solve a problem, or when for reasons of scale or spillover effects, actions could be better undertaken by the EU. International governance, like national or regional government, can only gain legitimacy when the public agrees with the rules of the game including the division of labour between different levels of governance. Therefore, the opinion of European citizens regarding the appropriate level of government in the EU (European, national, regional) is crucial for the legitimacy of a given or future division of power be-

The Scope ofEU Government

67

tween these levels. The level-of-government discussion also poses a question about the quality of political representation, as national MPs control the transfer of competencies from the national to the European level. Hence, the appropriate level of government constitutes an issue on which opinions of the represented and their representatives can diverge. Our analyses show that European citizens have a clear opinion regarding the appropriate level of decision-making for the problems they consider the most important in their society. They also perceive accurately the current level of decision-making for these matters, Twothirds of these problems are believed to be currently decided at the national level; the other third is about equally distributed between the European and regional levels. However, in normative terms one finds that nearly four out of ten respondents view the European level as the appropriate level to solve problems, which is more than twice as much as the perceived 'real' level of decision-making. Therefore, these findings contradict the all-too-popular visions of a growing Eurofatigue and Euro-scepticism amongst European publics. This desire to solve more problems at the European level is stronger among the elites than among the public at large, but differences are not so important as to pose a problem of inadequate representation. Policy domains that are already perceived as most Europeanized are those for which citizens demand further Europeanization: European integration, international conflicts, and the environment. However, there is not a single sector for which European citizens want less European government. Citizen and elite opinions are rather congruent in terms of the level of government they prefer. As a consequence, the current competencies of the European Union are somewhat at odds with both citizens* and elites' preferences. A number of hypotheses concerning the variation in citizens* preferences for EU governance have been tested. The length of a country's EU membership, its population size, the endogenous nature of policy problems, and national orientations are the most important variables for explaining the degree of support for problem-solving at the European level. These variables are followed by the sorrow one would feel if the EU were to be dissolved, level of education, the degree a country benefits from the EU budget, left-right attitudes, and general support for European integration. Least important is satisfaction with national democracy.

68

Legitimacy APPENDIX

Seven of the eleven policy problems included in the questionnaire were offered to all European respondents, while four were formulated by country specialists and presented to each national sample only. The latter problems represent 35% of all issues mentioned and therefore cannot simply be regrouped in a single category of 'national issues'. Including these national issues complicates an analysis per policy sector as well as country comparisons, as in some cases the countryspecific issues are difficult to regroup in larger categories, due to their unique character (such as the divorce issue in Ireland). Still, for nearly all issues we found a specific common denominator. Unemployment and economy include 'fighting unemployment' and the following country-specific issues: Denmark: 'reforming the labour market'; The Netherlands: 'strengthening Dutch industry'; Portugal: 'promoting economic development'. Public finance includes 'securing stable prices' and countryspecific issues: Denmark: 'reforming taxation'; Germany: 'solving the fiscal crisis'; Ireland: 'introducing a fairer system of taxation" and 'introducing sensible controls on public spending'; UK: 'keeping taxes low'; Greece: 'reducing the cost of living'; Belgium and Italy 'reducing public debt'. European integration includes 'further unifying Western Europe' and 'reducing agricultural surpluses'. Immigration includes 'solving immigration problems* and in Luxembourg 'allowing citizens of other EU countries to vote and run in general elections'. Environment includes 'protecting the environment'. Law and order includes "fighting crime', 'fighting terrorism' in Spain, and 'fighting drugs' in Greece and Portugal. Welfare-state issues include only country-specific ones: Denmark: 'preserving the welfare state within the EU*; Germany: 'securing the welfare state'; Spain: 'keeping social protection (pensions; social security)'; The Netherlands: 'fighting abuse of social security*, 'ensuring good old-age pensions'; Portugal: 'increasing social protection'; Belgium: 'securing pensions in the future'; Luxembourg: 'harmonizing the pension schemes in a socially justified manner'; France: 'fighting so-

The Scope ofEU Government

69

cial inequalities', 'securing social protection (health care, pensions, etc,)'; UK: 'reforming the health service'. Centre-periphery conflicts include only country-specific issues: Belgium: 'maintaining the unity of Belgium'; Germany: 'promoting equal living conditions in East and West Germany'; Spain: 'solving the problems between autonomous regions and the State'; France: 'securing the balance between cities and the countryside'; UK: 'solving the Northern Ireland problem'; Italy: 'alleviating north-south differences'. International conflicts include only country-specific issues: Germany: 'peaceful democratization of Eastern Europe'; Greece: 'dealing with the Macedonian question' and 'improving relations with Turkey'; Ireland: 'achieving peace in Northern Ireland'. Improving democracy includes only country-specific issues: Belgium and Spain: 'fighting political corruption*; Denmark: 'improving moral standards in politics1; The Netherlands: 'bringing politics closer to the citizens'; Italy: "effectuating institutional and electoral reform', 'establishing rules for the television sector'; Luxembourg: 'depoliticizing the written press'; Portugal: 'fighting corruption in politics and economies'; UK: 'enacting a Bill of Rights'. Education includes 'improving education and training to prepare for the future' in France and 'developing a national university' in Luxembourg. Finally, the Irish national issue 'tackling the divorce issue' could not satisfactorily be classified and has been dropped, given the fact that very few Irish respondents saw this as an important problem.

NOTES

1

2 3 4

For an overview of the opinions of classical philosophers (Aristotle, Thornas of Aquinas, Althusius, Hegel, de Toqueville, and Montesquieu) on the principle of subsidiarity, see Millon-Delsol (1992, 1993) and du Granrut (1997: 13-16), See for instance the European Charter of Local Self-Govemment, drafted by the Council of Europe, Strasbourg, on October 15, 1985 (Duff 1993), For an overview of the other factors that may explain the expansion of the EU, see Schrnitter( 1996). The president of the Commission, Jacques Delors, launched this possibility of introducing the principle of subsidiarity in the discussion on the di-

70

5

6

7

8

9

10 11 12 13

Legitimacy vision of competencies. For an overview of the meaning of subsidiarity in the context of European integration, see EIPA (1991), CEPR (1993), Maillet (1993), Hrbek (1995), Lourau (1997), Wilke and Wallace (1990), and Estclla de Noriega (1997). For an interesting attempt at a definition, see the Giscard d'Estaing report (1990) prepared for the Institutional Affairs Committee of the European Parliament, which applies criteria of effectiveness of task accomplishment of the Community vis-a-vis the member-states, and the transnational nature of the task, whose dimensions or effects extend beyond national frontiers, The fight against drugs has a stronger international or cross-border dimension in Belgium, France, and Germany than in Portugal or Austria, given the permissive Dutch policy on soft drugs. Pollution of rivers is a more international issue in Holland, as all its rivers pass first through other countries, whereas in the insular GB and Ireland this would be much more a national issue. Immigration may be more of a European issue in Schengen countries than in others, Luxembourg is probably too small to conduct an effective monetary policy. In federal, states like Germany, the state governments are much better equipped to deal with environmental problems than the sub-state level in non-federal states (Massart-Pierard 1998), Niedermayer and Westle (1995: 50) classify 'functional scope' as an element of the 'policies object' of orientations towards international governance. For an analysis of differences in elite and citizens* views on the ongoing federalization of the Belgian state, see De Winter and Frognier (1997). 'Which of the following issues do you consider the most important, which comes second, and which comes third? 1. "National issue. 2. Fighting unemployment. 3. Securing stable prices. 4. National issue. 5. Furthering political unification of Western Europe. 6. Solving the immigration problem. 7, National issue. 8. Reducing agricultural surpluses. 9. Protecting the environment, 10. National issue. I I . Fighting crime.* 'Do you think that this issue is mainly treated at the regional, national, or European level?' 'Which level of government do you think would be most appropriate in dealing with what you mentioned as the most important problem, the second most important problem, the third most important problem (regional, national, European)?* For an analysis of country differences of these non-response rates, see Schmitt and Scheuer (1996; 166-7). See for instance the Eurobarometer Trends (1994: 161—172). For an interpretation of this lack of political interest, see van Deth (1990). The Swedish national MPs were interviewed, but were not asked this question. In contrast to the face-to-face interviews of voters and MEPs, the national MP interviews conducted via mail surveys contained a small number of

The Scope ofEU Government

14

15

16

17

18

71

combined answers (2% of the valid responses): respondents had filled in two or three levels rather than one. The percentages for these multiple response cases have been proportionally allocated to the single-level categories (if mentioned in combination with one other level, the corresponding level has been inflated with half of the frequency percentage of that level). This method allows us to standardize the columns in Table 3.1. The strong preference of MEPs for the European level is corroborated by their attitudes on two similar matters. The elite survey first asked 'One of the major issues presently facing the EU is increasing the range of responsibilities of the European Union. Do you favour or are you against increasing the range of responsibilities of the EU?' Of the MEPs, 71% were in favour, 17% were against, while 12% were situated at the middle position. Surprisingly, national MPs are also predominantly in favour of such an increase (62% for, 22% against, and 16% keeping the mean). For a first analysis of the effect of the endogenous nature of problems of level preferences of citizens, see Schmitt and Scheuer. For a different conceptualization of endogenous nature and its effect on elite attitudes, see Wessels and Kielhorn (1999). Conclusions of the Presidency of the European Council of Edinburgh of 1992 (Bull. EC 12-1992). For a discussion of the practicality of these criteria, see Estella de Noriega (1997). They include the following country-specific issues: Belgium: 'maintaining the unity of Belgium', 'fighting political corruption'; Denmark: 'improving moral standards in polities'; Germany: 'promoting equal living conditions in East and West Germany'; Italy: 'alleviating north-south differences', 'effectuating institutional and electoral reforms', 'establishing rules for the television sector'; Spain: 'fighting political corruption', 'solving the problems between autonomous regions and the State'; France: 'improving education and training to prepare for the future*; Ireland: 'tackling the divorce issue*; Luxembourg: 'developing a national university", 'depoiiticizing the written press', 'fighting political corruption'; The Netherlands: 'fighting the abuse of social security', 'bringing politics closer to the citizens'; Portugal: 'fighting corruption in politics and economies'; the UK: 'enacting a Bill of Rights', These issues include 'fighting unemployment' and 'securing stable prices* and the following country-specific issues: Belgium: 'reducing public debt*, 'securing pensions in the future'; Denmark: 'preserving the welfare state inside the EU', 'reforming the labour market", 'reforming taxation*; Germany: 'solving the fiscal crisis', 'securing the welfare state*; Greece: 'reducing the cost of living"; Italy: 'reducing public debt*; Spain: 'keeping social protection (pensions, social security)*; France: 'fighting social inequalities', 'securing social protection (health care, pensions, etc.)*; Ireland: 'introducing a fairer system of taxation*, 'introducing sensible controls on public spending'; Luxembourg: 'harmonizing the pension schemes in a socially justified matter*, 'ensuring pensions'; The Nether-

72

19

20

21

22

23 24 25 26

27

Legitimacy lands: 'strengthening Dutch industry*, 'ensuring good old-age pensions'; Portugal 'promoting economic development', 'increasing social protection*; the UK: 'keeping taxes low", 'reforming the health service*. These issues include 'solving the immigration problem*, 'protecting the environment', and 'fighting crime" and the following country-specific issues: Greece: 'fighting drugs', 'dealing with the Macedonian question', 'improving relations with Turkey'; Spain: 'fighting terrorism*; Ireland: 'achieving peace in Northern Ireland'; Luxembourg: 'allowing citizens of other EU/EC countries to vote and run in general elections'; Portugal: 'fighting drugs': UK: 'solving the Northern Ireland problem'. For instance, with regard to the environment, Golub (1996) rather convincingly argues that most issues of environmental protection can be solved at the level of the national state, as many do not have a significant cross-border character. Also in terms of market distortion, competitive strategies based on ecological dumping have negative effects for those member-states that employ them, and therefore market distortion only occurs in the short run. These include 'furthering political unification of Western Europe', 'reducing agricultural surpluses', and one country-specific issue 'peaceful democratization of Eastern Europe' in Germany, In some countries such as Spain, Belgium, and Italy also sub-national actors and authorities challenge the legitimacy of the nation-state as decision-making level. For a discussion of the substantive meaning of this variable and its impact on European integration attitudes, see for many others Martinotti and Stefanizzi (1995). For a discussion of the substantive meaning of this variable, see Bosch and Newton (1995). Operationalized as the degree of feeling European (in addition to one's identification with the nation). Measured by the share of exports in the Gross National Product in 1994 (OECD 1996: 192-5). The import shares correlate strongly with export shares, so either one could be used. We do not use the difference between imports and exports (as used by Wessels and Kielhorn in the second book) because the absolute degree of openness can be expected to have the major impact on citizen attitudes[, not the balance of payments (mostly unknown to most citizens)]. When a country's imports and exports range around a fifth of GNP, which is the case in Greece, France, and Germany, the need to be internationally competitive (and all its consequences in terms of wage and price setting) is less a constant ordeal and worry than in countries where imports and exports constitute three-fourths or more of the GNP (as in Belgium and Luxembourg). As measured by the differences between contributions to and returns from the EU budget of 1994 (data calculated by Kerremans (1996) based on the figures used for calculating the amount of repayments within the 'British

The Scope ofEU Government

28

29 30

31

73

arrangement', published by the Official Journal of the EC, C 303, 14 November 1995). The dependent variable counts the number of mentions of the European level for all three, two. one, or none of the most important problems; respondents mentioning the regional level are excluded, so that the dependent variable essentially measures preferences for the European vs. the national government level as most appropriate for policy-making. The following are metric independent variables: duration of membership, budget benefits, population size, openness of economy, age, feeling European, and left-right attitudes. For the calculation of these measures, see Long (1997). In this analysis the size of the country samples has been set equal in order to prevent the risk of contaminating structural country characteristics by population size. For the variables measured at the individual level this procedure may certainly be questioned. Therefore, we also analysed the direct net effects of the non-structural variables only. In this step, the cases were weighted taking the size of the country fully into consideration. The model based on these variables and weights retains exactly the same individual-level variables as the first one. Parameter estimates were only slightly different, standard errors being identical. The relative importance of the variables retained did not change in order, but the overall Pseudo R2 diminished slightly to 0.38, as can be expected from the omission of the structural variables in the second model. See Martinotti and Steffanizi (1995) for a similar result.

4

The Political Regime

PIPPA MORRIS

THE third element of legitimacy distinguished in the introductory chapter refers to the political institutions of the European Union and its policy processes. In this chapter we will first elaborate on these aspects of legitimacy or regime support. We will go on to examine crossnational variations in regime support, including public satisfaction with the basic principles, decision-making processes, and institutional arrangements of the European Union. The subsequent section analyses alternative explanations for these variations in regime support, including the role of cultural, social background, performance-, and leadership-based factors. Lastly, the conclusion considers the implications of this analysis for understanding the future of the European Union. A CONCEPTUAL FRAMEWORK FOR POLITICAL SUPPORT The most common way of understanding the concept of political support derives from David Easton, who distinguished between attitudes towards the political community, the regime, and the authorities (Easton 1975; see also Niedermayer and Westle 1995). Subsequent work has expanded this framework (Norris 1999a; Dalton 1999). This chapter uses the following classification of the objects of support: (1) political community, (2) regime principles, (3) regime processes, (4) regime institutions, and (5) political authorities. Though it has not been clearly established, it is generally believed that these objects of support are mutually related. For example, the ero-

The Political Regime

75

sion of trust in politicians may eventually undermine confidence in parliament and general elections; and deficient regime support in the long run may even affect the evaluation of the political community. While other parts of this volume analyse different components, this chapter focuses on political support for the EU as a political regime, including attitudes towards the core principles, processes, and institutions of the European Union, Let us start by clarifying these conceptual distinctions. According to Easton, support can be expressed at the most general level through a sense of belonging to, and identification with, $\& political community, whether at local, regional, national, or international levels. This aspect was taken up by Scheuer in Chapter 2; she found a substantial 'sense of community' among EU citizens which, however, is not independent from disturbances. Support for the basic principles of the regime concerns agreement with the predominant values which underpin the political system. Where the regime is democratic, adherence to core values is often gauged in surveys by approval for democracy 'as the best form of government', or 'as a good way of governing'. Support for the basic principles of the EU however are commonly measured by attitudes towards European unification, membership, and co-operation since these values lie at the heart of the European project (Niedermayer 1995a). Support also concerns evaluations of the way the policy process functions. This component is gauged in this chapter by satisfaction with the way democracy works within the European Union and within each member state. This item is not an ideal performance indicator. As Inglehart (1999) argues, it taps both support for democracy as a value (i.e. regime principles) and satisfaction with regime performance. Nevertheless this question has a track-record since 1973 in the Eurobarometer series, it allows comparison between national and European levels of governance, and the emphasis on 'how democracy works' means it does tap evaluations of processes. Another component of regime support relates to the major institutions of the political system. In the EU, institutional support can be measured by trust in bodies such as the European Parliament and Council of Ministers. This focus emphasizes formal structures, not specific incumbents or office holders. Lastly, although not discussed in this chapter, support may also be directed towards political authorities and their policies. This aspect

76

Legitimacy

will be dealt with in the following chapter. It is relevant for the present chapter because some degree of mistrust of authorities can be regarded as a normal part of democratic politics; a more general disenchantment with politicians however may gradually erode confidence in the core institutions of representative government, faith in the democratic process, and perhaps even support for the basic principles of the regime. Our data show that the general public is able to distinguish between regime principles, processes, and institutions. Factor analysis identifies three dimensions in citizens' EU orientations: attitudes towards regime principles (based on evaluation of EU membership, European unification, and EU dissolution) fall into the first; attitudes towards regime processes (measured by satisfaction with the working of democracy within the EU and within each country) fall into the second; and confidence in EU institutions (the European Commission, European Court, Council of Ministers and European Parliament) fall into the third (Table 4,1). TABLE 4.1. Dimensions of EU regime support Regime principles Membership Unification Regret dissolution Satisfaction democracy own country Satisfaction democracy EU Confidence Confidence Confidence Confidence

in Council of Ministers in European Parliament in European Commission in European Court

Regime processes

Regime institutions

.84 .80 .75 .86 .74

.80 .79 .78 .70

Source: Eurobarometer 40 (Fall 1993). Factor analysis based on principle component analysis, Varimax rotation with Kaiser's normalization. Coefficients betow ,3 not shown. The analysis of EB42 data produced an almost identical solution. Evidence from the World Values Study 1995/7 suggests there are different trends in public support for each of these components at the level of the national State (Norris 1999). Throughout Western Europe faith in democratic values and principles appears to be widespread and

The Political Regime

77

pervasive, and these beliefs seem to have strengthened over time (Klingemann 1999; Thomassen 1995). Support for regime processes, measured by satisfaction with the working of democracy in one's country, displays a broad pattern of trend-less fluctuations over time, with persistent cross-national differences between countries such as Italy and Denmark (Fuchs, Guidarossi and Svensson 1995). In contrast, confidence in governing institutions, and perhaps also trust in political leaders, seems to have been seriously eroded in many Western countries, and to be at relatively low levels in many transitional and consolidating democracies (Klingemann 1999; Dalton 1999; Inglehart 1997; Listhaug and Wiberg 1995). In the following section we will analyse over-time and crossnational variations for these dimensions of regime support.

CROSS-NATIONAL AND OVER-TIME VARIATIONS IN REGIME SUPPORT 1. Support for Regime Principles Support for the basic principles of the European Union can be measured by attitudes towards European unification, membership, and dissolution (Niedermayer 1995a). Standard items with a long pedigree in the Eurobarometers include whether respondents were for or against European unification, whether they believed their countries' membership in the Union had been a good or bad thing, and whether they would be sorry, indifferent, or relieved if the EU was dissolved. Figure 4.1 displays the time series of these indicators, EU-wide, for more than a quarter of a century. We find that the evolution of the membership and the dissolution indicator run very much in parallel, while the unification item is somewhat more robust. In very broad terms, and disregarding minor composition effects due to enlargements of the Union, we identify a declining trend of support for EU principles during the 1970s and, even more pronounced, the 1990s, while the 1980s saw a steady growth of support of EU citizens for these regime principles. There is also remarkable cross-national variation, as shown in Table 4.2. Support for European principles varied widely across member-

78

Legitimacy

Source: Eurobarometer surveys.

FIG 4.1. The development of support for EU principles: unification, membership, and regret dissolution (displayed are yearly averages)

The Political Regime

79

TABLE 4.2. Support for European principles (per cent 'for' unification, membership 'good', and 'sorry* if EU dissolved)

Luxembourg Ireland Italy Greece The Netherlands France Belgium Germany Portugal Spain Finland Denmark

UK Austria Sweden

EU15

Unification

Membership

Dissolution

Average

84 89 93 88 82 79 82 74 87 84 70 60 69 64

82 84 75 69 79 60 62 60 58 51 51 54 51 36

73 65 62 54 48 48 36 44 32 40 34 41 28 39

80 79 77 70 70 62 60 59 59 58 52 52 49 47

58

43

29

43

77

60

43

60

Source: Eurobarometer 42 (Winter 1994).

states, with Luxembourg, Ireland, and Italy as the most Europhile and Sweden, Austria, and the UK as the most Eurosceptic. 2, Support for Regime Processes Support for the general processes of EU decision-making is commonly measured as satisfaction with the working of democracy in the European Union compared with the working of democracy in the different member-states.1 The results, in Table 4.3, show that overall less than half the public (48%) were satisfied with the workings of democracy in the European Union, while slightly more (57%) expressed positive evaluations of the democratic process in their own country.

80

Legitimacy TABLE 4,3. Satisfaction with democracy (per cent 'very' and 'fairly' satisfied)

The Netherlands Denmark Luxembourg Germany France

UK Finland Belgium Portugal Greece Spain Ireland Italy EU 12 + Finland

EU

Own country

Difference

40 61 67 39 46 44 50 6! 54 39 44 81

64 84 86 53 59 54 59 60 49 32 35 71

24 23 19 14 13 10 9 -1 -5 -7 -9 -10

49

26

-23

49

57

8

Source: Eurobarometer 42 (Winter 1994).

But the pattern varied again substantially among member-states. In many of the Mediterranean countries there was greater satisfaction with the EU than their own country, including Italy, Spain, Greece, and Portugal, as well as Ireland. In contrast citizens in Northern European states expressed more positive attitudes towards democracy in their own country rather than in the EU, with this pattern particularly marked in The Netherlands, Denmark, and Luxembourg. 3. Support for Regime Institutions To examine trust in EU institutions we can compare trust in the core decision-making bodies in the Union (including the European Parliament, European Commission, the Council of Ministers, and the European Court of Justice) with confidence in comparable national bodies.2 The following table compares confidence in parliaments.

The Political Regime

81

TABLE 4.4, Confidence in parliaments (per cent 'can rely on') European Parliament

National parliament

Difference

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

39 26 29 38 24 55 55 36 29 53 50 47 51 51 46

62 47 48 55 39 68 67 48 37 56 53 42 46 45 29

-23 -21 -19 -17 -15 -13 -12 -12 -8 -3 -3 5 5 6 17

EU15

39

42

-3

Source: Eurobaroracter 45 (Spring 1996).

In Spring 1996, only 41% of European citizens felt they could rely upon the European Parliament. Slightly more (48%) expressed confidence in their national parliament; this level was also fairly low. There were clear country differences in the results. People expressed more trust for their national parliaments than for the European Parliament in eleven countries, often by a large margin. This feeling was particularly strong in the newer member-states with a long tradition of democracy in the post-war period, including Denmark, Sweden, and Austria, but this was also the view in some of the states which are traditionally the most pro-European, such as Luxembourg. This pattern was only reversed, with greater trust in the EP than in national parliaments, in Spain, France, Belgium, and Italy.

82

Legitimacy TABLE 4,5. Confidence in European and national government institutions (per cent 'can rely on') European Commission

Council of Ministers

National government

Difference (a)

Greece France Belgium Spain Italy

30 39 58 21 34 28 58 40 50 28 50 50 48 50 43

28 38 55 21 36 28 54 35 46 26 49 45 41 46 40

49 56 73 37 50 40 68 46 55 33 51 46 40 38 27

-20 -18 -17 -16 -15 -12 -12 -9 -7 -6 -2 2 5 10 15

EU 15

38

35

39

-3

Austria Denmark Luxembourg Germany Portugal Sweden Netherlands Finland Ireland

UK

Source: Eurobarometer 45 (Spring 1996). Difference between average confidence in European Comraission+Council minus confidence in national government

If we compare levels of confidence in national governments with the European Commission and Council of Ministers, we find a similar response (Table 4.5). Again just under half of the European public (48%) expressed confidence in their national government, while fewer citizens felt the same about the Council of Ministers (40%) and the European Commission (38%). People trusted their national governments far more than the European government institutions in most member-states including Austria, Denmark, Luxembourg, Germany, and Portugal. Only Spain and Italy saw a marked reversal of this pattern.

The Political Regime

83

EXPLAINING REGIME SUPPORT

1, Systemic Factors On this basis we need to explore alternative explanations to account for variations in support for the European regime. The first hypothesis to be considered focuses on cultural legacies in member-states. Evaluations of the process of decision-making and core institutions within the European Union can be expected to reflect long-established political traditions within each country. Countries in the European Union have widely different patterns of democratic development. Adopting Huntington's periodization, the 'first wave' of democracy swept over many advanced industrialized societies in Europe during the mid- to late nineteenth century, including France, Germany, Britain, and several smaller European countries (Huntington 1991). Italy followed shortly before World War I, Ireland on independence in 1922, and Spain in the very early 1930s, Yet the first reverse wave began in 1922 in Italy, followed by the rise of authoritarian governments in Germany, Austria, Greece, Portugal, and Spain. The second wave of democratization occurred after the Second World War, with new regimes in Germany, Italy, Austria, and (temporarily) in Greece. The final wave manifested itself in the early 1970s in southern Europe with the transitions in Portugal, Greece, and Spain (Gunther, Diamandouros, and Puhle 1995; Morlino and Montero 1995; Morlino and Tarchi 1996), followed in the late 1980s with the transitions in Central and Eastern Europe, notably in East Germany following the fall of the Berlin Wall. Given differing cultural legacies this has produced, evaluations of the regime legitimacy of the EU might be expected to vary substantially among member-states. In particular, we might expect that citizens in countries that developed democracies in the third wave may have come to associate membership in the European Union with the process of modernization and democratization, as part and parcel of the rejection of their authoritarian past. In contrast, countries that developed democratic regimes in the first and second waves, well before membership in the EU, may not make this association, and indeed may compare the institutional structures of representation in the EU unfavourably with their well-established domestic arrangements. We can therefore explore the first hypothesis that the length of time which a

84

Legitimacy

country has been democratic (measured by the number of years with a universal franchise for citizens) may prove part of the explanation for different evaluations of the legitimacy of the EU regime.

2. Social Background Alternatively, differences in the social background of citizens, which have usually proved significant indicators of political attitudes, may help explain support for the EU. Based on an extensive body of literature, regime support can be expected to vary according to the education, age, gender, and socio-economic status of citizens (see, for example, Wessels 1995; Handley 1981). Post-materialist theory suggests that generational differences are likely to prove particularly important (Inglehart 1997, 1977, 1990). We might expect the young and welleducated to be most supportive of the international ideals of European co-operation, while the older generation and the less educated may prove more nationalistic. We therefore examine how far the social background of citizens—-particularly their age, class, and education— helps to explain support for the European regime. 3. EU Performance Both systemic and social background explanations imply that if there is a widespread crisis of legitimacy in the EU, there is little that proEuropean political leaders can do to overcome this phenomenon. In this regard, low support for the EU may reflect a general crisis of public confidence in governance which afflicts many advanced industrialized societies. Yet there is an alternative explanation which draws on the Citrin/Miller debate in the United States (Miller 1974; Citrin 1974; Miller and Borrelli 1991), and the broader literature on 'government overload' of the 1970s (Crazier, Huntington, and Watanuki 1975; Huntington 1981; Lipset and Schneider 1987). The main explanation for dissatisfaction with the EU in this perspective relates evaluations of the regime to perceptions of its performance, particularly on the economy. In support of this thesis, Clarke et al. found that in Western Europe attitudes towards democracy responded to judgements concerning gov-

The Political Regime

85

ernments' effectiveness in managing the economy (Clarke, Dutt, and Kornberg 1993). Lockerbie found economic dissatisfaction to be significantly related to political alienation in Europe (Lockerbie 1993). Yet after a thorough review of the evidence, Bosch and Newton (1995) were unable to establish that economic performance per se was significantly related to support for the EU. The precise linkages between public expectations of the EU, evaluations of actual performance, and diffuse support for the regime therefore remains poorly understood. When applied to the context of the European Union, performancebased theories suggest that the problem lies with increasing public demands and the inability of the Union to meet these expectations. During the affluent 1980s, when Europe seemed to be booming, this did not seem to be a problem. But in the 1990s, the economic situation may have called into question the benefits of EU membership, and in turn this may have eroded diffuse regime support. The perceived failure of performance could be for many reasons, including the growing complexity of decision-making in the enlarged community, more intractable problems of European integration over issues like monetary union, and the problems of persistent unemployment, sluggish growth, and escalating welfare costs in most member-states. This suggests that we should examine the hypothesis that evaluations of the economic record of the EU are related to diffuse support for the regime. Perceptions of EU policy performance are measured by another standard Eurobarometer question, whether one's country has benefited from EU membership or not.3 If public confidence has declined because of the perceived performance of the European Union, then the solution lies in European leaders reducing public expectations, or improving policy performance, or both. 4. Opinion Leaders The last approach is provided by the classic 'two-step' model of public opinion, first proposed by Lazarsfeld, Berelson, and Gaudet (1944), and developed further into a 'cascade' theory of communication by Deutsch (1967). These theories stress that public attitudes towards the EU may be heavily conditioned by the leadership of political elites. The social psychological work of Paul Lazarsfeld suggests that the public develops fairly stable and well-formed attitudes towards certain

86

Legitimacy

issues, like jobs and prices, where they have direct experience in their everyday life. On the other hand, on certain complex and technical issues, particularly those of foreign affairs, the public may be influenced far more strongly by the cues given by opinion leaders. These may be trusted figures within the local community, media commentators and policy experts, or elected representatives. The complexities and the lack of transparency of EU decision-making may make it difficult for the public to evaluate the Union directly. As a result they may be strongly influenced by the elite (for a detailed discussion see Wessels 1995). To explore this issue we can analyse the hypothesis that the views of political elites about Europe are significantly related to those of the general public within each country. While Cross-sectional evidence at one point in time cannot by itself establish whether elected elites actually lead public opinion, or only follow. Nevertheless according to the two-step theory evidence of similarity of viewpoints can be regarded as indicative of elite influence. To estimate the influence of opinion leaders we measured how far members of the European Parliament in each country expressed strong support for the European Union. MEPs were asked how far they agreed or disagreed (on a seven-point scale) concerning a range of statements about Europe. After checking for internal reliability four items were scaled into a measure of support for Europe, and the mean scores for each country's MEP elite were then used as an indicator of opinion leaders' position within that country. The systemic and structural factors were entered first into the ordinary least squared regression analysis models, as long-term variables unlikely to change, followed by measures of EU performance and support for the EU among MEPs, Systemic explanations suggest that countries which democratized in the decade before they joined the Union would be more likely to support the EU regime than countries with long-established democratic traditions. The results of the regression analysis confirm that democratic legacies were significantly related to approval of EU democracy, although only weakly. Evidence from the 1994 Eurobarometer shows that citizens were more satisfied with democracy in the EU than in their own countries in many countries with experience of authoritarian regimes in Southern Europe, including Italy, Greece, Spain, and Portugal, as well as Ireland, In contrast, citizens were more satisfied with democracy in their nation-state than in the Union in many well-

The Political Regime

87

established democracies in Northern European countries, including Denmark, The Netherlands, Luxembourg, and France, although also in Germany, TABLE 4.6. Models of support for EU regime processes (figures are standardized coefficients from multiple OLS regressions) Model 1

Model 2

.04**

.02*

,08** ,04** .03** ,01**

.02* .02*

SYSTEMIC FACTORS Years democratic

SOCIAL BACKGROUND Gender Social class

Age Education (years)

POLICY EVALUATIONS OPINION LEADERS PARTISANSHIP Adjusted R2

.64** .05** .03*

.03

.40

Source: Eurobarometer 1994 and European Representation Study. See text for details of measures used. ** = p < .005; * = p < .05,

Turning to the social background of respondents, none of these variables proved to be strongly related to regime support for the EU, although gender, class, and age (but not education) remained statistically significant. The model suggests that the attitudes of opinion leaders were significantly related to public support for the EU, although the relationship remained modest. By far most important, however, is the policy performance of the European Union as perceived by the citizens.

88

Legitimacy

CONCLUSIONS The explanations explored in this chapter have different implications for the future of the Union. The most important conclusion from this analysis is that overall policy performance lies at the heart of (dis-) satisfaction with EU democracy. If people feel that there are direct benefits from membership then they are more likely to feel positive about the broad principles of the regime. In this regard, it can be argued, approval of the Union reflects 'normal' politics, which should not engender undue concern. If the Union manages to improve policy performance, for example by reducing the record levels of unemployment in many countries and stimulating economic growth, then dissatisfaction with the EU regime may fade. But if the serious policy problems facing European societies continue to be exacerbated under pressure for both enlargement and the project on European monetary union, then the problem of legitimacy seems likely to continue. This general pattern is open to alternative perspectives (Morris 1999). If the erosion of support for the core institutions of European government spreads upwards and starts to undermine faith in the democratic process and principles, this may eventually have serious consequences for the long-term stability of the EU. Alternatively, if this development signifies the growth of a more critical citizenry, who feels that traditional political institutions in Europe need to be brought into line with democratic ideals, this may ultimately strengthen European governance. NOTES 1

2

are: (1) very satisfied, (2) fairly satisfied, (3) not very satisfied, (4) not at all The questions used to assess this in the Eurobarometer are: 'On the whole, how satisfied are you with the way democracy works in the European Union?' *On the whole, how satisfied are you with the way democracy works in [YOUR COUNTRY]?' Answering categories for both questions satisfied. This can be measured by the following Eurobarometer item: "Many important decisions are made by the European Union. They might be in the interests of people like yourself or they might not. To what extent do you feel you can rely on each of the following institutions to make sure that

The Political Regime

3

89

the decisions taken by this institution are in the interests of people like yourself? The European Commission The European Parliament The European Court of Justice The Council of Ministers of the EU representing the national governments The (national) government The (national) parliament' Confidence in these bodies is measured on a ten-point scale, ranging from 'cannot rely on it at all' (1) to 'can rely on it completely' (10), which is receded into a dichotomy in the analysis reported in the text. This indicator is the central measure of EU policy performance of the following chapter, and is described and analysed there in greater detail.

5

Policy Performance

MICHAEL MARSH

INTRODUCTION

THE previous chapter has shown how important the policy performance of the EU is for the legitimacy of its institutions. This chapter will expand on this, examining evidence on how citizens judge the impact of the EU, and asking how these evaluations can be explained. It then goes on to consider the implications of the answer for the legitimacy of theEU. As suggested in the previous chapter, support for any system is likely to be affected by how well that system performs. In the most general terms David Easton's (1967) model of the political system included a 'feedback loop*, from outputs to inputs, inputs including both specific and general support for the political system, while Brian Barry's classic analysis of democratic stability argued for the importance of performance over political culture (Barry 1969). More specifically, functionalist and neo-funetionalist theorists argued that the development of supranational institutions, including the EU itself, would depend in part on how well tasks were performed, A central functionalist premise, used later by neo-functionalists, was that more authority would be given to an institution that performed well (Mitrany 1965), States and pseudo-states may be judged by how well they perform a service. Concerns about legitimacy 'crises* advanced in recent years by

Policy Performance

91

those on the left and the right have claimed that states are increasingly less able to do this, whether because of over-demand, because of the impossibility of squaring the circle of capitalism, or the weakness of its political institutions (Weil 1989). In any evaluation the reference group is important. Neofimctionalists originally assumed that performance would be judged in terms of an impact on functional social groups, who would then, via group leaders, press for more decision-making resources for the EU. This view was always problematic. Certainly it is possible that groups other than functional ones might serve as the unit for evaluation (Putnam 1981), or at least that national cleavages might cut across functional ones. Functional social groups—like fishermen—might be defined largely within a national rather than a European context. In that case political developments might -take a rather different form, with national governments as mediating institutions and people pressing their own government to secure even more national benefits from Brussels. Governments would then do their best to take credit for good things while shifting blame for bad ones onto the EU. Recent interpretations which have cast the EU as the saviour of the nation-state rather than its gravedigger (for example Moravscik 1994} argue that national governments have been able to confirm their own authority with the help of economic and other resources of the European Union. Even so, this argument surely does no more than relegate policy effectiveness from being a 'sufficient' condition for further integration to being a 'necessary' one. In more probabilistic terms, policy effectiveness is certainly not enough in itself to increase the likelihood of the EU becoming much more legitimate in the eyes of its citizens, but more legitimacy is unlikely to develop in its absence. Most studies have certainly demonstrated that popular evaluations of the EU differ significantly by nationality, but different conclusions are drawn from this. One recent analysis asserted that the fact that country was much more important in predicting popular support for European unification than socio-economic (functional) variables, or (postmaterial) value priorities, confirmed the importance of national rather than functional identity as a basis for the EU, refuting the 'myth' of 'postnationalism' (Deflem and Pampel 1996). However, it is a little dubious to see national differences in individual assessments as solely a function of national identity, even with various controls for social structure and political values in place. As many chapters in this volume

92

Legitimacy

illustrate, 'country' is a variable which can be unpacked, and it remains crucial to our understanding of the processes involved to do so (see Przeworski and Teune 1970). Simply noting that countries are different is little more than putting labels on our ignorance. Hence, while this chapter will examine the extent to which evaluations vary by country, it will also try to establish the source of that variation. The fact that countries vary is not simply a fact of academic interest. National units serve as one channel of representation in the European political system (see Chapter 1 of this book), and to the extent variations in support are between countries rather than within them, this channel is particularly appropriate. Policy outputs take a variety of forms. The central plank of functionalist theory was that people would be moved by economic outputs, or outputs with a direct impact on economic matters. Obviously, a growing volume of EU policy falls under this heading, especially if we include things like the social charter legislation impacting on conditions of employment (Schmitter 1996). However, as Stanley Hoffmann (1965) pointed out in a robust attack on the materialist bias of functionalist theory, there are other areas of policy whose impact and significance may be quite separate. He argued that matters of 'high polities'—foreign and security policy—would not be brought under supranational authority through the 'spillover* process. Events may have gone some way to proving him wrong, but what is clear is that these areas of policy-making have proved even more intractable than some others, and have certainly necessitated different, essentially intergovernmental arrangements for policy-making. Furthermore, political agendas are increasingly characterized by the appearance of 'new politics* issues, such as concerns about the environment, which accord neither with traditional economic political divisions nor necessarily with concerns about high politics. Certainly the scope of EU policymaking has been extended by the Single European Act and the Maastricht Treaty, and it is possible that evaluations of the EU's performance in some new areas may be very important for some groups unconcerned by other issues. The growth itself is seen by some to account for what is seen as significant opposition to the European project. One interesting analysis has argued that dissatisfaction with Europe is actually a direct response to increases in EU decision-making (and that satisfaction rises as fewer decisions are made) (Wlezien and Franklin 1997).

Policy Performance

93

While the growing range of policy outputs might be important, a significant body of research has argued that people's views of the EU have a strong materialist basis. The argument is hardly a new one (for example Shepherd 1975; Inglehart and Rabier 1978). However, with the time-series data now available for econometric analysis, several studies have concluded that variations in support for the EU over time can be explained in economic terms (Dalton and Eichenberg 1993; Eiehenberg and Dalton 1994; Anderson and Kaltenthaler 1995; Palmer and Gabel nd) although they differ over what economic factors are important. Anderson and Kaltenthaler (1995) argued that national evaluations of the EU could be seen as driven partly by national economic performance: national rates of growth, inflation, and unemployment. In good times the EU would get some of the credit; in bad times it would get some of the blame. This model of the impact of EU outputs on EU support is essentially a very indirect one, and because of that may be thought implausible in the absence of more direct evidence that people do perceive the EU to have a major impact on national economic wellbeing. Gabel and Palmer (1995) questioned the assumption that people did see the EU as having much impact on national economies, and cited evidence from a 1983 Eurobarorneter to the contrary. However, whatever the situation in 1983, the assumption seems reasonable in the late 1990s, as governments cite the imminence of European Monetary Union as a factor inhibiting their freedom of action in framing national economic policy. An alternative, more direct model underpins Dalton and Eichenberg* s analysts which examined the impact of both domestic economic performance and more direct measures of EU contributions to that performance: a country's benefits from the EU budget and the proportion of its trade that is intra-EU. Findings were mixed. Only inflation and intra-EU trade had significant effects on support.1 A more recent study by Palmer and Gabel (nd) which employed a much greater variety of estimation techniques achieved results similar to those of Dalton and Eichenberg, although the importance of inflation was not very robust. The link between objective well-being and political satisfaction is thought by many to depend on individual perceptions of well-being: a 'feel-good' factor (Nannestad and Paldam 1995), Eichenberg and Dalton (1994) explored the impact of perceptions of economic well-being on EU support and argued these largely mediated the effect of objective economic conditions: *in general it is more important to know what Europeans think about economic conditions rather

94

Legitimacy

than to measure the economy in objective terms' (Eichenberg and Dalton 1994:20). However, not all researchers have managed to discover any economic roots for EU support. Bosch and Newton, eschewing multivariate methods in favour of bivariate tables, argued against the thesis that the performance of the national economy prompted people to view European integration in more positive terms. However, their findings, when considered in conjunction with those of others, are better taken to indicate that the economy is only one of several effects rather than that it does not matter at all. With their methodology, and at best a dozen cases, any factor needs to be very strong to show up. Andrew Gamble's suggestion that 'popular attitudes towards Europe will usually be indirect, mediated through attitudes towards national elites* (Gamble 1989: 169; see also Franklin, Marsh, and McLaren 1994)—notwithstanding his assertion of a link between economic trends and perceived benefits—is a plausible explanation for why the link might not always be pronounced, and why the next step—support for more EU policy— might not always get taken. Even so, it is important to establish how far economic circumstances do impact on evaluations of benefits, and to identify the particular economic circumstances that are important. For the most part these analyses have been concerned with support for the EU, measured with a variety of questionnaire items, tapping support for European unification as well as the EU itself: 1, In general, are you for or against efforts being made to unify Europe? 2, If you were told tomorrow that the European Community had been scrapped, would you be very sorry, indifferent, relieved? 3, Generally speaking, do you think (your country's) membership of the Community has been a good thing, a bad thing, or neither good nor bad? These items have been analysed separately by some people (Anderson and Kaltenthaler 1995; Dalton and Eichenberg 1994); others have built an index of the interrelated items and use that as the variable to be explained {Palmer and Gabel nd). However, the validity of these items as an indicator of policy satisfaction is disputable. Another question asked in Eurobarometer polls since 1983 is the 'benefits' question: 'Taking everything into consideration, would you say that (your country) has on balance benefited or not from being a member of the European

Policy Performance

95

Union?' This has a clear national referent, unlike items 1-2. Moreover, item 3 is diffuse in its object, asking about the principle of membership as much as the practice. By contrast, the benefit question is a specific measure of policy satisfaction. The previous chapter, as measured by this indicator, showed significant implications of policy satisfaction for EU support in general. Here we will examine the public perceptions of benefits accruing from EU membership and seek to explain them.

P U B L I C PERCEPTIONS OF BENEFITS

I . Description The 'benefit* question is the most direct question about policy-outputs and their impact, and the only one asked on a regular basis. It should be emphasized that while the nature of the benefit is left up to the respondent, the reference group explicit in this question is the national unit rather than a functional social group such as farmers, or the respondent personally. It would certainly not be surprising were results to vary substantially by country. Figures 5.1 to 5.3 show the pattern of responses, recorded as those saying their country benefited minus those who say it did not. Hence this is treated as a national measure of relative benefit. The series runs from 1983-97, starting with EB 20 and running up to and including the EB of Spring 1997. With fifteen countries now members, any graph that tries to show them all individually is unlikely to do an adequate job. The countries are presented in three groups according to the timing of their joining the EU, but we have also presented on each graph the trend line for EU members as a whole. Figure 5.1 shows the picture for the original six member-states, Figure 5.2 for the three who joined in 1973, and Figure 5.3 the six who joined at various times after 1981.

96

Legitimacy

Scwee: Eurobarometers surveys.

FIG 5.1. Relative benefit from EU membership: the original six members

Policy Performance

Source: Eurobarometers surveys.

FIG 5.2. Relative benefit from EU membership: Denmark, Ireland, and the UK

97

98

Legitimacy

Source: EuFobarometers surveys. FIG 5.3. Relative benefit from EU membership: Greece, Spain, Portugal, Austria, Sweden, and Finland

Policy Performance

99

The overall pattern of assessment is positive for the whole period, but is certainly not stable. Whereas it was rising for most of the 1980s, the series has been declining since Spring 1991, and by 1997 positive evaluations outweighed negative ones by only a few percentage points; the index now stands at a lower point than at any time since it started. Quite clearly, the public perception of the EU's impact is increasingly negative. Furthermore, while there was a time when dissatisfaction was largely the preserve of newer members, it is clear that this is not the case in the 1990s. The original six have all followed a broadly similar path, fluctuating around a consistently high level of perceived benefit until the late 1980s and then declining in the 1990s. Positive opinion in The Netherlands and Luxembourg has declined only marginally, but it has done so rather more sharply elsewhere, and now hovers around the zero point in Germany, Italy, France, and Belgium. The experience of the three countries who joined in 1973 has been very different. Irish opinion was already positive in 1983, and has become steadily more so ever since. The net perception of benefit is now overwhelming. Notwithstanding strong opposition to the SEA and the Maastricht Treaty, Danish opinion has fluctuated around the EU average but has always stayed positive, and remains well above it, despite a fall in the last few years, British opinion followed a similar pattern of rise and then fall, but has tended to be negative. Although it moved into the 'black' between 1988 and 1992, a sudden drop post-Maastricht has left it well in the 'red' over the last few years. The experience of the post-1981 intake has been variable. In Greece and Portugal opinion grew more positive until around 1991, and although support has since declined, it remains very positive overall. By contrast, Spanish opinion has tended to be negative, if trending upwards. It was much more positive between 1989 and 1991, after a remarkably sharp increase in the second part of 1989. The three latest adherents, Finland, Austria, and Sweden, joined with opinion at best marginally positive. This then declined sharply, only to start rising again slowly. 2. Explanations There are several different aspects of these charts that require explanation. One is the relatively durable difference between countries in the levels of perceived benefits. Another is the fluctuation that all countries

100

Legitimacy

experience. A third is the long-term trend demonstrated in some countries, and a fourth, and perhaps most important, feature is the widespread decline in the last few years. In addition to the materialistic explanations discussed above, some other reasons have been given for durable country differences and some distinctive trends. One explanation popular in the 1970s was a socialization hypothesis, which asserted that positive evaluations increased with time (Handley 1979), as people learned to approve of the EU in the same way that they acquired party identification. Countries* support would thus differ according to when they joined the EU, and how long they had been members, but all would trend upwards. Quite clearly, this trend has either stopped (see Anderson and Kaltenthaler 1995), or been outweighed by other factors. A more politically focused development of this contrasts the circumstances of accession, with later joining members (all but the original six) necessarily experiencing a more extensive public debate about membership which may have tended to politicize membership and polarize opinion. This could account for why levels of support vary across (groups of) countries: arguably these different groups might not all respond in the same way to policy outputs. The 'direct' materialistic explanations discussed above, such as the proportion of a country's trade that is intra-EU, could well account for enduring national differences as well as upward trends in new members, but the widespread recent decline, as well as patterns of national fluctuations over time, really require some other explanation, and domestic economic trends, real or perceived, provide a possible account. In the analysis that follows we focus on the extent to which economic factors provided important explanations of the national variations and overall trends displayed in Figures 5.1 to 5.3. If economic factors are important, which ones are they: those relating to national economic performance, or those which are more international, such as trade flows and budgetary transfer payments? And if the national economy is important, how much do perceptions of national well-being mediate the link between objective circumstances and policy satisfaction?

101

Policy Performance 3. Analysis

Drawing on previous econometric studies cited above, we estimated a simple model of EU policy satisfaction which included 'direct' and 'indirect' objective economic effects, as well as length of membership (for which we use the log value to allow the effect to diminish with time). Data are annual, with the policy satisfaction variable annualized by combining Autumn and Spring Eurobarometers, to approximate an end-of-year figure. A full description of sources and measures is in the Appendix. Results are in Table 5.1. TABLE 5.1. Regression estimates of EU policy satisfaction Model I GLS estimates b t-value 0.42 5.19 102.23 5.40 0.66 1.77 -1,89 -4.78 O.E4 2.68 12.61 7.07

Net budget contributions Proportion of trade within EU Growth rate Unemployment rate Inflation rate Length of membership (log) UK Ireland Italy Constant -62.14 R2 0.66 Log likelihood -507.30 Chi square 158.81 166 N Source: Eurobarometers Mid OECD 1996.

-4.75

Model II GLS estimates b t-value 0.32 3.49 79.10 3.85 0.49 1.36 -2.15 -5.12 0.75 1.98 11.84 2.09 -27.17 -5.18 18.43 2.09 10.67 1.56 -39.81 -2.17 0.75 -513.99 266.94 165

The results of estimating this first model support the argument that the economy matters. The coefficients for both of the direct economic benefit measures have t-values well in excess of 2.0, indicating a significance well above .05, and the model as a whole explains 66 per cent of the variance. Within-EU trade and greater net gains from the EU budget are connected with policy satisfaction. The measures of national economic well-being also perform well. All have the expected

102

Legitimacy

sign, and unemployment has an effect comparable to the trade and budget variables, with policy satisfaction adversely affected by higher levels of unemployment. The length of membership also matters, particularly in the early years. To get an idea of how much these variables impact on policy satisfaction it is helpful to consider how much change in policy satisfaction is predicted to follow from a change of one standard deviation in each of the independent variables (these are shown in Appendix 1), Trade moves policy satisfaction 10 points, unemployment 8 points, budget share 8 points and inflation 3 points, while growth moves it only 1 point. Altogether, a change of one standard deviation in each of these variables would move policy satisfaction 30 points, which is close to the standard deviation of the policy satisfaction variable. Time has an even greater effect, with one standard deviation shift having a 10-point impact. Model I does not fit all countries equally well. Inspection of predicted and observed policy satisfaction indicates that the model consistently overestimates support in the UK ( by an average of 21 points), and underestimates it in Italy and Ireland (on average, by 16 points). Elsewhere there is no consistent pattern of under- or overestimation, (although Sweden and Finland provide a poor fit on their two observations). There is some difference between the several groups of countries as defined by the timing of EU membership. The original six have an average (absolute) residual of 7 points, the second group, UK, Ireland, and Denmark, one of 16 points, and the third group, Spain, Portugal, and Greece, one of only 3 points. However, there is diversity within each group, notably in group two with the UK at one extreme and Ireland at the other. Similarly amongst the original six, the model underestimates Italian satisfaction by 17 points and overestimates German satisfaction by 8 points. Particularly deviant countries can be dealt with by including dummy variables for them, not so much to improve the fit of the model but to check that these cases are not having a distorting effect. This is done in Model II, also shown in Table 5.1. This does improve the overall fit (explained variance increases to 76 per cent) but it also alters one of the coefficients noticeably, with that for budget dropping from 0.42 to 0.32, This is due largely to the inclusion of Ireland as a dummy. Policy satisfaction is particularly high in Ireland, and Ireland is the biggest net beneficiary from the EU budget. With the exception again of Finland and Sweden, there are no countries whose support is consistently overestimated or underestimated to

Policy Performance

103

any great extent. The range of fit is reasonably even, although it is particularly good in The Netherlands, where the standard deviation of the residuals is only 5.3. Elsewhere it ranges between 9,8 (Greece) and 18.1 (Portugal). The poor fit of the newest entrants is evident but this may or may not be enduring. What may be more significant is the extent to which the model overestimates policy satisfaction in the larger countries of the EU, most notably in France and Germany. We will return to this point later. Many studies of government popularity in individual countries have argued that perceptions of economic well-being are more important than objective indicators. Dalton and Eichenberg (1994) have made this argument with respect to support for Europe, using a variety of measures available in Commission consumer surveys. The Eurobarometer also asks a question in each Autumn survey to tap people's perceptions of improvement in the national economy over the past year. This is a measure only of retrospective, collective (or sociotropic) perceptions. Including this measure along with the objective measures in Model I allows us to see whether economic perceptions have an impact on policy satisfaction. Table 5.2 displays the results. Unfortunately, these cannot be compared directly with those in Table 5.1 since the series available is shorter. Eurobarometer did not ask the perception question after 1994. Hence we have re-estimated Model I on the same time period for comparison. The major questions this part of the analysis should answer is whether or not perceptions are important, and if they are, to what extent do they mediate the effects of the more objective measures of national economic well-being. The first question can be answered by pointing to the coefficient of 4.91 for perceived economic improvement, which has an associated t-value of over 2, There is a significant effect, though hardly a strong one. This variable has a standard deviation of .41, so a change in it of one standard deviation would move policy satisfaction up or down only 2 points. Not surprisingly given this relatively slight effect, there are no clear signs that perceptions mediate relationships between the more objective economic measures and policy satisfaction. Model III and Mode! I perform almost equally well, so it is clear that economic perceptions do not add explanatory power. However, there is no obvious drop in any of the coefficients for the objective measures. Certainly the unemployment coefficient is

104

Legitimacy

equally strong in both models. It is possible that the under- or overestimates with respect to some countries which we found with Model I might have been explained by perceptual differences, but this is not the case. Even with perceptions included, the UK, Ireland, and Italy were again most deviant. Adding them in left only trade and unemployment as significant economic variables, reinforcing the lack of robustness of the budget variable. TABLE 5,2, Regression estimates of EU policy satisfaction Model III GLS estimates b t-value 0.37 4.11 Net budget contributions 102.69 5.39 Proportion of trade within EU -0.003 -0.007 Growth rate -2.37 4.60 Unemployment rate 0.40 Inflation rate 1.00 10.85 3.06 Length of membership (log) 2.10 Perceived economic improvement 4.91 -57.93 -3.02 Constant R2 0.63 -427.04 Log likelihood Chi square 131.89

N

138

Model I GLS estimates b t-value 0.39 4.37 104.80 5.43 0.25 0.63 -2.40 -4.66 0.80 0.33 10.33 2.90 -46.01 -2.63 0.64 -428.57 111.88

138

Source: Eurobarometers and OECD 1996.

Moving back to the European level, away from consideration of individual countries, all of these models perform quite well in the primary task of accounting for the particular pattern followed by policy satisfaction. As was evident in Figures 5.1 to 5.3, the 1980s saw policy satisfaction rise and the 1990s has seen it fall. This can be explained at least in part by the downturn in the economies of many countries after a period of expansion. Figure 5.4 shows the predicted policy satisfaction (averaged across the 15 members) compared with the actual (average) satisfaction. The model captures both the upward and downward movements, while underestimating quite how far satisfaction would rise and how far it would fall.

Policy Performance

105

Source: Eiirobaroineter surveys,

Fio 5.4. Predicted (Model II) and actual policy satisfaction, 1983-96 Finally, we should address the issue of whether policy satisfaction has become less predictable as the EU has extended its authority post-SEA and post-Maastricht, Whilst economic evaluations might provide the primary drive, it could be that as the EU has expanded the basis for evaluation has become wider. Possibly people consider more things when judging whether or not the EU has been beneficial for their country, A simple way to address this is to test the stability over time of the various models estimated above. If the question of policy satisfaction is becoming more broadly based then the model should become less efficient over time. We examined the hypothesis that the residuals from the model are greater in the second half of the period covered than in the first half: i.e. greater post-1989 than up to and including that year. A Chow test on Model I, taking 1989/90 as the cut-off point, produced an F (7,154) statistic of 20.3, a strong indication that there is a lack of stability in the model over time. Further results indicated that the null hypothesis, that residuals were not greater in the second half of the period covered, could be rejected for Models I and II at the .01 level and for Model II at the .10 level. This growing lack of fit is a robust finding

106

Legitimacy

which survives experimentation with various cut-off years and subgroups of countries. All this suggests that the explanations offered above for variations in policy satisfaction are weaker for the 1990s than the 1980s, and that the argument that a wider range of considerations underlie policy evaluations now has some support.

CONCLUSIONS

This chapter set out to describe and explain how citizens of the EU judge the specific policy outputs of the community. It is clear that national differences in satisfaction are strong, and relatively enduring, although there is considerable fluctuation in most countries. In general, satisfaction rose in the 1980s but has fallen in the 1990s, a period which is coincident with a rapid growth in the scope of EU policymaking. Our analysis suggested that notwithstanding the probable change in the salience and scope of EU policy from the middle of this period, much of this growth and decline could be explained by economic factors, most notably variations in levels of unemployment and by growing within-EU trade, together with how long a country had been an EU member. The recent interest taken by the EU in the unemployment issue seems to be well founded, if the aim is to improve its image. Somewhat surprisingly, the effect of national economic circumstances did not appear to depend on a 'feel-good' factor. While this largely economically-driven model accounted quite well for the national levels of policy satisfaction between 1983 and 1996, it was nonetheless evident that it performed less well in the second half of this period than it did in the first, and it failed to account well for the particularly striking decline in support in recent years in France and Germany. It could be that citizens are evaluating the impact of the EU across a wider range of issues, not tapped in this model.

APPENDIX

Perception of benefits: the difference between those who see the EU as benefiting their country and the proportion who see no benefits.

Policy Performance

107

Spring and the preceding Autumn Eurobarometers were combined so as to give an average figure for the end of a year. Economic perceptions: 'Compared to twelve months ago, do you think that the general economic situation in this country is a lot better, a little better, a little worse, a lot worse?* These are coded 1,2,3,4. We have simply taken the country mean value, Trade: the proportion of a country's trade going to other EU member countries. Data from OECD Main Economic Indicators. Growth: the national growth rate. Data from OECD Main Economic Indicators. Unemployment: national unemployment rates. Data from OECD Main Economic Indicators, Inflation: national inflation rates. Data from OECD Main Economic Indicators. Budget: the extent to which the country is a beneficiary rather than a contributor to the EU budget. This is measured by taking the difference between contributions and receipts in ECUs, and dividing this by national GDP. Budget data from the appropriate volumes of the Official Journal; GDP from Office of Official Publications, Eurostat Handhook, Luxembourg. Years of membership: log (years of membership to the EU).

NOTE 1, Some further studies came to hand as this chapter was in proof stage. Of particular importance is Eichenberg (1998) which also contains references to the most recent literature. He examines the benefit question dealt with here, and despite some differences in specification and operationalization also demonstrates the strong impact of unemployment.

This page intentionally left blank

PART II REPRESENTATION

This page intentionally left blank

6

Distinctiveness and Cohesion of Parties

HERMANN SCHMITT AND JACQUES THOMASSEN

INTRODUCTION IN this and the following three chapters, we will put the requirements of the Responsible Party Model to an empirical test. The question we will investigate in the present chapter is whether EU parties are sufficiently distinct and cohesive to fulfil the partisan prerequisites of an effective system of political representation. But before we do so, we first turn to a more formal description of the present structure of the EU party system and, secondly, elaborate the partisan prerequisites of the Responsible Party Model. EUROPEAN UNION PARTIES: A QUANTITE NEGLIGABLE1 The party groups of the European Parliament are the nucleus and main manifestation of the European Union party system (Gaffhey 1996; Hix and Lord 1997; Jacobs, Corbett, and Shackleton 1995; Morgan and Tame 1996). They structure the agenda of the European Parliament; they oversee the EU budget; they together accept or reject the investi-

112

Representation

ture of a new Commission and its president; they are a powerful colegislator in the complex EU system of policy-making particularly— but not only—under the co-decision clause; last but not least, they coordinate and try to determine the political behaviours of 'their' national governments in Council and summit meetings (Ladrech 1996: 292). Unlike some of its national counterparts, the EU party system is characterized by processes of concentration rather than fractionalization. While affiliations between national party delegations and EP political groups have never been very stable and re-alignments have been manifold from the first direct election in 1979 on, the group structure of the European Parliament over the last twenty years has become simpler. The most striking examples of more recent crossovers are the merger of the Spanish and British Conservatives with the European People's Party; the integration of former Euro-communists (the ex-PCI in particular) into the socialist group; the merger of (Northern European, more leftist) Green and (Latin European, more centrist) ecologist parties; and the proto-integration of the European extreme right into a group of formally independent, non-aligned members. The end result of these more recent changes of group affiliation is a significant reduction in the number of EU parties. In the fourth directly-elected European Parliament, the following players are in the game (Table 6,1): the socialist group is the strongest, closely followed by the European People's Party (what we used to call the Christian democratic group). Together, these two groups account for nearly two-thirds of the members of the house which come from 43 national parties. Third comes a conservative group made up of Berlusconi's Forza Italia, the French Neo-Gaullists, and six smaller parties. Next in line is the liberal group integrating representatives from 18 parties and a variety of liberal outlooks, sometimes originating in one and the same country, as in the case of VVD and D66. Next comes the communist group made up of those who remained faithful to communist ideology after the end of the cold war and the breakdown of the USSR; 'non-affiliated' representatives from extreme right-wing parties; the Greens and ecologists; regionalists; and a small group of antiEuropeanists (see Table 6.1). A number of party systems of EU member-countries today are more fragmented than this—particularly if we consider the combined strength of the two main groups, socialists and Christian democrats.

Distinctiveness and Cohesion of Parties

113

What do these processes of concentration and political integration inside the European Parliament mean for the broader question of the functioning of representative democracy in the European Union? In order to assess the effectiveness of the system of political representation in the European Union, we will rely on the criteria stipulated by the Responsible Party Model,

THE REQUIREMENTS OF THE RESPONSIBLE PARTY MODEL

Let us recall a distinction introduced at the beginning of this volume. There we identified a federal and a confederal system of political representation co-existing in the European Union. While the present part of the book aims at a comprehensive assessment of the effectiveness of both systems of political representation, the purpose of this chapter is more limited. It concentrates on the assessment of partisan prerequisites for the functioning of either of the two systems of representation. Irrespective of which system of political representation we look at, these preconditions are twofold: competitive and cohesive parties are needed either way. This at least is the point of view of the Responsible Party Model of political representation (see Thomassen, 1992 and 1994 for a systematic discussion). The model rests on the premise that government policy should be 'in line' with the preferences of (a majority of) the electorate. Being *in line* conventionally is operationalized in terms of a Downsian smallest-distance voting model. According to this model, voters must have a choice between different policy proposals offered by cohesive parties; and voters are assumed to vote for the party whose policy proposals are closest to their own policy preferences. More specifically, Thomassen (1994: 251-52; similarly Esaiasson and Holrnberg, 1996; Pierce, forthcoming) enumerates these requirements as follows: 1. Voters do have a choice, i.e. they can choose between at least two parties with different policy proposals. 2. The internal cohesion, or party discipline, of political parties is sufficient to enable them to implement their policy.

TABLE 6,1. The group structure of the fourth directly-elected European Parliament Group

PSE

PPE

Austria

SPO6

OVP7

L1F1

GAL 1

CVP4 PSC2 CSP 1

VLD3 PRL2 FDF 1

Agalev 1 Ecolo 1

Belgium

UPE

ELDR

CUE

Denmark

SD3 SF1

KF3

V4 RV i

Finland

SDP4

KOK4

KESK.4 SFP 1

Vas 2

France

PS 16

UDF 12

Rad 1

PCF7

Germany

SPD40

CDU 39 CSU8

Greece

Pasok 10

ND9

Ireland Italy

RPR 14 UDF 2

V

ARE

EDN

VU 1

NI

S seats

F6

21 25

VB 2 FN 1

16

JB2 FEF2

16

Vihr 1 ARE 12

AE 1 RPR I

FN 1 1

99

B90/G 12 POLA2

24

KKE2 SAP 2

Lab 1

FG4

FF7

1

PDS 16 PSf-AD 2

PPIS Segni 3 SVP 1 F12

FI25 LN I PSD1 1

LN5 PRI 1

GA2 RC 5

87

Verdi 3 Rete 1

LP 2

AN 1 1

15 87

Luxembourg

LSAP2

CSV 2

Netherlands

PvdA7

CDA9

1

Portugal

PS 10

PSD 9

PP 3

PSOE 21

PP28 UDC 1 EAJ1

CDC 2

IU 9

Sweden

SAP 7

MS

Cp 2 FpL 1

Vp 3

Great Britain

Lab 62 SDLP 1

Cons 18 UUP 1

LD 2

215 19

1S2 24

Spain

I seats S parties

56 8

DP 1

G 1

VVD 6 D66 4

GL 1

6 31

SOP 1 GPV 1

25 64

CDU3

43 18

CEP 1 1

22

MPG 4 SNP 2

33 8

27 11

20 5

20 6

DUP 1

87

32 6

626 101

Legend: PSE = Party of European Socialists {social democratic and socialist); PPE = European People's Party (Christian democratic and conservative); UPE = Union for Europe (conservative); ELDR = European Liberal, Democrat, and Reform Party (liberal and centrist); OUE = European Unitary Left (socialist and commynist); V = Greens in the European Pariiament (green); ARE = European Radical Alliance (centrist); EDN = Europe of the Nation States (anti-communitarian); NI = non-affiliated (extreme right).

116

Representation

3. Voters do have policy preferences. 4. Voters are aware of the differences between the programmes of different political parties, 5. Voters do vote according to their policy preferences, i.e. they choose the party that represents their policy preferences best. Former research has made it abundantly clear that this is not the most realistic view of voter-party relations in a representative democracy (Green and Shapiro, 1994; Thomassen, 1994). Survey research has established that while most voters have clear policy preferences, many of them are unaware of parties' policy positions and thus of programme differences between the parties (for example Thomassen, 1976; Schmitt, 1995). Moreover, voters tend to consider other factors than—or in addition to—policy distances when choosing the party to vote for (Schmitt, 1998). However, as we explained in chapter 1, the requirements of the Responsible Party Model can still be used as a benchmark. Also, we developed an alternative and more realistic conceptualization of the Responsible Party Model, based on weaker assumptions. Inspired by the valence issue logic introduced by Stokes some three decades ago (Stokes 1963, 1992), and the issue salience theory introduced by Budge and Farlie (1983) and reinforced by mass communication studies (for example lyengar and Kinder, 1987), a possible alternative list of requirements reads as follows: 1. Voters can choose between at least two parties that differ with regard to the problems they claim to be able to solve when elected. 2. Parties are sufficiently cohesive to be able to effectuate these claims. 3. Voters are concerned about problems. 4. Voters are aware of differences in issue emphases of the parties. 5. Voters vote their concerns, i.e. choose the party that they believe will best be able to deal with the problems central to them. In this view, it is not so much specific policies that parties compete on. Rather, they fight about the salience of issues. Their goal is to make 'their own* issues the most salient—the ones which a majority of voters perceives are best taken care of by them.' Effective representation is no longer indicated by the congruence of policy preferences between party elites and party voters, but by the congruence of political agendas of representatives and represented.

Distinctiveness and Cohesion of Parties

117

DATA AND MEASUREMENTS

Our data enable us to test both variants of the Responsible Party Model. The surveys among the members of the European and national parliaments include both position and valence issues. Position issues were chosen to represent the most salient EU policy dimensions at the time of the 1994 European Parliament election: currency (national vs. European), borders (rebuild or remove), and unemployment (state intervention vs. market economy),2 However, the effectiveness of the federal system of political representation in the EU must not exclusively3 perhaps not even primarily, be judged upon issues relating to the further development of the Union,3 As a consequence, most of our valence issues have no such explicit link with EU-—or with national or sub-national—policies.4

Union issues

Commonconcern issues

Valence issues

Position issues

unify Europe

common/national currencies open/closed borders (employment programme/ Single Market)

fight unemployment protect the environment fight crime

left-right (employment programme/ Single Market)

FIG 6.1. Types of issues analysed A truly federal system of political representation implies that European political parties compete on a European scale for the votes of a European electorate on the basis of issues that are relevant at the European level—be these Union issues such as the common currency or issues of common concern such as unemployment, environment, or organized crime. The four valence issues analysed in this chapter are the most salient both among members of the European Parliament and members of the national parliaments studied. How can we determine how distinct and cohesive parties are? We generally assume that party elites take the more distinct and distinguishable positions the farther apart they place themselves from one

118

Representation

another. In European multiparty systems, the party differential is not restricted to the distance between the two major parties; the more distinct and distinguishable party positions there are, the richer we assume here the choice options of voters to be. We will pursue this empirically by determining partisan distinctiveness not just by the distance of the two largest parties but additionally establish the overall distinctiveness of the relevant parties in the respective party system. In operational terms there is of course no absolute measure of distinctiveness available, and we need to rely on relative judgements. An obvious and realistic relative standard is the distinctiveness and cohesion of national parties. If parties at the European level are no less distinct and cohesive than national political parties, there is at least no reason to attribute any shortcomings from a more absolute point of view to the particular circumstances of the European Union. In order to be able to make such relative judgements the issue questions introduced above were asked to both members of the European Parliament and members of the national parliaments. Also, they were asked to the voters; this will enable us to assess in the chapters that follow, with the same kind of data, additional conditions of the Responsible Party Model. In more technical terms, and in the case of position issues, two parties are the more distinct, the farther apart they are from one another on the respective policy dimension. We use both the arithmetic mean5 (to determine the distance between two parties) and the eta statistic (to determine the distinctiveness of the overall party system). A similar logic is applied with regard to issue emphasis (though the substance logic is somewhat different). We measure the relative importance of a party's issue emphasis with an index ranging from 0 to 10.* Arithmetic means and etas are computed for this issue emphasis index. The larger the difference between two parties, and the higher the eta for a particular party system, the more distinctive they are with regard to that particular issue. Given the basic decision on the type of data to work with,7 the second partisan precondition of the Responsible Party Model is much easier to determine. Parties are the more 'cohesive', the more alike the positions of elite members of a given party are on dimensions relevant for the vote choice. Therefore, cohesion is measured with the standard deviation of distributions. The smaller the standard deviation, the more cohesive a party is considered to be.8

Distinctiveness and Cohesion of Parties

119

FINDINGS A first visual impression of the distinctiveness of the European party system on several issue dimensions is presented in Figures 6.2 to 6.4, Even without the national frame of reference it is clear that the distinctiveness of the European party system is high on the left-right dimension (Fig 6.2). Distinctiveness of the two extreme parties—excommunists and extreme right—is overwhelming, and the positions of the two major parties are very distinct, too.

Source: European Representation Study. The length of party-elite mean position marks corresponds with their party's electoral strength (measured in per cent of votes obtained in the 1994 European Parliament election). Party abbreviations are explained in Table 6.1.

FlG 6.2. Left-right orientations of EU citizens and mean placements of party elites The most salient valence issue in recent years is unemployment. The relative emphasis of EU parties on the issue is depicted in Figure 6.3. The parties that traditionally represent the working class—ex- communists and socialists/social democrats—emphasize the importance of the issue most; Green and liberal party elites stress it least, Distinctiveness of the two most extreme parties (ex-communists and Greens) is im-

120

Representation

pressive, and distinctiveness between the two major parties is still considerable.

Source: European Representation Study. The length of party-elite mean position marks corresponds with their party's electoral strength (measured in per cent of votes obtained in the 1994 European Parliament election). Party abbreviations are explained in Table 6.1.

FIG 6.3. Unemployment concerns of EU citizens and mean issue emphasis of party elites Our final look here is on the European currency issue which most typically demonstrates the distinctiveness of EU parties with respect to union issues (Figure 6.3). Here again, overall distinctiveness signified by the positioning of the two most extreme parties (the anticommunitarian EDN—Europe of the Nation States and the centrist ARE—European Radical Alliance) is impressive; it actually could hardly be more pronounced. However, the positioning of the two main parties of the system, PSE and PPE, is almost identical on the Europositive pole of the scale. This is of particular importance; within the spatial logic, the size of parties does matter: distinctiveness of large parties is more important, while cohesion is probably harder for them to achieve.

Distinctiveness and Cohesion of Parties

121

Source; European Representation Study, The length of party-elite mean position marks corresponds with their party's electoral strength (measured in per cent of votes obtained in the 1994 European Parliament election). Party abbreviations are explained in Table 6.J.

FiG 6.4. Common currency preferences of EU citizens and mean issue positions of party elites Table 6,2 allows us to compare the distinctiveness of the European party system with national party systems. It displays the differences in issue positions of the two major parties9 for all 12 systems under study—the EU and 11 nations—while Table 6.3 presents the etas for the overall party system analysis. The main findings can be summarized as follows: as regards their positions on the basic left-right dimension, European Union parties are rather more than less distinct than national parties on average are; our visual impression from Figure 6.I is confirmed. Major party comparisons show European Union parties somewhat farther apart than the average major national parties, while the etas are at an equally high level. Nevertheless, in a number of national systems the two major parties are farther apart from one another than PES and EPP are. By contrast, major national parties on average take more distinct positions on the two Union issues 'European vs. national currency' and 'open vs. closed borders' than PES and EPP do. Overall distinctiveness, however, is greater in the EU than in national party systems on average.

TABLE 6.2. Main party distinctiveness (figures are differences of means) parties

B

F

D

Parties in the national parliaments of I NL GR IRL L

Left-right

3.4

2.9

4.3

2.8

4.2

0.5

4.3

4.1

Currency Employment Border

0.2 4.1 0.9

0.2 2.8 1.2

S.5 2.9 2.7

0.1 4.9 1.1

0.4 1.8 0.5

0.1 0.1 0.6

0.2 4.2 2.0

3 position issue average

1.7

1.4

2.4

2.0

0.9

0.3

Unemployment Environment European unification Crime

2.6 1.0 1.4 0.8

2.8 0.8 3.0 0.2

0.8 0.2 0.6 1.2

0.8 1.8 0.8 1.4

0.1 0,2 0.8 0.0

4 valence issue average

1.4

1.7

0.7

1.2

0.3

EP

P

E

S

_Ay^j>e_

1.5

2.4

3.0

4.0

3.0

0.8 1.8 1.3

0.2 3.4 1.5

1.0 0.8 0.1

0.5 3.7 1.0

2.3 na 1.0

0.7 2.5 1.4

2.1

1.3

1.7

0.6

1.7

1.7

1.5

1.2 0.6 0.2 1.0

2.4 0.6 0.8 2.0

na na na na

1.4 0.1 1.0 2.0

0.4 0.4 0.1 0.4

1.2 0.6 0.6 0.4

na na na na

1.3 0.6 0.7 0,9

0.8

1.5

-

1.1

0.3

0.7

-

0.9

Source: European Parliamentarians* Study. Figures are differences of arithmetic means. na= not ascertained; the currency question was not asked in the MNP study of Sweden, and the agenda question was not asked in the MNP studies of Sweden and Luxembourg. The larger the mean difference, the more distinctive are the parties.

TABLE 6.3. Overall distinctiveness (figures are etas) Parties in the national parliaments of

EP parties

B

F

D

GR

IRL

I

L

NL

P

E

S

Av'ge

Left-right

.73

.83

.77

.72

.76

.64

.88

.79

.81

.71

.76

.85

.77

Currency Employment Border

.56 .60 .45

.73 .62 .64

.43 .41 .28

.21 .73 .34

.38 .26 .09

.44 .47 .27

.48 .57 .43

.33 .49 .39

.88 .73 .71

.52 ,18 .27

.36 .58 .30

.68 na .51

.50 .50 .38

3 position issue average

.54

.66

.37

.43

.24

.39

,49

.40

.77

.32

.41

.60

.46

Unemployment Environment European unification Crime

.37 .32 .44 .21

.43 .34 .73 .38

.11 .31 .16 .20

.24 .17 .51 .45

.12 .11 .05 .09

.51 .34 .67 .30

.49 .50 .37 .35

na na na na

.39 .33 .52 .69

.10 .11 .57 .25

.25 .22 .54 .18

na na na na

.29 .27 .46 .31

4 valence issue average

.34

.47

.20

.34

.09

.46

.43

-

.37

.26

.30

-

.33

Source: European Parliamentarians' Study. Figures are bivariate etas with issue scales as dependent variables and party affiliation as independent variable. na= not ascertained; the currency question was not asked in the MNP study of Sweden, and the agenda question was not asked in the MNP studies of Sweden and Luxembourg. The larger the eta, the more distinctive are the parties.

124

Representation

Major parties In both the European Union and the member-states take considerably less distinct positions on Union issues than on issues of common concern. However, this is not true for overall party systems. Eta coefficients demonstrate that there are considerable differences on Union issues within the national party systems, and in the EU party system in particular. Looking at partisan distinctiveness in issue emphasis rather than in position-taking, the result of our comparisons is somewhat simpler; the two major EU parties are more distinct than major national parties in their emphasis on unemployment. This holds as well (but is somewhat less pronounced) for environmental protection and European unification, and they are equally distinct in view of the emphasis on the issue of crime. Major-party comparisons and the overall party system analysis produce the same result. Distinctiveness in issue emphasis does not systematically vary between Union issues and issues of common concern, neither for national nor for EU parties. The Responsible Party Model requires parties to advocate distinct policy proposals and to do so in a cohesive way in order to be able to implement their proposals when elected to office. Smaller and more extreme parties are bound to be more cohesive than larger parties which need to integrate a greater variety of viewpoints. We are therefore safe when we restrict our analyses to the two main party alternatives in each system-—a lack of cohesiveness should show up here first of all. How cohesive then are European Union parties as compared to national parties? And does cohesion vary according to the different types of issues considered? Table 6.4 provides answers to these questions. As regards cohesion in position-taking, the PSE and EPP stand out as being less cohesive on the common currency dimension than any pair of main national party alternatives surveyed here. At the same time, van der Eijk's agreement scores (data not shown) indicate that they are more cohesive than most national party pairs as regards the 'open border* question, about which they are very positive at the same time. Overall, these differences are not really pronounced, and for the two left-right measures, EU parties are exactly as cohesive in their position-taking as average national parties are. To summarize, EU parties are about as cohesive as national parties in their position-taking on common-concern and European Union policy dimensions.

TABLE 6,4, Main party cohesion (figures are mean standard deviations) EP parties

B

F

D

Parties in the national parliaments of ML I GR L IRL

Left-right

1.3

1.1

1.6

1.7

1,8

1.6

1,3

1.4

Currency

2.3 2.2 2.6

1.0 2.2 1.8

2.1 2.4 3.2

1,9

Employment Borders

2.5 2.4

1.8 3.3 3.3

2.1 3.1 3.2

2,0 2.6 2.4

1.3 1.6 2.0

3 position issue average

2.4

1.9

2.6

2,3

2.8

2.8

2.3

Unemployment Environment European unification Crime

3,8 2.6 3.6 1.9

1,8 1.2 3.0 1.5

3.8 1.6 1.8 1.8

2.6 2,0 2.6 1.8

4.6 2.6 3.0 0,8

3.6 1.4 0.4 3.8

4 valence issue average

3.0

1.9

2.2

2,2

3.0

2.4

P

E

S

Av'ge

1.1

1.2

1.3

1.0

1.3

1.0 2,5 2.2

0.7 2.7 1.9

1.8 2.6 2.7

2.1 na 1.9

1.7 2.4 2.4

1,6

1,7

1,8

2.4

2.0

2.2

2.4 1.6 2.0 2.0

na na na na

4,2 3.2 3,6 2,8

2,8 1,8 2.6 1.2

2.6 1.6 2.4 1.2

na na na na

3,0 1.8 2.4 1,8

2,0

-

3.4

2.2

2.0

-

2.3

Source: European Parliamentarians* Study, na- not ascertained; the currency question was not asked in the MNP study of Sweden, and the agenda question was not asked in the MNP studies of Sweden and Luxembourg. The smaller the standard deviations, the more cohesive are the parties.

126

Representation

The situation is different with regard to our issue emphasis measures. European Union parties are considerably less cohesive than average national parties in their emphasis on fighting unemployment, protecting the environment, and farther unifying Europe. They are just as cohesive as national parties on average in view of their emphasis on fighting crime. This difference in cohesiveness is somewhat surprising considering the fact that the distinctiveness of national and European parties on these same valence issues was very similar, and is probably best explained by different political agendas in the various membercountries, SUMMARY What does this mean for the functioning of representative democracy in the European Union? Our findings are briefly summarized. First, the two major European Union parties are somewhat farther apart than major national parties on issues that relate to the basic left-right dimension. Second, major EU parties are somewhat closer together on European Union issues than major national parties are. But this is not the full story; distinctive positions in EU policy questions are taken by small polar parties rather than by the major party alternatives in the EU and the national political systems. Third, major EU parties are again somewhat more distinct than national parties in their emphasis on fighting unemployment, and they are no less distinct than national parties in their emphasis on the other valence issues analysed. Generally, distinctiveness on valence issues is somewhat less pronounced than on position issues. Fourth, EU parties are about as cohesive as national parties in their position-taking on common-concern and European Union policy dimensions. Fifth, they are considerably less cohesive in their issue emphasis. Therefore, the answer to the question whether European party groups meet the requirements of the Responsible Party Model of political representation must be differentiated. If we accept national parties' distinctiveness and cohesion as a baseline for cross-level and cross-issue comparisons, we cannot but conclude that European party groups do meet the stipulated criteria equally well. However, this comparative judgement should oot camouflage our findings on European

Distinctiveness and Cohesion of Parties

127

Union issues. On these issues, the two main European parties take hardly distinct positions. In the next two chapters we will see to what extent the European electorate meets the requirements of the Responsible Party Model, We can expect that voters' perceptions and behaviours will be related to our findings on political parties. The more distinct the parties are, the easier it is for voters to see the differences between them and vote according to their issue orientations.

NOTES 1 In the case of European parties, these efforts in selective issue emphasis originate in their representation of cleavage positions, of particular segments of society, and hence of the specific interests, for example, of a specific social class, a particular religion, and so on. It is known that the social bases of cleavage politics are weakening as a result of ongoing social change (for example Dalton, Flanagan, and Beck 1985), and that the social distinctiveness of part}' electorates is declining correspondingly (for example Franklin, Mackie, and Valen 1992). Issue saliency strategies of political parties, however, are not dependent on the continuing relevance of a particular social basis of the parties. 2 The currency question reads as follows: 'Should [country] keep its [national currency] and make it more independent from the other European currencies or should the aim be a new common European currency? What is your position?' Answers were requested on a ID-point scale ranging from (1) independent national currency to (10) new common European currency. The unemployment question: 'The former President of the European Commission, Jacques Detors, has proposed to raise funds for a massive programme to fight unemployment. Others argue that the completion of the Single European Market alone will be the best remedy for unemployment. What is your position?' The scale endpoints are (1) launch a massive employment programme and (10) concentrate on completion of Single Market. The border question: 'Should the EU continue to remove national borders or should tighter border controls be reintroduced to fight crime effectively?* The scale endpoints are (1) continue to remove national borders and (10) reintroduce close border controls. 3 In consolidated polities, party competition does not normally involve the organization of government. Constitutional issues are only contested if a system of government is at stake—as was the case on the occasion of German reunification, the federalization of Belgium, or in the Northern Ireland conflict (Schmitt 1994, Thomassen and Schmitt 1997). For a contrary view

128

4

5

6

7

8

9

Representation

see for example Bogdaaor, 1989; van der Eijk and Franklin 1991; van der Eijk, Franklin et al, 1996, The valence issue question goes as follows: "Which of the following issues do you consider the most important, which comes second, and which comes third?' Eleven issues were listed, including four nation-specific: nationspecific 1, fighting unemployment, securing stable prices, nation-specific 2, furthering political unification of Western Europe, solving the immigration problem, nation-specific 3, reduction of agricultural surpluses, protection of the environment, nation-specific 4, fighting crime. Nation-specific issues were introduced to fully capture the most important issues for each membercountry; they are fully documented in the notes to Chapter 3, In addition interpolated medians have been calculated. In the following we report mean scores only, as the two measures of central tendency produce identical results. The most important issue of any respondent is assigned the value '3\ the second most important the value '2*, and the third most important the value T. Issues not considered important are coded '0'. As this is meant to be a measure of the relative importance of issues for individual respondents, these assigned scores are divided by the overall number of 'points' each respondent had to assign (i.e. 6). This again is multiplied by 20 to standardize the scale to a range of 0 to 10. A study of party manifestos obviously would have been a viable alternative strategy for the assessment of partisan preconditions of effective systems of political representation. The present study relies on a comparison of issue orientations of representatives in the European and in national parliaments, because comparable indicators of the issue orientations are available for both groups of representatives as well as for the voters. This will enable us to assess in the chapters that follow, with the same kind of data, further conditions of the Responsible Party Model We used an additional algorithm again and calculated van der Eijk's agreement scores, which are probably better suited to the measurement level of our data (van der Eijk 1998) In the following we report standard deviations only, as the substantive conclusions from the two sets of measures of dispersion are identical. The two main parties are in the EU: PES and EPP; in Flanders: SP-CVP; in Wallonia: PS-PSC; in France: PS-RPR; in Germany: SP1>€DU/CSU; in Greece: PASOK-ND; in Ireland: FF-FG; in Italy: PDS-Forza Italia; in Luxembourg: POSL-PSC; in The Netherlands: PvdA-CDA; in Portugal: PS-PSD; in Spain: PSOE-PP; in Sweden: SP-M.

7

The Cognitive Basis of Voting

WOUTER VAN DER BRUG AND CEES VAN DER EUK

INTRODUCTION

IN the previous chapter Schmitt and Thomassen apply the logic of the Responsible Party Model to the European Parliament, When European Parliament elections are viewed from the federal perspective, a number of requirements for political representation can be deduced from this model. One of these requirements is that voters are 'aware of the differences between the programmes of different political parties'. When voters lack accurate perceptions of what the various political parties stand for, they may unintentionally give an electoral mandate to policies they disagree with. Conversely, to the extent that voters are well aware of party positions, elections are more likely to function as a meaningful vehicle for mass-elite communication (Converse 1975; van der Brag 1996), The focus of this chapter is to assess the extent to which this specific requirement for representation—that voters have adequate perceptions of the choices offered to them—was met in the European elections of 1994. Political parties differ from one another in their substantive political stances in a number of ways, each of which will—to the extent that the available data allow us to do so—be examined in this chapter. First of all, parties may be regarded as occupying different positions on dimen-

130

Representation

sions that represent a spectrum of ordered policy positions between two polar extremes. Some parties may, for example, favour the abolition of national borders within the EU, whereas others may take the position that national borders are essential and that their significance within the Union should be enhanced rather than diminished. In the previous chapter, three such 'position issues' were introduced, which reflect conflicting views on the policies the EU and its member-states should pursue. A second type of substantive political differences between parties concerns so-called 'valence issues', where disagreement is not about the goal to be achieved, but about the means by which a generally desired goal should be reached, or the priority it should be assigned in comparison to other desirable objectives (Stokes 1963). According to some scholars, parties compete for votes not so much by occupying different positions, but by selectively emphasizing issues, particularly those on which they think they have a good reputation. To the extent that they are successful in claiming competency in some policy domain, parties would make certain issues theirs, and thus become 'issue-owners' (Budge and Farlie 1976). Successful attempts from parties to 'own' issues would result in well-defined public images of which parties are most competent to handle particular issues. The third type of substantive political differences between parties is ideological differences. Many scholars have argued that voters orientate themselves to party politics largely through general ideological predispositions. Because of cognitive limitations and because of the need to save on information costs, it is rational for voters to use ideological yardsticks to inform themselves about general political differences between parties (Downs 1957; Sartori 1976; Enelow and Hinnich 1984; van der Eijk and Niemoller 1983, 1992; Granberg and Holmberg 1988; van der Eijk and Franklin 1996). Ideological positions on a leftright dimension have been shown to effectively summarize party policies in the member-states of the EU on a wide range of specific issues, both in terms of position-taking and in terms of assigning priorities. Consequently, voters who use ideological positions as information shortcuts to assess parties' stands on specific issues may thus acquire adequate perceptions even if they have no direct information about party positions on these issues (Fuchs and Klingemann 1990; van der Brag 1996). In this chapter we do not intend to explore individual characteristics of respondents related to the quality of their perceptions. Rather we

The Cognitive Basis of Fating

131

want to investigate whether and in which respects the political world as perceived by citizens in general sufficiently resembles that same world as experienced by patty elites to allow citizens to use elections as an effective communication channel for their political preferences— should they desire to do so. For this purpose we assess (variations in) the 'quality' of perceptions at the aggregate level. European elections provide an excellent context to explore such variations because they allow for comparisons between contextual variables and policy dimensions. To study the 'quality' of aggregate perceptions, three aspects will be taken into account: (1) the extent to which voters have any perceptions whatsoever of party policies; (2) the extent to which voters agree about the political stands of various parties on different issues; and (3) the extent to which their perceptions are accurate, i.e. in accordance with an 'objective' measure of parties' positions. Moreover, we will explore the extent to which issue positions of parties on various policy dimensions are related to each other, and whether this relationship is reflected in voters' perceptions.

THE DATA

This chapter uses part of the survey material introduced in the previous chapters, the data of the European Election Study 1994 and the European Candidates Study 1994. In both surveys questions were asked about perceptions of party policies in a number of policy domains. As professional politicians, the candidates may be considered wellinformed about policy positions taken by their party. Their answers may be used therefore as a criterion for judging the quality of the perceptions of the electorates at large. This criterion may be derived from different bits of information contained in the elite study. We could, for example, make use of the candidates' own policy preferences to characterize their parties' stance. We chose not to make use of this possibility for a number of reasons. First of all, not all candidates are equally politically relevant: some were almost certain to be elected, while others were almost equally certain not to be, for instance because of their low rank on a party list. It is less plausible to characterize parties by the personal opinions and preferences of low-ranking candidates than of high-ranking ones, yet this information is not unequivo-

132

Representation

cally available to be takea into account. Moreover, one may wonder how indicative personal opinions of candidates, even high-ranking ones, are for parties' actual political stances. After all, party discipline defines the political position of a party, which in general takes precedence over the personal opinions of its members or representatives. This is not to say that the two are separated by unbridgeable gaps, but rather that the reality of inter-party conflict and co-operation in political arenas produces actual party stances that may be quite different from what all the party's politicians would claim as their personal preference in a politically abstracted and de-contextualized questionnaire. Finally, the candidates of a party for the European Parliament elections may be atypical for their own parties. Particularly with respect to policy dimensions relating to European integration it seems plausible that EP candidates will more enthusiastically subscribe to pro-European positions than other politicians of their own parties. In view of all these considerations, it did not seem appropriate to characterize parties by the personal policy opinions of their candidates. Instead, we used the candidates* perceptions of their own party, which were arrived at somewhat indirectly, namely by asking the candidates where they thought the national members of parliament of their own party stood on a number of policy areas. As this question refers not to particular national MPs, but to the national MPs of the respective parties in general, we may assume that the candidates' responses are directed by their perception of the political positions of the parliamentary faction of their national party, which seems the functional equivalent of the question about the position of (national) parties that was asked in the voter surveys. Perceptions of the three types of part)' policies described in the introduction were measured with different question formats in the EES '94. Perceptions of parties* positions on the left-right dimension and on three bi-polar position issues (currency, unemployment, border control) were measured by asking respondents to indicate on 10-point rating scales the position of a series of parties. Perceptions of the competence of parties in various areas of policy-making were measured in two stages. First, voters were asked to pick the three most important problems from a set of 11. Then they were asked which of the political parties they thought best able to deal with each of the issues they picked from the list of 11.1

The Cognitive Basis of Voting

133

In the present chapter, the units of analysis are the various parties in the member-states. Each of these will be characterized by summary measures of political perceptions (such as percentages of missing data, coefficients of perceptual agreement, and median perceptions). Voters' perceptions of party positions are measured for 98 different parties.2 Four factors complicate comparisons between voters and candidates: (1) for some of the parties for which information from voters is available, the elite sample includes no candidates or an insufficient number of them; (2) some parties competed in the European elections with a joint list, which is a different entity than its separate components about which voter information is available; (3) in some instances different question formats were used in the elite and the voters' studies; and (4) there are differences in the way the surveys were administered, with likely consequences for the incidence of missing data. As a result, for each of the analyses a separate decision had to be made on which parties (cases) to select. A short appendix to this chapter elaborates on the reasons for the different numbers of parties (N) in each of the analyses.

DO VOTERS H A V E PERCEPTIONS OF ISSUES?

As a first step in assessing the quality of political perceptions of European voters, we first examine whether voters have such perceptions at all. An indicator for the presence of such perceptions is the percentage of informative responses to the survey questions about party positions and competence. For different types of parties and for different systems, the average percentage of non-responses—indicating absence of perceptions—is presented in Table 7.1/ Across the European Union as a whole, averaged for all parties for which these questions were asked, nearly 40% of voters are unable to say where they would place that party on each of the three position issues; national borders, national currency, or employment programme versus Single European Market (SEM), This average reaches over 50% in Greece, Portugal, Spain, and Wallonia. Relatively tow percentages of non-response to these questions were found in Denmark, Germany (both West and East), and The Netherlands,4 By comparison, the

TABLE 7.1. Average proportion of respondents who do not locate parties on various issue dimensions (figures are per cent if not otherwise indicated) Valence issues

EU position issues

All parties Parties from: Flanders Wallonia Denmark Germany (W) Greece Italy Spain France Ireland Northern Ireland Luxembourg The Netherlands Portugal Great Britain Germany (E)

N

Currency

Employment

Borders

98

41

40

38

19

6 5 8

48 59 23

48 52 24

46 59 27

6 5 7 6 8

23 52 39 54 36

28 56 41 53 40

7 8

50 38

6 9 5 6

6

Second most

Third most

20

22

22

25 25 11

19 23 15

22 26 18

23 29 19

20 49 40 52 34

13 8 13 26 9

18 16 25 16 15

18 16 27 18 16

19 14 27 18 17

34 35

43 30

25 27

25 29

27 36

27 36

40 37 56 45

50 31 51 46

35 31 56 48

19 17 23 35

21 10 32 12

18 12 33 14

21 16 33 16

25

27

22

17

21

23

22

Left-right

Most important

Governing parties 28 Party families Christian democratic 8 Soda! democratic 16 Green 12 Extreme left 12 Extreme right 8 Other 40 Party size Small 25 Medium-small 23 Medium-large 23 Large 27

34

33

33

14

32 36 46 45 38 41

32 33 43 43 45 40

31 34 42 42 35 39

14 14 24 15 15 21

46 41

47 41 40 33

43 38 39 33

28 18 16

41

35

14

Source: European Election Study 1994. The average non-response percentages were computed on the basis of the number of parties listed in the second column, except for the party competence items, which are computed for each country. The all-countries average for the party competence items are the unweighted averages of the 15 non-response percentages presented in each of the respective columns.

136

Representation

percentage unable to place the same parties in left-right terms is less than half as large (18.7% on average, with a maximum of almost 35%). Scattered information from other sources (such as national election studies) indicates that the percentages unable to indicate party positions on each of these three position issues are comparatively high,5 This probably reflects low salience of these position issues in the public debate of the various political systems. Since more than 90% of the European electorates expressed opinions on these European issues, many of those who have policy preferences on these matters have no perceptions of where parties stand (Schtnitt 1995). To the degree that European elections are about providing MEPs with a mandate on Union issues, this implies that large segments of the European electorates cannot use their ballots to express opinions in a politically relevant fashion. In contrast to the three position issues concerning European integration, lack of salience is not a problem for the set of valence issues reported in the three right-hand columns of Table 7.1. Respondents in the voters' study were asked to state which party would be 'best at' dealing with each of the problems they themselves had chosen as the three roost important ones. The right-hand part of Table 7.1 shows that the non-response to these items is substantially lower than for the position issues. On average, about 20% of the voters in these 15 systems are unable to name a party that is most competent to handle a problem.6 This percentage is only slightly higher than for positions on the leftright dimension. When discussing the absence of perceptions of parties, it may be useful to distinguish various (types of) parties.7 We find that systematic differences exist between non-response for perceptions of parties of different ideological colour; non-response is comparatively lowest for Christian democratic and social democratic parties. Also, we find that the percentage of absent perceptions is lower for government parties than for other ones. These differences reflect to a considerable extent differences in size of parties, which are shown in the bottom panel of Table 7.1. Extremist parties of either the left or the right and Green parties are generally much smaller than mainstream social democratic and Christian democratic parties.8 The bottom panel of Table 7.1 shows that non-response decreases systematically with the size of a party. Conceivably, this reflects the fact that larger parties draw more media attention, with the

The Cognitive Basis of Voting

137

result that voters will be better informed about their positions than about those of smaller ones.

PERCEPTUAL AGREEMENT

The previous section showed that in general, majorities of voters have perceptions of party positions on the policy dimensions we study here; in some cases these are rather slim majorities, in other cases they are more substantial. However, the mere presence of perceptions does not yet say much about their adequacy. One aspect of the adequacy of perceptions that we may study at the aggregate level is the extent to which voters agree about positions of parties, i.e., perceptual agreement. Such perceptual agreement is a necessary, although not sufficient, condition for the electoral process to function as an effective channel of communication, command, and control between citizens and elites. In the absence of any agreement about what parties stand for, individual voters' choices that are guided by policy preferences will together be indistinguishable from random noise. Consensus or agreement about parties* positions allows the collective outcome of such choices to be intelligible. To describe the extent of perceptual agreement regarding the positions of parties on the three position issues and in left-right terms, we use the coefficient of perceptual agreement developed by van der Eijk (1998), which is bound between -1 and +1 (see also van der Eijk and Franklin 1996: 433~4).9To describe agreement among voters about the party they deem most competent to handle each of the respective valence issues, we computed Goodman-Kruskall's Lambda, which is bound between 0 and 1. It describes the degree to which those who name a particular problem as most important name the same party as 'best able* to deal with it.10 Table 7.2 presents these coefficients of perceptual agreement for left—right, the three position issues, and the party competence items. When looking at Table 7,2 we find that perceptual agreement on left-right positions is considerably higher than agreement on issue positions or parties' competence to handle the most important problems. In all member states of the EU, except maybe for a part of Belgium (Flanders), the left-right dimension is obviously a cognitive

TABLE 7,2. Perceptual agreement about party positions and competence European position issues

Valence issues

N

Currency

Employment

Borders

Left-right

All parties

98

.27

.26

.26

.52

Flanders Wai Ionia Denmark Germany (W) Greece Italy Spain France Ireland Northern Ireland Luxembourg The Netherlands Portugal Britain Germany (E)

6 5 8 6 5 7 6 8 7 8 6 9 5 6 6

.26 .19 .39 .31 ,26 .18 .17 .31 .18 .24 .43 .31 .27 .25 .28

.18 .37 .34 .21 .24 .19 .16 .25 .38 34 .17 .31 .26 .33 ,16

,22 .16 .30 .36 .03 .18 .13 .29 ,30 ,31 .35 .34 ,21 .18 .31

.32 .53 .60 .53 .65 .62 .40 .65 .44 .47 .51 ,47 .54 .50 .61

Most important

Second

Third

.14 .16 .17 .21

.16 .12 .09 .20

.20 .23 .05 .15

.04*

.02*

.00*

.05 .03 ,11 ,04 .07 .17 .30

.09

.14

,02*

.04*

.19 .08 .15 .23 .26

,18 .12 .03*

.07*

.04*

.05*

.19 .08

.14 .12

,17 .12

Parties from:

.22 ,25

Government parties Party families Christian democratic Social democratic Green Extreme-left Extreme-right Other Party size Small Medium-small Medium-large Large Candidates

28

.26

.21

.24

.51

8 16 12 12

.32 .22 .18 .29

.25 .28 .30 .44

.29 .30 .28 .20

.49 .51 .54 .67

8 40

.47 .25

.10 .24

.41 .22

.63 .47

25 23

.31 .22

.27 .28

.29 .20

.50 .55

23 27

.26 .29

.28 .23

.27 .26

.54 .51

65

.50

.45

.57

.71

Source: European Election Study 1994, * Lambda does not deviate significantly from 0 (alpha < 0.05).

140

Representation

yardstick that voters comprehend and use in a highly similar way to place parties. On the other hand, perceptual agreement about party positions on the three bi-polar issue dimensions is in all systems much lower, sometimes even extremely low (as in Greece, Spain, Wallonia, and Italy). These low coefficients reflect large degrees of either uncertainty or disagreement about party positions on these issues; in the light of their low saliency in most systems, we expect the first of these to be predominant, In order to judge whether a particular coefficient of agreement is small or large, one needs a standard for comparison. Therefore, for left-right and for the three position issues, we also computed agreement coefficients for the responses provided by the parliamentary candidates." These are presented in the bottom row of Table 7.2, As could be expected, the parliamentary candidates—who can a priori be assumed to be well-informed about the position of their party—agree more about where parties stand than do the electorates at large. Yet, even among these parliamentary candidates there is very little consensus about the positions of their parties on the European issues. For instance, these parliamentarians agree less (on average) about where their own parties stand on national currencies and employment versus a Single European Market, than the entire European electorate does about the left-right positions of parties. When the party candidates are not very certain (or when they possibly even disagree) about the position of their own party, we can hardly expect the electorate to be certain. The right-hand part of Table 7,2 shows that there is just as little agreement about which party is 'best at' handling each of the various valence issues. In some cases the agreement between voters is not even significantly different from chance results, and only in a few cases does the coefficient exceed even a value as low as 0.20, The average value of these coefficients across all systems is slightly below 0.13. These low values may come as a surprise in view of the fact that the issues concerned are the ones designated by voters themselves as the most important from a set of 11 issues, some of which are of prime salience in the various countries of the EU. The explanation we gave earlier for the low perceptual agreement for the three position issues— low salience of EU policies—therefore does not apply here. A quite different phenomenon is likely to be responsible for the low agreement between voters displayed in the right-hand part of Table 7.2. As indi-

The Cognitive Basis of Voting

141

cators of perceptions, these 'party best at' questions are likely to be contaminated by political preferences, as the term 'best' does not refer to an interpersonally agreed-upon state of affairs, but rather to the combination of such a perception with a personal preference. Although this does not necessarily lead voters to mention the party they actually voted for—after all, voters may very well be able to admit that the party they voted for is not the very best for handling each and every political problem—when the question refers to the few problems voters themselves consider to be the most important, it creates a circular logic that will induce voters to ascribe competence to parties they prefer on other grounds. Let us now compare degrees of perceptual agreement between systems and party types. Except for the relatively high degrees of agreement about issue positions of extreme right-wing parties, perceptual agreement varies more between political systems than between types of parties. As the figures in Table 7.2 are computed on the basis of different numbers and kinds of parties for each separate political system, comparisons between systems are not straightforward. Yet, the large differences between, for instance, Flanders and Greece in agreement about left-right positions of parties reflect real differences in the saliency of the left-right dimension in both systems. There are also quite substantial differences between degrees of perceptual agreement in different systems about party positions on European issues. In some systems voters are obviously better informed about party positions than in others. Degrees of perceptual agreement are relatively high in Denmark, France, Germany, The Netherlands, and Luxembourg. In Denmark and France, European integration was a salient issue as referendums were held about the ratification of the Maastricht Treaty. Ireland also conducted a referendum on the Maastricht Treaty, but there the issue was considerably less politicized (either for its own sake or for domestic political reasons; see Franklin, van der Eijk, and Marsh 1995) than in either Denmark or France. Degrees of non-response and of perceptual agreement were both introduced as indicative of the adequacy of perceptions. To the extent that they both reflect the same phenomenon, they will be related. Studying the relationship between these two indicators might help to understand the processes that generate non-perceptions as well as perceptual agreement. We therefore computed correlations between nonresponse and degrees of perceptual agreement (across 98 parties). For

142

Representation

left-right the product moment correlation between non-response and perceptual agreement is -.57, for national borders this correlation is -.52, for national currencies it is -.24, and for employment versus SEM it is -.08. These negative correlations indicate that non-response and absence of agreement both reflect, to some extent, the degree of uncertainty among voters about the position of a party. When respondents are uncertain about the position of a party, they may decide not to answer, or alternatively to provide a rather uncertain response. The uncertainty in responses becomes reflected in larger variations in perceptions, i.e., in lower degrees of perceptual agreement. These correlations are, however, evidently not of equal magnitude, owing to the fact that in addition to uncertainty, other reasons also contribute to either the percentages non-response or to perceptual agreement. An important difference between patterns observed for non-response and for perceptual agreement is that non-response decreases with the size of a party, whereas perceptual agreement is not related to party size. Voters are less likely to have perceptions of small parties than of larger parties, but when they do have perceptions these are just as (un)dear for small parties as for larger ones. These findings confirm those observed for The Netherlands in a longitudinal study of political perceptions (van der Brug 1996: 29). Perceptual agreement depends on a combination of unambiguity of party positions, party size, and contemporary issue salience. Salience in turn is related to politicization and (partly as a consequence thereof) to the degree of media attention to some issue. These two factors may be sufficient for yielding responses on perception questions. Whether media attention leads to common perceptions depends also on the extent to which the information about party policies is consistent (see also Pomper 1972; Page 1978). This interpretation is supported by the relatively low degrees of perceptual agreement about the positions of British parties on European integration, in spite of the fact that such issues have been amongst the most salient political issues in recent years. However, they not only yield conflicts between parties, but particularly also within the thengoverning Conservative Party. Internal division makes it nearly impossible for voters to know where a party stands, because it does not know itself. It seems that the extent to which parties are unambiguous in their position on some issue is more important for yielding clear perceptions than are party conflict and media attention per se.

The Cognitive Basis of Voting

143

One final interesting finding in Table 7.2 is the relatively high degree of perceptual agreement about the position of (small) extreme right-wing parties on European integration. We suspect that the nationalist ideologies and rhetoric of these parties provide a perceptual cue for citizens that reduces their need for direct information about where these parties stand on these specific issues and allows them to deduce this from more generic knowledge. Summing up the results of the analyses so far, we find a mixed verdict regarding the question of whether European voters fulfil one of the requirements of the Responsible Party Model of representative democracy: that they adequately perceive substantive political differences between parties. We find that large groups of respondents are unable to indicate where the political parties stand in terms of specific position issues, whereas much smaller, yet still sizeable groups are unable to locate parties in terms of a left-right continuum. Moreover, those who did respond to these perception questions do not agree very much with one another about the location of parties on the specific issues, although in most systems agreement on parties' ideological (left—right) positions is considerable. The reason for the relatively high nonresponse rates and low degrees of perceptual agreement is probably twofold: first, European issues are not (yet) very salient, as evidenced by the fact that even among candidates for the European Parliament low degrees of perceptual agreement were observed. Second, parties are sometimes badly divided internally, which also makes it difficult to indicate what they stand for. We also examined to what extent the requirement of accurate perceptions is met from the perspective of 'issue-ownership'. The evidence here is mixed. On the one hand we found (in comparison with the three position issues) that comparatively few voters are unable to indicate which party is most competent to handle the problems that voters themselves consider important. On the other hand, however, there is virtually no agreement among voters about which parties are most competent to handle particular problems. Two different, yet not mutually exclusive, explanations may be given for this finding. First of all, although parties may attempt to become owners of issues, they may not be successful, particularly in multiparty systems. Secondly, voters may judge a party best at handling some problem because they agree with this party about the best means to handle it, while not agreeing among themselves about these means and thus not about which party is

144

Representation

best at the issue. In other words, various position issues are hidden behind each of these valence issues, and the party competence question taps not a perception, but a combination of perceptions and opinions. As these 'hidden' position issues were not measured (which would be virtually impossible anyway), we have no way of testing this proposition.

A C C U R A C Y OF M E D I A N PERCEPTIONS

The third aspect of the quality of (aggregate) political perceptions we set out to assess here is how 'accurately' voters perceive the positions of parties. In order to assess this, we will confront medians of voters* perceptions with parties' 'true' positions. This 'true' position of a party is measured by the median of European Parliament candidates* perceptions of the position of their own party's national MPs.'2 For the three position issues as well as for the left-right dimension we will compare medians of perceived party positions with these indicators of their 'true' positions. To compare the medians of voters and candidates we performed four linear regressions, one for each policy dimension (the parties being the units of analysis). The dependent variable of each regression is the median of voter's perceptions of party positions, and the independent variables are parties' 'true' positions on the policy dimensions. These regressions are not interpreted in an inferential way— no assumptions will be made about causal relations—but merely to describe how and how strongly voter perceptions are related to the parties' true positions. The results of the three regressions are presented in Table 73,n They show that voters' perceptions of party positions are moderately to strongly related to parties' 'true' positions. This conclusion flows from the positive slopes and from R2s that range between 0.40 and 0,80, However, voters' perceptions of parties' left—right positions are much more strongly related to where these parties actually stand than are voters' perceptions of party positions on European issues, as becomes clear from a comparison of the R2s.

The Cognitive Basis of Voting

145

TABLE 7,3, Regressions where the dependent variable is the 'median perceived position by citizens' and the independent variable is the parties' 'true' positions. Intercept

Slope

R2

N

Left-right

1.08

.94

.80

74

Open vs, closed borders

3.15

.45

.47

74

Common vs. national currency

2.93

.45

.40

74

Employment programme vs, single market

3.18

.36

.50

74

Source: European Representation Study 1994.

If no systematic differences existed between the perceptions of voters and the parties' actual positions, the regressions would yield an intercept of (close to) 0 and a slope of (close to) 1. Instead, in each case we find a positive intercept and a slope smaller than 1. To interpret how voter perceptions are related to parties' true positions, we may insert the two extreme values of the rating scales (1 and 10) in the regression equation, thus arriving at the predicted positions by the regressions. For left-right these values are 2.02 and 10.48, which show that, although only minor systematic differences exist between 'true' and perceived left-right positions, voters tend to perceive parties at a somewhat higher (more right-wing) position than where they actually stand (as indicated by how candidates perceive them). Such differences are much larger for the perceptions of party positions on the two issues. For instance, the two predicted extreme values for positions on the border dimension are 3.6 and 7.65.14 Thus voters perceive much smaller differences between parties on these European issues than exist in reality, albeit with no systematic bias in either direction of the issue scale. In order to observe the relationships between perceived and 'true' party positions in more detail, Figures 7.1 to 7.4 present the scatterplots of the data of the three regressions in Table 7.3. Figure 7.1 presents this plot for left-right positions. It shows that voters' median perceptions of left-right positions correspond well with the positions as perceived by candidates of each party. Political candidates and voters

146

Representation

apparently agree largely about positions of parties in left-right terms. The largest differences involve some very small parties (the Dutch Centre Democrats and the Danish Progress Party, winning 1% and 2.9% of the votes in their respective countries in the 1994 European election).

FIG 7.1. Left-right: perceptions and 'true' positions Figure 7.2, which presents the scatterplot of perceptions and 'true' party positions on national borders, tells us first of all how misleading overall measures of association, such as R2, can be. Although the R2 of the regression is not very high, most parties are quite adequately perceived. The relatively low R2 (compared to left-right) is partially due to the fact that there is relatively little variation in party positions on this issue. To observe which parties are and are not adequately perceived, two dotted lines are included in the graph to indicate party positions relative to the centre of the scale. The large majority of European parties are situated in the lower left quadrant, which means that these parties wish to continue to remove national borders, and that they

The Cognitive Basis of Voting

147

are perceived by voters as taking such & stance, A small group of (mainly extreme right-wing parties) is also adequately perceived to be against the tendency to remove borders (located in the upper right quadrant).

FIG 7.2. Borders: perceptions and 'true' positions Two Luxembourg parties (PD and PSC) are somewhat against removing national borders, yet are perceived to be in favour of it. For three parties the reverse is clearly true. The Danish Progress Party is the largest outlier. Somehow its ideology combining notions of free enterprise and libertarian anti-statism leads to this gross misperception. The Dutch VVD (conservative liberal party) is incorrectly perceived as opposed to opening the borders. A tentative explanation is that, while the party is in favour of opening the borders for economic reasons, it opposes an 'open border' policy for refugees and immigrants from outside the EU. The German PDS is perceived by the West German electorate as opposed to opening national borders, although they are in fact

148

Representation

strongly in favour of it. The reason for this misperception is probably that the party radically changed its position on this issue shortly before the elections, Figure 7,3 displays the scatterplot of perceived and 'true' part}' positions on the issue of a common currency. Perhaps the most vital observation to be made is that perceptions of party positions on this issue appear to be less adequate than on the issue of national borders. We find voters' perceptions and 'true' positions on opposite sides of the middle of the policy continuum considerably more often than in the case of the borders issue. Despite the fact that parties appear to be more polarized on the issue of national currencies than on the other European issues, voters perceive party positions less accurately.

FIG 7.3. Currency: perceptions and 'true' positions As was the case with the issue of national borders, the cases deviating from the 45° line that would radicate perfect concurrence between perceptions and actual positions are most interesting, as they suggest how

The Cognitive Basis of Voting

149

misunderstandings between the public at large and political parties come into being. The six parties indicated with labels in the lower right quadrant are the German PDS (as perceived by West German voters), the British regionalist parties Playd Cymru and SNP, and the Wallonian and Flemish extreme right parties; these parties are perceived to be against the new European currency while they were in fact (in the eyes of their candidates for the EP) in favour of it. For the PDS the same explanation applies as for the general misperceptions of the West German electorate of its position on national borders. Inadequate cuetaking provides a likely explanation for the other five misperceptions,16 These involve nationalist parties of a separatist (or regionalist) nature. Apparently, the fact that these parties denounce their centralist nationstates and demand more autonomy for their region does not imply that they are against a common European currency. However, most voters seem to have deduced from their nationalist ideologies that these parties would be against transnational policies. The group of parties situated at the bottom left quadrant of Figure 7.3 are mainly centralist-orientated extreme right-wing parties (such as the Dutch CD, the French Front National, and the German Republikaner), and a few others (such as the two Danish splinters, the Progress Party and the SF), who are adequately perceived as opposing a common European currency. At the other end of the scale, two more outliers can be found: the German CDU and CSU are perceived by East German voters to be strongly in favour of a new European currency, while these parties take a somewhat centrist position on the issue. Possibly this is the doing of then-chancellor Kohl for whom European integration seemed to be a leitmotiv. However, many politicians from his party are less enthusiastic about sacrificing the Deutschmark to this goal. Figure 7,4 shows where parties stand on the issue of 'employment versus SEM', and where voters perceive them. Although parties differ substantially on this issue, voters perceive relatively few differences between them. Voters perceive most parties to have a slight preference for launching an employment programme rather than concentrating on a Single European Market. Parties perceived to hold this slight preference range from those who take an extreme position in favour of doing so (such as the environmentalist parties AGALEV and Les Verts from Flanders and France respectively), to parties that radically prefer the other option (such as the Danish Progress Party). Two socialist parties

150

Representation

(the Spanish communists IU-IC and the Walloon PS) are 'wrongly* perceived as being strongly in favour of launching an employment programme. Again, wrong 'cue-taking' is a likely explanation.

FIG 7.4. Employment programme: perceptions and 'true' positions A more general view of the results presented in this section tells us that party positions in left-right terms are perceived quite accurately. Voters agreed to a large extent among themselves about positions of parties on this dimension. Hardly any large outliers were observed between 'true' and perceived positions. Moreover, only small systematic differences were found between these positions. For the position issues things are somewhat different. First of all, much larger systematic differences between 'true' and perceived positions were observed than on the left-right dimension. Secondly, although the positions of the larger political parties are generally perceived accurately, some cases were found in which the electorate generally misperceived positions of (smaller) parties. Such situations could in principle generate large-scale misunderstandings in mass-elite communication.

The Cognitive Basis of Voting

151

RELATIONSHIPS BETWEEN THE POLICY DIMENSIONS

Thomassen argued that policy voting can only yield meaningful representation as conceived by the Responsible Party Model when 'political parties, in the composition of their programs, and voters, when they decide which party to vote for, are constrained by the same unidimensional ideology' (1994; 254). Various studies of electoral behaviour in European elections have demonstrated that ideological left-right positions of voters are good predictors of patty choice in most systems (Franklin, Mackie, and Valen 1992; Oppenhuis 1995; van der Eijk and Franklin 1996). In the previous sections we demonstrated that European voters have in general accurate perceptions of parties' left-right positions. To the extent that European issues are structured by the leftright dimension, voters may still be well represented on these issues when casting their vote on the basis of ideological orientations. Moreover, adequate perceptions may be the result of the fact that an issue is structured by an ideological dimension. To the extent that party positions on the three policy dimensions are related, this will be reflected in relationships between the 'true' positions of parties on these dimensions. Such a structure may or may not exist in the perceptions of voters. On the one hand, voters may, in case they do not know where a party stands on some issue, use left-right positions as a surrogate for these issue positions. In such a situation we would observe relationships between perceived issue positions and the left-right positions even if the 'true' positions are not related. On the other hand, it is possible that party positions on these issues are structured to some extent by a left-right dimension but that the voters are not aware of such a relationship, in which case we would find no correlation between perceived ideological and perceived issue positions. Table 7.4 provides the information to evaluate these hypothetical situations, in the form of squared correlations of median party positions on the four policy dimensions.

152

Representation

TABLE 7.4. Squared product-moment correlations between median positions on four different policy dimensions Vi

V2

V3

Vi, Left-right

voter perceptions candidate perceptions

1.00 1,00

V2.

voter perceptions candidate perceptions

031 0,32

1.00 1.00

voter perceptions candidate perceptions

0.04 0.06

0.71 0,42

1.00 1.00

V4. voter perceptions Employment candidate perceptions

0.18 0.30

0.01 0.04

0.14 0.01

Borders V3.

Currency

V4

1.00 1.00

Source; European Representation Study 1994. The correlations between the medians of candidate perceptions are computed with weighted data. The correlations between the medians of voters' perceptions are computed on a selection of parties for which candidate data are available (N=75, see Appendix). The squared correlations presented in Table 7.4 show that little constraint exists in the four policy dimensions, in reality as well as in the eyes of the voters. Two issues, national borders and employment, are moderately related to the left-right division, whereas the issue of the national currency is not. A stronger relationship exists between party positions on national borders and national currencies, which indicates that these issues are structured by a more deeply-rooted concern with national sovereignty. These two issues are virtually unrelated to the matter of launching an employment programme. Perceived and 'true* relationships between the three policy dimensions are roughly of a similar magnitude. A salient detail is that voters perceive more linkage between party positions on the two European issues than actually exists. The apparent lack of organizing dimensions could well be the reason that, judging by low levels of perceptual agreement and high nonresponse, voters in general do not have very clear perceptions of where parties stand on these three issues.

The Cognitive Basis of Voting

153

CONCLUSION This chapter has dealt with one of the requirements of the Responsible Party Model: that voters adequately perceive the substantive political differences between the parties. It set out to assess to what extent European electorates met this requirement at the time of the 1994 elections for the European Parliament. Three types of such policy differences were distinguished: differences between party positions on three specific dimensions of EU policy; differences in priorities parties wish to give to solving various problems; and ideological differences in terms of left-right. The analyses showed that voters in most European countries are not well aware of the positions parties take on three EU policy dimensions. Consequently, using their European ballots to express preferences concerning European policies is unlikely to be effective. An important factor that contributes to this situation is that parties themselves are not very clear about their positions on these issues. In 1991 van der Eijk and Franklin concluded, on the basis of survey material collected in the 1989 European Elections Study, that, although voters did not choose parties on the basis of policy stands on European integration, the election results could be interpreted as 'an ex post facto endorsement of the policies these parties pursued' on this matter (p. 125). This was the consequence of the fact that, for most parties, their electorates concurred with party stands that in general moderately to strongly favoured further European integration. Although in 1994 the major parties throughout Europe were still generally inclined to favour further integration, as were their voters (see Chapter 3 of this volume), political conflicts on this matter, between as well as within parties, seem to have become more frequent. Internal divisions within parties about European integration on the one hand, and the fact that positiontaking on this matter is not (yet) structured by generic (ideological) dimensions of conflict on the other hand, produce a context in which adequate perceptions of voters are unlikely to develop. The consequence is that, possibly even more than in 1989, any representation that does exist on these policy dimensions is more the effect of coincidence than of the operation of any set of conditions (such as the Responsible Party Model) that generate representation. Analyses of voters' perceptions of parties' competence to handle a number of political problems, which is the second type of policy differences between parties looked into in this chapter, did not indicate

154

Representation

either that the requirement of adequate perceptions was met. In 15 of the systems we studied, very little agreement exists about which parties are most competent to handle the problems voters themselves consider important. Amongst other things, this shows that, to the extent that parties attempt to become owners of issues (Budge and Farlie 1976), they are mostly unsuccessful. It also implies that 'electoral messages' of votes cast on the basis of perceived competence of parties (which is rational at a micro level), are unlikely to be interpreted correctly (at a macro level). The necessary (though not sufficient) condition for electoral messages to be effective—perceptual agreement—is clearly lacking in this respect (van der Brag 1996: 5), This brings us to the third type of substantive political differences between parties: ideological differences. We found that voters perceive parties quite accurately in terms of left-right ideology. As positiontaking of parties on various common concern issues (Chapter 6, this volume) is structured to a large extent by this dimension, left-right positions contain important information for voters about what parties stand for. In order to save on information costs, it is rational for voters to orientate themselves to their respective party systems on the basis of left-right ideology. Because of the good fit between perceived leftright positions at the mass and elite levels, left—right ideology appears to be a suitable vehicle for meaningful mass-elite communication. Chapter 8 will investigate to what extent this ideological dimension is actually used by voters to determine their electoral choice, and thus to what extent its potential to establish the conditions for representation implied in the Responsible Party Model is realized in practice.

The Cognitive Basis of Voting

155

APPENDIX Section 7,2 referred to practical problems in linking the elite and the voters' surveys. In the voters' study (EES '94), respondents were asked to give their perceptions of policies of national parties. In the elite study respondents were asked to express their perception of where the national MPs of their party stood. These elite respondents can be identified by their European party label. In many cases these do not coincide with the national party labels. Table A. 1 gives an overview of how the information in the two surveys was combined. As can be seen in Table A.I, for most of the parties a comparison between voter and elite data is rather straightforward, A few nonstraightforward situations have to be discussed, however. Separate voter studies were carried out in West and East Germany, whereas at the elite level no such distinction was made. Moreover, in the elite study a distinction was made between the CDU and CSU candidates, whereas in the voters' study these two parties were treated as one. These differences resulted in the following decisions: (1) when analysing only voters' perceptions, we analyse East and West Germany separately, and in each of these the CDU/CSU is a single party; (2) when analysing only candidates* perceptions, we make no distinction between East and West Germany, but we treat the CDU and CSU as separate 'cases' (i.e., parties); (3) when linking voter and elite data, we have four cases, as indicated in Table A, 1. A similar problem arose in France and The Netherlands, but in reverse form. The voter survey contains data on perceptions of national parties that participated in the European elections with a joint list. In the elite study the candidates can only be identified by the label of the joint list, not by its constituent parties. Therefore, the perceptions of candidates of the UDF/RPR joint list are included twice and linked to voters' perceptions of both the UDF and the RPR. Similarly, candidates of the SGP/GPV/RPF combination in The "Netherlands are included three times and linked to the SGP, GPV, and RPF. To summarize, the aggregate data set contains perceptions of voters of 98 parties, of which the CDU/CSU combination is included twice (in order to be linked to the separate elite perceptions of CDU and CSU candidates), yielding a total of 100 observations. Whenever information about voters' perceptions is presented that is not linked to elite

156

Representation

TABLE A.I. Overview of how voter and elite data are linked Voters

Elites

Voters

Elites

0K-SDP DK.-RLP DK-Con FP DK-CD DK-SF DK-Chr FP DK-Lib P DK-Progr. EQE-CDU/CSU EOE-CDU/CSU EOE-SPD EGE-FDP EGE-Griinen EGE-Republikaner

DK-SDP DK-RLP DK-Con FP DK-CD DK-SF DK-Chr FP DK-Lib P DK-Progr. GER-CDU GER-CSU GER-SPD GER-FDP GER-Griinen GER-Republikaner GER-PDS GER-CDU GER-CSU GER-SPD GER-FDP GER-Granen GER-Republikaner GER-PDS FLA-VLD FLA-CVPR FLA-SP FLA-AGALEV FLA-VB FLA-VU WAL-SP WAL-PRL/FDP WAL-PSC WAL-ECOLO WAL-FN

FR-PC FR-ES FR-MRG FR-UDF FR-RPR FR-FN FR-Verts FR-GE BRI-Conserv BRI-Labour BRJ-LibDem BRJ-SNP BRI-PC BRI-Greens

missing FR-ES missing FR-UDF/RPR FR-UDF/RPR FR-FN FR-Verts FR-GE BRI-Conser BRI-Labour BRl-LibDem BRI-SNP BRI-PC BRI-Greens missing missing missing missing missing IRE-FF missing missing IRE-Greens 1RE-PD IRE-SF missing

EGE-PDS WGE-CDU/CSU WGE-CDU/CSU WOE-SPD WGE-FDP WGE-Criinen VlGE-Republihmer WGE-PDS FLA-VLD FLA-CVPR FLA-SP FLA-AGALEV FLA-VB FLA-VU WAL-SP WAL-PRL/FDP WAL-PSC WAL-ECOLO WAL-FN

GRE-Pasok GRE-Ncodem GRE-Pol Ann GRE-KKE GRE-Sysnasp 1RE-FF IRE-FG IRE-Labour IRE-Greens 1RE-PD IRE-SF IRE-Dem L. ITA-FI ITA-PDS ITA-PPI '72 ITA-AN ITA-RC ITA-LN 1TA-FDV

ITA-FI ITA-PDS ITA-PPI '72 ITA-AN ITA-RC ITA-LN 1TA-FDV

157

The Cognitive Basis of Voting

TABLE A, 1 cent.

Voters

Elites

Voters

Elites

LUX-PSC/CSV

LUX-PSC/CSV

NIR-DUP

missing

LUX-POSL LUX-PD/DP LUX-GAP-G LUX-ADR LUX-NationB NL-CDA NL-PvdA NL-VVD NL-D66 NL-GroenL NL-CD NL-SGP NL-GPV NL-RPF

LUX-POSL LUX-PD/DP LUX-GAP-G LUX-ADR missing NL-CDA NL-PvdA NL-VVD NL-D66 NL-GroenL

N1R-SF

missing missing missing missing missing missing missing missing missing missing ESP-PSOE ESP-PP ESP-IU-IC ESP-CDC-Ciu ESP-Nat. Co missing

NIR-Ailiance NIR-UUP

NL-CD NL-SGP/GPV/RP, NL-SGP/GPV/RP. NL-SGP/GPV/RP. missing missing

NIR-SDLP NIR-Labour NIR-Conserv N!R-Grcen POR-PSD POR-CDU/PC POR-CDS/PP POR-PS POR-PSN ESP-PSOE ESP-PP ESP-IU-IC ESP-CDC-Ciu ESP-Nat. Co ESP-HB

Note: DK = Denmark, WGE = West Germany, EGE = East Germany, GER = Germany, FLA = Flanders, WAL = Wallonia, FRA = France, BRI = Britain, ORE = Greece, IRE = Ireland, ITA = Italy, LUX = Luxembourg, NL = Netherlands, NIR = Northern Ireland, FOR = Portugal, ESP = Spain.

data, the two extra observations were filtered out, so that we retain the original 98 parties. At the elite level information is included in the aggregate data set from the candidates of 65 parties. By including the seven German parties and the UDF/RPR list twice, and the SGP/ GPV/RPF list three times, ten extra observation units were obtained. Whenever data on elites and voters are compared, those analyses are performed on 75 observation units (or as many of these 75 for which valid data are available). When elite data are analysed separately, the extra ten observations are filtered out, so that 65 observations remain. Whenever lower numbers of observations are reported, this is due to missing data.

158

Representation NOTES

1

2

3

4

5 6

7

These latter questions about the perceived competence of parties are not ideal measures of perceptions, however. Responses to these questions may not only reflect voters' perceptions of parties, but, owing to the evaluative phrasing of the question ('best at') also voters' preferences for parties, unrelated to how they are perceived on the specific issue; Chapter 8 reports analyses that give rise to such suspicion of contamination of perceptions with preferences, A related phenomenon with regard to the placement of parties on ideological or policy dimensions is known as the projection problem; Chapter 11 refers to this. In the analyses we distinguish between the member-states, between the Flemish and Walloon parts of Belgium, Northern Ireland and the rest of Britain, and West Germany and the former GDR. These latter distinctions are made to capture any differences that may exist along these lines and that have been found to exist in related studies (for example, van der Eijk and Franklin 1996), No information about missing data in the elite survey is presented because differences in the way the voter and elite questionnaires were administered makes a valid comparison of missing data in the two data sets impossible. Particularly in terms of non-response percentages do we find differences between countries that are not easily explicable at first sight. It seems likely that in addition to differences in the relative salience of issues in the different member-states of the Union, so-called house effects also play a role in these percentages. The term 'house effects' is used to refer to differences in the outcomes of identical surveys that are generated by different operational procedures used by different survey research agencies. Such effects are known to be much more pronounced in terms of response and non-response than in terms of distributions of valid responses and associations between valid responses. Consequently, when interpreting the results presented in Table 7.1, it seems wise to focus particularly on differences between non-response percentages for different questions within each single country, rather than drawing conclusions on the basis of comparisons between countries. See for example van der Brug (1996: 22). But see Schmitt (1998: 157) for substantial non-response rates from German national election studies. The questionnaires allowed respondents to indicate that they considered none of the parties capable of handling a problem. These responses are not included in the 'missing' categories of Table 7.1. Such comparisons cannot be made for the party competence question because in that case the respondents were not asked to indicate how much competence they ascribed to each part)', but rather which party they thought most competent for particular issues. When a particular party has

The Cognitive Basis of Voting

159

been named by the respondent, we still do not know whether that respondent considered other parties also to be competent, incompetent, or whatever, 8 The category 'other parties* contains both mainstream secular-conservative and liberal parties, as well as sundry other parties that are hard to classify under a single heading. 9 It attains its minimum of-1 when half the sample places a party at the one extreme and the other half places it at the other, i.e., a situation of maximum disagreement. When all respondents place a party in the same category of a rating scale, the measure attains its maximum value (i.e., +1), A uniform distribution yields an agreement coefficient of 0. 10 Goodman-KruskaH's Lambda describes the relationship between two nominal variables, and is bound between 0 and 1, The coefficients have been calculated on the basis of a contingency table that displays on the one hand the problem that respondents mentioned as most important (functioning as the 'independent* variable), and on the other hand the party they mentioned as best able to handle that problem (the 'dependent' variable). If for each problem only one party is mentioned by all voters, Lambda is 1.0. If the proportions mentioning each of the parties as 'best at' are identical for all problems (and hence also identical to the marginal totals), the value of this coefficient is 0. 11 Calculating perceptual agreement of candidates is somewhat problematic when the number of candidates is very small for particular parties. To avoid capitalizing strongly on analyses pertaining to only a few cases, we weighted the various parties for which these coefficients were calculated by the square root of the number of candidates who provided valid responses to the respective items. As the sampling error decreases by the square root of the sample size, this procedure ensures that cases (i.e. parties) are weighted more strongly when they are less affected by sample size. As the problem of the small numbers of candidates for some of the parties occurs in all analyses involving perceptions of candidates, we will weight parties in the same manner in all analyses that make use of information from the elite survey. For the particular analyses in Table 7,2 one additional decision was made: those few eases where only one candidate of a party provided an answer (automatically yielding a meaningless agreement of 1) were deleted, 12 The question of the preconditions for considering the perceptions of the candidates as accurate will not be taken up here, nor will the question of whether such preconditions are empirically met. For a discussion of such preconditions, which hinges particularly on the comparison of perceptions on the one hand with actual party behaviour on the other-—both in terms of what parties say in their manifestos and what parties do in their legislative roles—refer to van der Brag (1996: Chapter 5, 1998).

160

Representation

13 Cases (i.e., parties) have been weighted by the square root of the average number of parliamentary candidates who provided valid responses to the items included in these analyses (see also note 11). 14 For the currency issue these values are 3.38 and 7.47, and for employment versus SEM they are 3,54 and 6.78. 15 A similar instance of an aggregate misperception, probably caused by a misleading cue, was reported by Granberg (1985), In this case most voters misperceived the stand of Ted Kennedy on abortion in 1980. Granberg provides evidence indicating that voters used the fact that Kennedy was a Roman Catholic as a misleading cue for his (liberal) stand on abortion.

8

Policy Preferences and Party Choice

CEES VAN DER EIJK, MARK N, FRANKLIN, AND WOUTER VAN DER BRUG

INTRODUCTION

IN Chapter 6 of this volume Schraitt and Thomassen list five requirements for political representation as conceived by the Responsible Party Model, The two preceding chapters assessed to what extent three of these requirements were met in the 1994 elections for the European Parliament: whether parties offer voters a choice between different programmes, whether they are sufficiently coherent to be able to carry out these programmes (Chapter 6), and whether voters know what parties stand for (Chapter 7). Here we focus on the two remaining conditions for effective representation: whether voters have preferences on the issues at stake, and whether those preferences motivate their electoral choices. Policy voting is a necessary but not sufficient condition for political representation. When voters elect a party or candidate on the basis of their policy preferences while they are misinformed about the stands of the party or candidate in question, policy voting does not produce meaningful representation, but voters may have been motivated by policy nonetheless. In that case voters' preferences will not 'fit' the positions of the parties they voted for (Converse 1975), The opposite

162

Representation

situation is also conceivable. Voters may not care enough about certain issues to let these issues motivate their party choice, even though positions of voters and the parties they voted for correspond well. In such a case the respective issues are of little consequence for our evaluation of the extent to which preferences of voters are represented in the political domain. An example of this can be found in the 1989 European elections (van der Eijk and Franklin 1991 and 1996). Thus, for an overall evaluation of how well voters' preferences are represented in the European Parliament, one must also consider the evidence presented in the previous two chapters, which is a matter that will be discussed in the next chapter. The previous chapters introduced various issue areas about which questions were asked in the surveys. First of all, the surveys include a number of valence issues where the focus of political conflict is not the goal to be reached, but where voters and parties may differ in terms of how important they find each of these generally desired objectives, or in terms of how (not whether) they are to be reached. According to the theory of issue-ownership, voters evaluate parties largely on the basis of their perceptions of the priorities parties will give to issues or the competence of parties in certain policy areas (Budge and Farlie 1976). To the extent that this model correctly predicts electoral decisions, differences in the importance voters give to issues will be reflected in their party choice. Such differences were measured by asking respondents which of these valence issues they found most important, and which were the second and third most important issues.1 In addition to the valence issues, three position issue questions were asked where voters had to indicate both their own positions and those of the political parties in their country on a continuum defined by two poles reflecting different and opposite goals to be attained. These three position issues refer to policy options on matters of currencies, borders, and employment.2 Finally, respondents were asked to indicate their own position, as well as that of the various parties in their country, on the left-right dimension. Although this dimension does not refer to any single specific policy concern, the left-right continuum is known to be strongly related to the most important issues in most western democracies, reason for some to refer to it as a 'super issue' (for example, Inglehart 1984). Various studies have demonstrated that left-right positions reflect to a large extent the positions of voters and parties on several concrete is-

Policy Preferences and Party Choice

163

sues (for example, Fuchs and Klingemann 1990; van der Brag 1996 and 1998), Moreover, it has been argued that the Responsible Party Model requires the behaviour of parties and voters alike to be constrained by the same uni-dimensional ideology (Thomassen 1994: 254), The information available in the European Election Study 1994 permits us to make two contrasts while studying determinants of party choice. We can contrast nation-specific with common issues, and valence with position issues, seeing which of these types of issues are most important in explaining party preferences. In addition we can contrast degrees of policy voting in each of the political systems. First, however, we need to know whether voters in the various European countries have preferences on the various policy dimensions included in this study.

DO VOTERS HOLD POLICY PREFERENCES?

The extent to which voters have preferences for different policies is indicated by the percentage that is or is not able to express such preferences when asked about them in surveys. Table 8.1 presents the percentage of respondents in each country (and over the EU as a whole) unable to express such a preference for the various issues in the European Election Study 1994.J These percentages should be interpreted as lower-bound estimates of the proportion of electorates in each country that have preferences.4 As we can see from Table 8.1, the number of respondents who do not provide a substantive answer to the different questions varies considerably, between countries as well as—for our present discussion more importantly—between the different kinds of questions. Naming which of the 11 valence issues are the three most important ones is the simplest task of all: in no country do we find more than five per cent of the respondents unable to do this; averaged across the various systems it amounts to no more than slightly over two per cent. When we turn to the three position issues and to left-right, we find rather more failures to respond, with as many as 15 to 20 per cent 'don't knows' on one or the other of these policy dimensions in some countries. Averaged over

164

Representation

the political systems of the Union, however, the group without preferences is smaller than ten per cent for each of these scales.5 TABLE 8.1, Percentage of respondents without preferences for various issue areas Valence issues

European position issues Currency Flanders Wallonia Denmark West Germany Greece Italy Spain France Ireland Northern Ireland Luxembourg Netherlands Portugal Great Britain East Germany

EU

Employment

8 15 3 6 10 9 14 4 17 2 6 5 15 2 7

13 15 8 12 14 9 10 7 9 3

8

Borders Left-right

Three most important

3 5 2 2 2 0 3 1

14 6 7 9 9

4 4 9 12 I 19 5 1 3 18 4 5

11 12 4 7 5 11 6 5 15 13 10 6 12 6 10

10

7

9

2

8 18 2

j-»

2 4 2 4 1 1

Source: European Election Study 1994.

Because of differences in the ways polling institutes in each of the countries conduct their interviews which affect the proportions nonresponse (so-called house effects), we must be very cautious with our interpretation of differences between countries. Nevertheless, differences in the proportions of non-response between countries seem to indicate differences in the country-specific significance of these policy concerns. For instance, left-right is not as important in Ireland as in Denmark or Greece. Since large majorities of European voters have opinions on the issues they were questioned about, we may conclude that the condition that voters hold policy preferences is largely met. Although minorities

Policy Preferences and Party Choice

165

of voters seem not to have such preferences, more than 90% (on average) of the electorates in these 15 countries volunteer preferences for these policy concerns. Having assessed this, we now turn to the second condition of the Responsible Party Model to be addressed in this chapter: whether policy preferences motivate party choice.

PARTY CHOICE AND HOW IT IS MEASURED

Any attempt to evaluate the independent contributions made by policy preferences to the party choice of voters is always plagued by two problems: (1) policy positions and party preferences may both be the consequence of deep-seated values generally encapsulated in ideological orientations such as voters* left-right position, or of affinities with social or religious groups, so that we run the risk of attributing effects to issues on the basis of spurious correlations; and (2) policy preferences of voters may be the consequence (rather than the cause) of party choice, where voters are educated into accepting the policies of the party they prefer (persuasion) or where they tend to overestimate the proximity between their own policy preferences and those of political parties that are preferred on other grounds (assimilation). In this chapter we make a number of assumptions about these matters that allow us to arrive at tentative answers to our research questions. Different assumptions might well yield different answers. The assumptions we make, however, are not unique to this research. The same assumptions have been made by many researchers, including ourselves in past work. They are that (1) ideological dispositions and other group and religious norms derive from socializing processes (both in childhood and in later life); (2) that voters' preferences on specific issues (and on the relative importance of those issues) derive in part from ideological predispositions and socializing processes; and (3) that in explaining party preferences at any particular moment in time causal effects run from ideological and issue preference to party preferences and not the other way around. Thus, when looking for the effects of policies we only have to consider their additional effects over and above those of factors such as religion and class.6 In addition to these problems there is another matter which has serious repercussions for the research design, namely that party choice (the

166

Representation

dependent variable) is both a nominal and an ipsative variable, as voters in (almost) all political systems are only allowed to choose one party. From a substantive point of view, this causes two major problems. First of all, any explanatory statements about party choice imply an intra-individual comparison of parties which, in multi-party systems, cannot be observed when only analysing actual party choice. A theory that states, for example, that left-right proximity is an important determinant of party choice implies that voters evaluate all parties on the basis of such value orientations. When attributing actual party choice to this factor one should ideally be able to demonstrate empirically that (ceteris paribus) the second-most proximal party is also the respondent's second choice, that the party most distant from the voter is also the least attractive to vote for. Such observations are impossible when analysing actual voting behaviour, since that relates generally only to a voter's first choice.7 Secondly, when analysing party choice we run into a problem that arises from the (sometimes exceedingly) small numbers of respondents who voted for small parties. If we measure the respondent's party choice in a specific election, we have no information about the electoral attractiveness of other parties. The specific context of our research into European elections causes an additional problem because party choice cannot be compared between countries that each have their own particular party system. These problems—deriving from the ipsative character of the dependent variable, which does not reflect in any kind of detail the various degrees of electoral attractiveness of different parties, and the noncomparable character of party choice—are dealt with by employing as the dependent variable the electoral attractiveness of a political party. This variable, which we will refer to as party preference, is a characteristic that can be measured for all parties, irrespective of their particular traits and irrespective of the political system in which they are located. Respondents to the 1994 European Election Study were asked, 'Please tell me for each of the following how probable it is that you will ever vote for this party?', after which they were presented with the names of practically all parties in their political system.8 These questions yielded a set of variables, one for each party contesting the European elections, measuring on a scale of 1 to 10 the likelihood that the respondent would vote for the party concerned. The open-ended reference to 'ever' in the question is intended to function only as a projective device encouraging respondents to ex-

Policy Preferences and Party Choice

167

press the extent of their current preference for each party. In this way it avoids the restrictions of the more usually employed question about actual voting behaviour, which does not allow respondents to report the extent of their electoral preferences for other parties as well,9 Moreover, the 10-point scale used in our question allows a continuous expression of preferences, rather than a merely dichotomous one. Because respondents were asked to evaluate all parties, even those they had not supported in the European elections, we have a score from each respondent for virtually every party that participated in the elections in that respondent's country. In the analyses to be conducted in this chapter, this measure of party preference is the phenomenon to be explained, the dependent variable. In order to achieve comparability between political systems we need to use an innovative research strategy which will be described in detail below. At this point it is important to note that the use of party preference as a dependent variable avoids the problems referred to earlier. The responses to 'probability to vote* questions provide a much firmer empirical basis for explanatory propositions about determinants of party preference than does actual party choice. A few objections to this approach might be raised. First of all, the 'probability to vote' question employed in the 1994 European election study asked about elections in general, even though it was posed in the context of a European election study. This is however a minor problem. Assessing a general vote intention at the time of a European election rather than at the time of a national election no doubt reduces election-specific national forces that might colour the vote in particular countries at particular times; but this might be an advantage when our objective is to discover what determines vote choice in general terms, since European vote choice is in many cases quite similar to national vote choice.10 A more serious objection could be that we substitute the explanation of party preference (which may be seen as electoral attractiveness of parties) for what was originally to be explained, actual party choice. This objection would, however, only be valid if the party that respondents actually voted for were often different from the one to which they gave the highest score. In fact our whole approach rests on the assumption that respondents do vote for the party to which they give the highest probabitity-to-vote score. Only on this basis can we confidently interpret the second-highest score as a voter's second choice, and so on. Table 8.2 demonstrates that this assumption is well-

168

Representation

founded by showing the extent to which party choice can be deduced from the probability-to-vote variables. When asked which party they would vote for, far more than 90 per cent of the voters name the party they awarded the highest preference score. This suggests that party preference scores are very accurate reflections of actual vote intentions. As a consequence, by analysing the former we can arrive at valid conclusions about the latter,11 TABLE 8.2. Percentage of respondents supporting party with highest vote probability Valid N*

Number of parties

94 95 98 96 98 95 97

409 245 799 735 586 601 599

6 5 8 6 5 7 6

93 93 93 96 96 96 95

597 597 305 322 848

8 8 8 6 9 5 6 4**

Percentage support Flanders Wallonia Denmark West Germany Greece Italy Spain France Ireland Northern Ireland Luxembourg Netherlands Portugal Great Britain East Germany

99

607 923 435

Source: European Election Study 1994. *A1I respondents who indicated that they intended to vote for one of the parties for which the probability-of-fiiture-vote (POV) question was asked, and who indicated their POV for at least one party. **In East Germany the party lists included in the item measuring actual party choice do not correspond in two cases with the parties included in the battery of POV items (B-90/£fe Gritnen, and Linke Liste/tDS). They are therefore not included in this table.

Policy Preferences and Party Choice

169

ASSESSING THE IMPACT OF POLICY PREFERENCES ON PARTY CHOICE

As indicated above, the dependent variable for our analyses is the electoral preference of voters for each of the political parties in their country, which is expressed as a score ranging from 1 (will never vote for that part)') to 10 (will at some time certainly vote for it). These preferences would normally be represented in a data matrix as different variables, one for each party, which would not readily lend themselves to simultaneous analysis. Yet it is not sufficient to analyse these variables one by one. It would not do conceptually, because what we are looking for are determinants of party preference in general rather than a specific model for one part)' or another. It would also not do for another reason: we mentioned earlier that one of the major advantages of using these non-exclusive party preferences is that for each voter they provide variation between parties in terms of their electoral attractiveness, which is needed for valid explanatory statements about the nature of party choice. Analysing these preferences one by one, however, would obscure this (individual-level) inter-party variation, as such a design focuses exclusively on the variation between individuals. An adequate analysis of these scores requires a research design in which inter-party and inter-individual variance is accounted for simultaneously. This can be realized by rearranging the original data into a socalled stacked form (Brown and Halaby 1982; Stimson 1985): viewing each preference score given by a voter as a separate case to be explained. In this way, each respondent is represented by a number of cases in the stacked data set, as many as the number of parties for which he or she gave a preference score (see van der Eijk, Franklin, et al. 1996: 345). The stacked data set can be analysed in the same way as any normal rectangular data matrix. The dependent variable is the preference score; appropriate identifiers allow characteristics of individual respondents and of parties to be added as explanatory or control variables. The independent variables have to be defined in an appropriate manner before they can be included in the analysis (see below), but once this has been done, the stacked data matrix allows us to examine the dependent variable using familiar and straightforward methods of analysis, such as regression.

170

Representation

This strategy can also be extended to encompass different political systems: we just combine the stacked data sets from various countries into one single data matrix, in which one single variable represents party preference for all respondents over all parties in all countries.12 But before any of this can be done, we first have to decide how to represent the independent variables in the regressions. THE INDEPENDENT V A R I A B L E S The policy variables available to us for inclusion in our analysis were already described in the introduction to this chapter. In order to employ these questions in an analysis that assesses their contribution to party choice, we derived for each of the three position issues a variable measuring the distance between the position preferred by each respondent and the position that respondent ascribed to each party. Thus for each combination of respondent and party there is a measure of distance. These can be added to the stacked data set described above, and the contribution made by each of these position issues to party preference can be assessed by calculating the impact of "distance from party'—a rather general concept. The overall importance of each of the three position issues can thus be assessed by means of an effect that has been calculated over all the parties in a country or, indeed, over all the patties in all the countries. For the valence issues, a rather different procedure had to be adopted in order to be able to determine their contribution to general party preferences. From each of them we constructed two independent variables, representing two different views of the way these issues work. First of all we considered the logic of 'issue-ownership*. If both voters and parties are differentiated by the extent to which they give high priority to one issue or another, and voters were to react to this in their behaviour, then we would expect that voters would prefer those parties that share their own priorities among the valence issues. Even though no explicit information is available about voters' perceptions of parties' priorities, to the extent that voters are aware of which parties 'own' such issues, we would find that giving high priority to each specific valence issue would contribute to high preference for those parties. Therefore, we constructed 11 variables, one for each of the valence issues, to capture any such effects; the method of construction was the

Policy Preferences and Party Choice

111

same as for several other variables and is described below.13 The second way in which we constructed variables relating to the battery of valence issues took advantage of the presence of the answers to the question, 'which party is best at' each of the three most important issues. Recall that the stacked data set has a case for each party for each respondent. For each of these records (i.e. party by respondent combinations) we injected a variable that indicated whether that particular party was named or not as best able to handle each of the three problems deemed 'most important' by that respondent. In order to assess the relative contribution of these issues to party choice, we also had to take account of other influences. In particular, variables were considered that are assumed to partly determine issue preferences of voters: their religious and group identities and their leftright position.M In the case of left-right position, the same procedure was followed as for the position issues. Respondents were asked about their own position on this scale and about the position they ascribed to each of the parties in their political systems. By subtraction we derived a distance of each respondent to each party in his or her political system that could be included in the stacked data set. Other independent variables were handled somewhat differently. For class, religious denomination, and the valence issues introduced above, no measure of distance was obtainable. Instead, for each of the parties separately (and hence for each of the countries separately) an artificial measure of 'closeness' was derived from the predictions of party preference for each of these variables in turn.15 The same procedure was adopted for social, class, using subjective social class and income as predictors, and for the valence issues, using whether or not the issue in question had been named as one of the three most important.16 An independent variable that has been added to our analyses without either referring to an: issue in the proper sense of the word, or being a necessary control variable is 'EU approval'. This variable is attitudinal in character and captures respondents' positive feelings (or the lack thereof) toward the process of European integration.17 The reason for including it in our analyses is twofold. On the one hand, the elections we investigate are European ones, providing room for the notion that attitudes and orientations toward the integration process itself may conceivably influence voters* electoral preferences for political parties measured in that context. Secondly, it allows us to replicate a similar

172

Representation

analysis of party preferences in the context of the 1989 European elections that has been reported elsewhere (van der Eijk, Franklin, and Gppenhuis 1996; Gppenfauis 1995). As indicated above, certain other control variables can be added to the stacked data without any special transformation being applied. In the analyses conducted in this chapter, only a single control variable was included: party size. Although this variable plays no part in the causal processes underlying issue voting, other research has found it to be a strong predictor of party preferences, representing the strategic considerations employed by voters who, other things being equal, prefer to support a party that has a good chance of achieving its policy goals (van der Eijk, Franklin, et al. 1996, Chapter 20; Tillie 1995). Consequently, in order to arrive at a model specified as properly as our data allow, we included it in our analyses. MISSING DATA AND WEIGHTING In the analyses reported below a number of records in the stacked data set were excluded because of missing values for the dependent variable. As a consequence, not all respondents have equal weight in each analysis. If, for example, the 'probability to vote' question was asked for six parties, and a particular respondent failed to answer this question for two of these parties, (s)he is represented in the stacked data set by four records, whereas another respondent who gave valid answers for all six parties is represented by six records. Other missing data were handled by means of pairwise deletion, except for the y-hat variables which were estimated from equations in which missing values for the predictor variables were replaced by their mean values. When analysing the data for the various political systems separately we weighted them in such a way as to reflect the results of the European elections themselves (cf. van der Eijk, Franklin, et al. 1996, Appendix B). In the Euro-wide analysis, these weights were multiplied by country-specific constants so as to ensure, first, that each political system contributed equally to the analyses by being represented by the same number of effective cases, and, second, that the effective number of cases in the entire (stacked) data set was reduced to 15,000—approximately the real number of interviews obtained.

Policy Preferences and Party Choice

173

FINDINGS We start by investigating the extent of variance in party preference (country-by-country and over the EU as a whole) that can be explained by the different (groups of) independent variables, taken one at a time. Because of their large number, Table 8.3 presents the joint effect of the set of 11 valence issues combined. As it turns out, voters' policy preferences in terms of the priorities they assign to various political problems have hardly any explanatory power over their party preferences. Taken all together, the highest R2 for the entire set is 0.08 in Flanders, and averaged across the 15 political systems only four per cent of variance is explained. At this low level of explanatory power it does not make much sense to consider the separate effects of the various issues, as these are minimal. This finding is somewhat surprising, because a very similar analysis of the data from the 1989 European election study yielded 12 per cent explained variance when using a set of 12 very similar valence issues (van der Eijk, Franklin, et al. 1996: 350).'8 It is not very likely that these differences are the consequence of real changes in the calculus of voters. Much more likely is that the difference arises from a change in question format, such that the effects of the 11 valence issues presented in Table 8.3 underestimate the effect of voters' issue concerns on their party choice by some eight per cent.19 On average (see the bottom row of Table 8.3) the strongest explanation of party choice is provided by the questions about party competence—the parties that, according to respondents, are 'best at' the three valence issues they had indicated as the most important ones. However, this relationship is not entirely convincing as we fear that the causal relation behind it runs at least partly the other way around: from party preference to party 'best at*.20 Ideological distance is more important than any one type of issue. The set of valence issues and each of the three position issues are all of roughly equal importance, each explaining approximately four per cent of the variance in party preferences. The strength of the effects of (distance to party in terms of) leftright, class, religion, and EU approval is almost identical to what a similar analysis of the data of the 1989 European election study yielded (van der Eijk, Franklin, et al. 1996: 350). Moreover, the relatively weak effects of class and religion are entirely in line with analyses of

TABLE 8.3. Explained variance In party preference by various independent variables Leftright distance Flanders Wallonia Denmark West Germany East Germany Greece Italy Spain France Ireland Northern Ireland Luxembourg The Netherlands Portugal Great Britain

EU

.09 .23 .21 .26 .19 .33 .28 .20 .25 .15 .19 .19 .20 .26 .14 .21

EmCurrency ployment distance distance

.03 .0? .04 .02 .01 .08 .11 .06 .05 .02 .04 .09 .04 .04 .03 .04

.02 .01 .05 .02 .03 .09 .06 .07 .04 .00 .02 .12 .02 .02 .05 .04

Borders distance

All valence issues

Class*

.04 .02 .03 .08 .03 .04 .04 ,05 .05 .03 ,13 .05 .03 .03 .02 .04

.08 .05 .06 .05 .02 .03 .05 .03 .07 .02 .07 .04 .04 .03 .06 .04

.02 .08 .05 .01 .01 .02 .01 .02 .02 .02 .04 .01 .02 .02 .04 ,03

Religion

EU approval

Party competence

Party size

.05 .03 .02 .02 .03 .02 .04 .04 .04 .02 .33 .03 .08 .03 .02 .06

.02 .01 .07 .01 .00 .03 .00 .01 .02 .01 .02 .02 ,00 .01 ,01 .02

.28 ,25 .30 .35 .17 .35 .30 .44 .18 .20 .19 .28 .24 .3! .41 .27

.03 .07 .10 .18 .08 .07 .04 .16 .02 .06 .02 .22 .21 .18 .19 .10

Source: European Election Study 1994. The effect of social class on party choice is estimated with two predictors (y-hats) of subjective social class and family income.

Policy Preferences and Party Choice

175

the socio-straetural bases of party choice as reported in Franklin, Mackie, and Valen (1992). Our findings here can thus be regarded as an independent confirmation of established insights, a fact that greatly increases our confidence in the results reported here. Incidentally, the sizeable differences between their respective R2s in Table 8.3 vindicate the decision to distinguish between Flanders and Wallonia, between the western and the eastern parts of Germany, and between Northern Ireland and Britain. The results of Table 8.3 do not give a decisive perspective on the strength of issues in the explanation of party preferences because issues are not independent of other predictors, especially left-right. Table 8,4 takes a more causal view of the voter's decision-making process by regarding issue preferences as being partly determined by variables earlier in the causal sequence, namely religious and class identity, left-right position, and party size.21 The table lists the explained variance for seven different models, starting with the simplest, which only incorporates religion and class, and adding other independent variables to this model in a series of steps. Models 2, 3, and 4 add first of all left-right ideology, then the three position issues, and next the set of 11 valence issue variables. This entire set of variables relates clearly to voters' policy considerations, and has a considerable impact on variance explained, raising it (EU-wide) from 0.08 in Model 1 to 0.28 in Model 4. By far the largest part of this increase is effectuated by the ideological (left-right) distance between voters and parties, while the more specific issues combined add about three per cent to variance explained." As was the case in 1989 (and as was foreshadowed by the results of Table 8.3), adding EU approval (Model 5) does not yield any appreciable increase in R2. The next addition (Model 6) concerns the variable that reflects voters' strategic considerations, party size, which EU-wide increases the share of explained variance considerably, from 0.28 to 0.36. Model 6 is the one that most closely resembles a similar model presented elsewhere for explaining party preferences in the 1989 European elections (van der Eijk, Franklin, and Oppenhuis 1996: 357). Yet its power is considerably less than in 1989, when the R2 reached almost 0.43. The cause for this difference is mainly the difference in independent variables that could be used. In 1989 a variable relating to respondents' approval of the government in power (reflecting either positively or negatively on the preferences of the parties holding government positions) added significantly to the explanation, while such

176

Representation

information was not available in the 1994 data. Furthermore, the change in question format for the set of valence issues (see above) robbed these of most of their explanatory power. Taking these considerations into account, we find our results for 1994 quite similar to those of 1989. Finally, to Model 6 we added party competence variables. This addition, however, gives rise to an important concern to which we now turn. TABLE 8,4. R2s of seven multivariate models to predict party choice Model Model Model Model Model Mode! Model 4 I 7 6 2 3 5 Flanders WaSlonia Denmark West Germany East Germany Greece Italy Spain France Ireland Northern Ireland Luxembourg The Netherlands Portugal Great Britain

.0? .11 .07

.14 .29 .24

.17 .30 .26

.22 .32 .28

.23 .33 .28

.25 .35 .39

.39 .44 .48

.04 .04 .04

.27 .21 .34

.30 .22 .36

.31 .23 .37

.31 .23 .37

.42 .28 .40

.52 .33 .53

.05

.30

.33

.34

.34

.37

.48

.06 .06

.22 .27

.26 .30

.28 .31

.28 ,32

.42 .34

.57 .41

.04 .34

.17 .40

.19 .41

.20 .43

.20 .44

.24 .46

.34 .50

.05 .10 .05 .06

.22 .26 .29 .18

.27 .27 .29 .19

.29 .28 .31 .22

.29 .28 .31 .22

.37 .43 .42 .43

.47 .49 .51 .53

EU

.08

.25

.27

.28

.28

.36

.45

Source; European Election Study 1994. The regressions contain the following independent variables: Model 1: religion and class; Model 2: Mode! ! plus left-right distance; Model 3: Model 2 plus three policy distances; Model 4: Model 3 plus rankings of I I valence Issues; Model 5: Model 4 plus EC approval; Model 6: Model 5 plus party size; Model 7: Model 6 plus party competence.

Earlier in this chapter we articulated our concern that these 'best at' or party competence variables might be circular, i.e., that they might be not the cause but the consequence of party preference. To assess the direction of the causal relationship between party preference (the dependent variable in the models displayed in Table 8.4) and these party competence variables, we used LISREL analyses to test two rivalling causal models, one founded on the premise that party competence is

Policy Preferences and Party Choice

177

one of the causes of party preferences, the other on the premise that it is the consequence. Neither one of these models fits the pattern of empirical relations between the variables satisfactorily, which indicates that a reciprocal relationship exists between them: to some extent the competence variable is the cause of preference, but to some extent it is also the result of those preferences. Therefore we added this variable to the regression models listed in Table 8.4 only after the more unambiguous causes were included. The addition of this variable increases (EU-wide) the variance explained by nine per cent to 0.45. Depending upon how one feels about the status of the different independent variables, one may derive from Table 8.4 different conclusions regarding the impact of issues. If one considers only the three position issues and the set of 11 valence issues as 'proper' issue variables, the contribution of issues to the explanation of party preferences amounts to a meagre three per cent. Considering left—right as sufficiently policy-related to be of similar status as specific issue concerns in the Responsible Party Model, the picture changes to one where issues contribute a full 20 per cent to variance explained. If one is willing to over-interpret the party competence variables, the result is that in Europe no less than 37 per cent of variance in party preferences can be explained by issues. All these figures are to some extent underestimates of the results we would get if more and different issues had been included in the surveys. Yet all available evidence from other surveys—such as national election studies in various countries—points to a decreasing contribution of additional issue variables to a total explanation as the number of issue variables increases. The reason for this is simply that in most (stable) political systems, voters' and parties' positions on issues are structured in only a few different dimensions, the main effects of which are rapidly tapped by including variables relating to sociostructural and ideological identifications. Having evaluated different models in terms of their explanatory power for each of the political systems of the EU, as well as for the combined electorates from these 15 systems, we are still left with two questions. First, how do the effects of the independent variables compare with one another when we do not focus on stagewise increases in R2, but rather on regression effects when variables are viewed in conjunction with one another? Second, to what extent does it make sense to combine the data of all the different political systems (as we did in

178

Representation

the EU-wide rows of Table 8.4) when the country-specific analyses (the rows for each of the political systems) seem to indicate that the importance of particular explanatory factors differs considerably between them? Table 8,5 presents the standardized regression coefficients for Models 6 and 7 when they are applied EU-wide, i.e. on the data set in which the surveys from all the member-states of the EU are combined. Model 6 shows clearly that only two factors have a strong impact on party preferences: the left—right distance between voters and parties (the larger the distance, the smaller the preference, hence the negative sign of the coefficient), and the size of a party, which represents that voters will, ceteris paribus, prefer large parties over small ones, presumably because they offer a better prospect for getting ideological or policy stands realized as actual government policies. All specific issue variables are weak, as is the effect of EU approval. Variables related to class are rather weak, as they have consistently been found to be in studies conducted since the late 1970s. Religious factors are of intermediate strength. Yet to turn to the second question formulated above, how should we look upon these EU-wide effects when we see in the country-specific analyses that the explanatory power of, for example, socio-structural factors varies between a high of 0.34 in Northern Ireland and a low of 0.04 in western and eastern Germany, Greece, and Ireland (see Table 8.4, Model 1)? Is it not necessary, one might wonder, to specify interaction effects that would tap differences between countries in the effects of the independent variables? Earlier research showed that indeed the regression effect of some variables is so different in the various countries that some interaction terms are warranted (van der Eijk, Franklin, and Oppenhuis 1996). The most important of these was an interaction between left-right distance on the one hand, and the perceptual agreement in a country about the position of the various parties on the left—right continuum on the other. Including this interaction term in Model 6 yields the results displayed in the middle column of Table 8.5. We find for 1994 the same value of this coefficient as in 1989: -0.04, which indicates that in countries where voters agree more with one another about parties* left-right positions, the effect of the leftright distance between a voter and a party has a stronger impact on voters' preferences for parties. A comprehensive scan was performed to assess whether or not any new and additional interaction terms involving the independent variables of Model 6 or 7 are necessary. The

Policy Preferences and Party Choice

179

independent variables of Model 6 or 7 are necessary. The results of this scan were overwhelmingly negative; in other words, there is no compelling need to complicate the models by including country-specific interaction effects, TABLE 8,5. Regression coefficients (betas) of the three most encompassing models of party choice (EU-wide) Model 6 + interaction

Model 7 + interaction

.16 .09 .03

.16 .09 .03

.13 .07 .02

-.33 -.06 -.06 -.04

-.34 -.06 -.06 -.04

-.28 -.05 -.04 -.04

.03 .03 .04 .04 .02 .03 .03 .03 .06 .03 .03 .04 .27

.03 .03 .03 .04 .02 .03 .03 .03 .06 .03 .03 .04 .27

.03 .02 .03 .02 .02 .02 .02 .02 .03 .02 .02 .03 .16 .17 .12 .11

-.04

-.04

.36

.45

Model 6 Religion Subjective social class Family income Left-right - ideological distance Currency - policy distance Employment - policy distance Borders - policy distance National issue J - ranking Unemployment - ranking Stable prices - ranking National issue 2 — ranking European unification - ranking Immigration - ranking National issue 3 - ranking Agricultural surpluses - ranking Environment - ranking National issue 4 - ranking Crime - ranking EU approval Party size Party competence (1st probl.) Party competence (2nd probl) Party competence (3rd probl.) Interaction left-right distanceperceptual agreement

R2

.36

Source: European Election Study 1994.

The differences in the effects of policy preferences on party choice in Tables 8.3 and 8.4 should therefore not be interpreted as differences between these countries in the process that generates preference (or

180

Representation

lack thereof) for political parties. They merely reflect differences in distributions of the respective policy preferences, not differences in how those preferences impinge on party choice. Having discovered this, we may indeed conclude that we are justified in combining the surveys from the different member-states of the EU into a single analysis, because the mechanisms by which policy preferences affect party choice is basically the same everywhere in these countries.

DISCUSSION AND CONCLUSIONS

How do these findings help us to assess the Responsible Party Model? To the extent that the model focuses on specific programmatic stances of political parties, deriving for instance from their manifestos or their observable political behaviour, our findings are sobering. We find hardly any motivational basis, in terms of specific issues and policies, for voters' party preferences. One might wonder whether this result is caused either by a paucity of relevant issue and policy questions, or by the way in which these questions were phrased in our surveys. Although more and different questions would probably make some difference to the overall picture of our findings, we have no reason to think that they would really alter it. The experience of analysing other surveys (the European Election Study 1989, as well as national election studies in various countries) leads to the conclusion that even the most extensive set of questions on the most salient issues contributes relatively little to the explanation of party choice once other, causally precedent, factors such as socio-structural factors and ideology have been taken into account. On the other hand, if we are willing to look at the Responsible Party Model in a somewhat more relaxed way—along the lines suggested by Thomassen (1994)—by focusing not only on specific policy and issue concerns, but more generally on substantive political concerns and orientations of voters relating to how parties position themselves in politics, then the Responsible Party Model becomes more persuasive. The strength of ideological motivations is considerable and, in all countries of the European Union, it is amongst the most significant motivational factors in party choice. Of course, parties decide less directly about their ideological position than about their stance on specific issues, as

Policy Preferences and Party Choice

181

their left-right position is the product of inter-party conflict on a host of specific issues. In spite of this, looking at left-right as akin to an issue, we find a strong basis for considering European elections from the perspective of the Responsible Party Model. This conclusion is reinforced when we consider the accumulated evidence of Chapters 6, 7, and 8. Chapter 6 showed that party groups of different ideological colours compete in the European Parliament along a left-right dimension. Thus at the level of the European Parliament a party system develops, which offers voters a programmatic choice that is clearly structured along the left-right dimension. Voters have rather sophisticated perceptions of party positions in left-right terms (Chapter 7). Here we find that left-right positions are not only relatively powerful determinants of party choice, but that the effect of left-right is basically the same in all countries. With respect to policy preferences of voters in generic ideological terms, elections for the European Parliament thus seem to satisfy all requirements of the Responsible Party Model reasonably well. This is what we will discuss in greater detail in the next chapter. NOTES 1 2 3

4

The wording of these questions was presented in Chapter 3 of this volume. The wording of these questions was presented in Chapter 6 of this volume. AH analyses reported in this chapter were performed on the post-election survey of the European Election Study 1994. For each country, the data have been weighted in such a manner as to reflect the actual outcome of these elections in that country. In the individual-level analyses performed in this chapter on the combined data from all countries, these weights were further adapted so as to maintain on the one hand their capacity' to reflect the actual election result in each country, and on the other hand to give each country as a whole an equal weight in the analysis. The weighting procedures used are identical to those used in van der Eijk, Franklin, et al. (1996, Appendix B). Moreover, as in the analyses reported in that volume, we distinguish more political systems than member-states of the Union, by differentiating between the two main Belgian regions of Flanders and Wallonia, between Northern Ireland and Great Britain, and by distinguishing between the two former German states. These percentages are lower-bound estimates of the percentages in the populations for two reasons. First, in our surveys we have no information about citizens who did not participate in the study, a group that consists of a larger proportion of people with little political interest than those who did

182

5

6

7

8

9

Representation co-operate with the survey. Secondly, there is an incentive in any interview to express a preference even if one does not really have an opinion, Elsewhere, non-response was found to be lower for specific issues than for left-right (van der Brug 1996). Such differences are also known from the European Elections Studies (Schmitt 1995). The reason that no such differences materialize in Table 8.1 is due to the fact that only one of two non-response categories is displayed—answering categories of the leftright scales provide a response 'refused* in addition to the usual 'don't know'. Since refusal does not measure the absence of preference but rather an unwillingness to reveal it, we decided to disregard it in the present context. If we were concerned with the evolution of issue preferences or ideological positions over time, the causal connection might well be reversed to at least some degree, with party preferences influencing policy preferences, and the latter influencing values and ideological predispositions. At any one election, such as that of 1994, however, the much higher short- and medium-term stability of ideological and value predispositions in comparison to specific policy and issue preferences ensures their causal priority (Campbell et al. I960; Fiorina 1988; Granberg and Holmberg 1988). This assumption may also be tested by using more explicit causal models, as we shall do below. To the extent that we find reciprocal causation this will indicate that the effects of issue preferences have been inflated. Tillie (1995) demonstrates that it is indeed quite possible to commit causal mis-attributions if only actual party choice is analysed. Of course, in strict two-party or two-candidate contests this argument does not hold, because then, but only then, the complete rank order of preferences is known by knowing which is first. 'All parties' has to be understood as all parties deemed sufficiently important to be included in the question. In general this includes all parties represented in the (national) parliaments, as well as those which, on the basis of contextual knowledge (such as opinion polls) could have been expected to gain such representation in a national election held on the same day. The number of parties about which this question (sometimes referred to as the 'probability to vote' question) was asked ranged from a low of five in Greece and Portugal to a high of nine in The Netherlands. The total number of parties included over all countries amounts to 98 (see the appendix to Chapter 7 for details). Some electoral systems allow the expression of multiple preferences. One can think of STV. the German dual ballot system, and the Luxembourg system of multiple votes with panachage. As the analysis in this chapter will show, however, the 'probability to vote* questions pose no special problems in these cases. To say whether one would ever vote for a party does not require the respondent to distinguish between the various methods by which he or she might do so.

Policy Preferences and Party Choice

183

10 See van der Eijk, Franklin, and Mackie (1996) for details about percentages of voters voting identically or differently in European and national elections. In most cases where party choice was not the same as it would have been in national elections at the same time, the party actually chosen in the European election was one that was also given a high 'probability to vote* score for national elections. We may therefore assume that distortions caused by such differences in choice between national and European contexts are minimized. Moreover, analyses of such differences in party choice by Oppenhuis, van der Eijk, and Franklin (1996) suggest that they are not so much the result of different sets of substantive considerations that voters use for each, but rather of different (strategic) weighting of the elements of one and the same set of political considerations. This implies that analysis of these responses yields results that are also valid for party choice in European elections. An advantage of focusing the 'probability to vote* question on national rather than European elections is that this avoids the risk that the low-stimulus character of European elections may cause high non-response. 11 Many other analyses can be performed to demonstrate that the replacement of actual party choice with this specific set of preference scores for each of the parties is justified. The most important of these is an unfolding analysis which demonstrates that the scores on the 'probability to vote* questions can be understood as emanating from the same latent factors for all parties, hence the origins of this score for the party which one actually voted for are the same as for all other parties. Such validating analyses have been reported in detail by Tillie (1995). 12 Obviously, these ways of arranging data yield a matrix containing a very large number of 'cases': the sum over all countries of the number of respondents in each country times the number of parties for which the 'probability to vote' question was asked. After deleting missing data (but before weighting—see below), this data set contains 85,853 records. 13 A number of conditions have to be fulfilled for these variables to have any appreciable effect on the explanation of party preferences. Consequently, to the extent that their explanatory power is low, this may be the result of a number of causes; parties not being sufficiently differentiated in terms of their issue priorities, voters being insufficiently cognizant of those differences, or voters not reacting to any such differences in their party preferences. Conversely, to the extent that these variables do contribute to the explanation of party preferences, the underlying logic of issue-ownership and its consequences is empirically vindicated. 14 Party identification measures were deliberately disregarded, as we consider the concept inappropriate for the analysis of electoral behaviour in the cleavage-structured European multi-party systems; see van der Eijk and Niemoller 1983. 15 Taking religion as the example to explain this procedure, we conducted a set of regression analyses (one for each party in each country) in which

184

16

17

18

19

Representation the preference score for that party was estimated from each voter's religious denomination and frequency of church attendance. The results indicate to what extent, in that specific country and for that specific party, religious differences between people determine their preference for that party. These effects of religion on preference are encapsulated in the regression predictions made for each respondent, which in statistical parlance are referred to as the y-hats. These y-hats are simply linear transformations of the original independent variable, and may therefore be used again as predictors of party choice. The y-hats are saved and added to the stacked data set, yielding a variable that is comparable for all parties (and countries) and that can be referred to as the "predicted religious effect*, or, more briefly, 'religion'. The actual variable added to the stacked data set is the deviation of the y-hats from their mean. This still encapsulates the variance in party preference caused by religious differences but prevents the average of these y-hats from being subsumed in this variable, an outcome to be avoided because in a stacked data set these averages cause correlations with party preference that cannot be attributed to religion and that would disturb subsequent analyses. See van der Eijk, Franklin, et al. (1996), Chapter 20, for a more detailed account of this procedure. In the procedure to construct these variables each of these issues was scored as follows: 0: not mentioned as one of the three most important problems; 1: mentioned as 3rd most important; 2: mentioned as 2nd most important, and 3: mentioned as the most important problem by the respondent. This variable was constructed using the same y-hat procedure explained above for religion. As predictor the cumulative score on a scale consisting of four items was used, referring to whether or not one would feel 'sorry' were the EU dissolved, whether or not EU-mernbership is regarded as beneficial for one's country, whether or not EU-membership is 'a good thing", and whether or not one is 'in favour' of European unification. For details on the scale and its construction, see van der Eijk, Franklin, et al. 1996, Appendix B. The 12 issues in 1989 included four country-specific issues and the following common issues: unemployment, stable prices, arms limitation, environmental protection, European unification, agricultural surpluses, EC membership of Turkey, and realization of the SEM. The difference in explanatory power between the 1989 and 1994 data cannot be attributed to the somewhat different selection of issues, as five of the common issues are identical in both years, while the differences in their respective explanatory power for those five is of the same magnitude as for the entire set (for details about these 1989 analyses, see Qppenhuis 1995, Appendix 8). In 1989 respondents were asked to indicate the importance of each of these issues in two different ways: by rating the importance of each, and— as in the 1994 survey analysed here—by indicating the three most impor-

Policy Preferences and Party Choice

185

tant ones (so-called rankings). By far most of the explanatory power of these issues in 1989 derived from the rating Information, while the rankings contributed only occasionally to the explanation of party preferences (Oppenhuis 1995, Appendix 8). When we analysed the 1989 data using only the ranking information, the average R1 value comes close to the four per cent reported in Table 8,3. 20 In Chapter 7 of this volume (Table 7.2), van der Brug and van der Eijk showed that exceedingly low agreement exists between voters about which party is best at which problem, particularly when compared with the perceptual agreement of parties' left-right positions, and even in comparison with the perceptions of parties* positions on the three non-salient position issues. This finding seems mainly explicable by interpreting the response to the 'best at' questions as preferential rather than perceptual. 21 Although party size is a consequence of the number of people who support a party, and thus in the long term may be seen in part as the consequence of issue preferences, at any one election the sizes of parties can be regarded as exogenous. For a more detailed discussion of the interpretation of party size as reflecting voters' strategic considerations, see van der Eijk, Franklin, et ai. 1996: 352-3, and Tillic 1995. 22 One could argue that adding left-right to the equation after (rather than before) the position and valence issue variables would increase the contribution to the explanation of the latter. Still, such an alternative order of inclusion of variables in the equation does not seem very compatible with the notion that left-right orientations function as political schema (Conover and Feldman 1984; Kerlinger 1984), super issues (Inglehart 1984), or ideological identification (van der Eijk and NiemSller 1983).

9 Issue Congruence

JACQUES THOMASSEN AND HERMANN SCHMITT

THE purpose of this section of the book is to assess the functioning of the system of political representation in the European Union. In preceding chapters, we have looked into processes that might possibly lead to a congruence between members of parliament and their constituents. However, the proof of the pudding is in the eating. So far nothing has been said about the outcome: the extent to which any of these processes, or other processes for that matter, really lead to a congruence between members of parliament and their constituents. The main research question in this chapter is to what extent such congruence does exist. It is a matter of dispute as to what extent the total agreement on policy matters between voters and their representatives is an indicator of a smoothly-functioning democracy. Perhaps in a perfect world, such total agreement would exist. However, it would be naive to expect perfect congruence in any real system of political representation. Demanding perfect congruence between represented and representatives would deny a basic characteristic of political representation and of modern society more generally: the division of labour. Representatives do something that the represented themselves cannot do. We have introduced in Chapter 1 of this book the Responsible Party Model of political representation as a benchmark to evaluate the system of political representation in the European Union. In this model, political parties are not the passive implementers of the will of the people. Quite the contrary, the initiative in the process of political representation is with them. Parties develop policy platforms and try to

Issue Congruence

187

mobilize support for it. By supporting one party's platform and rejecting others, voters generate agreement between the policy positions of political parties and their own policy views. The result of this process is that like-minded voters sort themselves out between the different parties. This is very much a process of'representation from above': political elites set the agenda and direct the process of opinion formation. It is perfectly consistent with this view that the elite marches ahead of the masses in developing policy views. However, inasmuch as this is the case, one might start to wonder what the democratic character of the system of political representation would be if agreement on political issues between the political elite and the mass public were irrelevant. Of course, it is relevant. As Hannah Pitkin in her classic book on political representation concludes about the mandate-independence controversy: The representative must act in such a way that there is no conflict, or if it occurs an explanation is called for. He must not be found persistently at odds with the wishes of the represented without good reason in terms of their interest, without a good explanation of why their wishes are not in accord with their interest (Pitkin 1967: 209-10).

Therefore, although the argument that representative democracy requires perfect agreement between representatives and represented cannot be sustained, one would turn the world upside down by arguing that the extent to which representatives and represented agree or disagree about policy matters is not relevant. However, not each and every instance of disagreement (or lack of congruence) is a pathological phenomenon. There is another consideration that is relevant for establishing how effective the complex system of political representation in the EU actually is. In Chapter 1, we distinguished two routes of political representation in the European Union: the European and the national channel. In brief, the European channel applies to the supranational mode of EU government, while the national channel applies to the intergovernmental mode of EU government. Both channels must be taken into account in order to assess the quality of the system of political representation in the European Union. Thus, while the emphasis in this chapter will be on the degree of consensus between the members of the European Parliament (MEPs) and their electorates, we will also see—

188

Representation

where possible and appropriate—to what extent the members of national parliaments (MPs) are representative of their voters. Establishing the degree of congruence between voters and MEPs is not the whole story. It is equally important to assess how this congruence is brought about. The real challenge for the European Union in general, and for its system of political representation in particular, is to overcome national borders. An important criterion to assess whether the EU system of political representation is developing into a truly European system of representation is the degree to which political debates and political alignments cross national borders. A truly European system of political representation can only exist when policy preferences are shared across borders. If that is not the case, if the citizens in each nation have policy concerns and preferences that are totally different from those of the citizens of other member-states of the Union, there might still be a perfect congruence between the voters of a particular nation and their representatives in the European Parliament. But we would hardly speak of a modern system of political representation, let alone of representative democracy, that is comparable to the system of representative democracy in any of the member-states. Such a parliament would have more in common with the medieval parliaments mentioned in Chapter 1, representing the interests and grievances of a particular territory vis a vis a central power responsible for the national interest (read: the common European interest).1 Therefore, the first question we will take up in the next section is to what extent people across Europe do have common concerns. Next, before looking into the issue congruence between voters and MEPs of the same party and nation, we will address the very general question as to what extent the European Parliament as a whole is representative of the European people as a whole. More than anything else, this seems to be the most essential question since the Treaty on European Union was signed at Maastricht: is the European elite still living in the same European world as the European people?

Issue Congruence

189

WHAT EUROPEAN CITIZENS ARE CONCERNED ABOUT

In order to find out what their major concerns are, Europeans were asked to select the three most important problems from a list of 11 preselected problems. To do justice to the possible differences in concerns between different countries, four out of this list of 11 problems were 'nation-specific'. The other seven were the same in all countries.2 As a result, it is difficult to compare issue priorities between countries. One strategy to overcome these problems is to compare synthetic issue domains rather than generic issues; this was pursued in Chapter 3, Here we will follow an alternative strategy by concentrating on the comparison of the seven common issues. While this ignores more or less important parts of the political agenda in each country, a few conclusions are still possible. Table 9.1 shows which issues are the most important for the people in the EU member-states. Everywhere unemployment is by far the most important. As far as unemployment is concerned, there can be no misunderstanding about a common concern among European citizens. Beyond that, however, comparisons become complicated: in many countries nation-specific issues are among the most important. If one leaves the nation-specific issues out, which for obvious reasons is a risky procedure, the correspondence between the political agendas of Europeans appears to be high. A computation of the complete matrix of rank-order coefficients of the issue agendas of the 11 countries from which data are available (data not shown) shows that 41 of the 55 coefficients are equal to or greater than 0,75. We take this as a provisional indication of a substantial degree of communality in EU citizens' concerns. THE R E S P O N S I V E N E S S OF THE EUROPEAN PARLIAMENT How responsive is the European political system to these concerns? A very direct way to answer this question is to see to what extent the members of the European Parliament share the concerns of the people.

TABLE 9.1. Issue priorities of EU citizens*

Unemployment Nation-specific 1 Crime Nation-specific 2 Environment Stable prices Immigration Nation-specific 3 Nation-specific 4 European unification Agricultural surpluses

Luxem- Nether bourg lands

Belgium

Denmark

Germany

Greece

Italy

Spain

France

Ireland

65 18 15 19 11 12 19 14 17 4 1

52 22 20 3 49 7 22 2 12 5 5

68 11 20 15 23 17 19 5 13 5 2

56 25 11 35 16 16 2 26 7 3 2

71 50 29 9 12 11 7 4 2

76 52 6 24 9 7 3 2 15 2 2

81 24 10 25 9 10 16 18

74 31 23 7 8 8 7 33

3

3 1

9 1 1 6

2

6

3 1

4 1

72 4 21 16 35 25

52 25 33 13 23 5 16 14 8 6 2

Portu-

gal

UK

70 28 7 33 6 16 2 20 30 1 2

64 27 31 25 12 9 7 14 4 2 3

Source: European Election Study 1994, weighted data. * Figures are scores averaging for any issue the national proportion of mentions as most, second- or third-most important in such a way as to count the 'most important' mentions three times, the 'second most important' twice, the 'third most important* once and divide the resulting sum by three.

Issue Congruence

191

Table 9.2 compares the importance which representatives and represented attach to each of the seven common issues. There is obviously a remarkable consensus between MEPs and the EU citizens about which of these issues are more or less important. The European people and their representatives in the European Parliament agree that unemployment is far more important than any other problem. Both groups also seem to agree on the importance of protecting the environment. TABLE 9.2, Issue priorities compared: EU citizens, MEPs, and MNPs EU citizens

67 19 16 11 11 3 2

Unemployment Crime Environment Immigration Stable prices European unification Agricultural surpluses

N

12,800

MEPs

73

9 11 6 1 23 2 314

MNPs

78 10 11

6 •-* J*

11 1 1,266

Source: European Representation Study,

However, MEPs have clearly different opinions than the public at large about the importance of two other issues. Firstly, and not surprisingly, MEPs stress problems relating to European integration more than the general public does. Secondly, the general public is more concerned about issues of physical security, crime in particular. The discrepancies are not dramatic; still, the pattern is not unknown from earlier representation studies. Political elites generally tend to be more libertarian than the mass public. This is less a reflection of their political role than of their belonging to the better-educated and wealthier part of society. Therefore, in this respect, the differences in policy concerns between the mass public and MEPs reflect a general discrepancy between political elites and the mass public and can hardly be seen as a specific failure of the European system of political representation. We can easily test this interpretation by comparing the degree of congruence between MEPs and the mass public with the congruence between the mass public and the members of the national parliaments, who were asked the same questions. As one can see in the third column of Table

192

Representation

9.2, the differences between any of the three groups are limited. But where there are differences of opinion, the consensus between the two elite groups is higher (r = 0.96) than between the mass public and either the members of the European Parliament (r = 0.66) or the members of the national parliaments (r = 0.60). Just like their colleagues in the European Parliament, the members of the national parliaments are inclined to give a little less weight to problems of crime than the mass public does, but again the differences are rather small. A major difference between national MPs and their colleagues in the European Parliament is that European integration is lower on their agenda than it is among the latter. But overall, these subtle differences do not add to much. Both members of the European Parliament and their national counterparts seem to be pretty much in touch with the concerns of the European people. Obviously, we can also compare the policy views of European voters and their representatives in the European Parliament by means of the three position issue scales and the left-right scale on which most of the analyses of Chapters 6 to 8 are based. The distributions of European voters and MEPs are presented in Figures 9.1 to 9.4. A few features in these figures stand out. On the left-right continuum, the distributions of voters and MEPs more or less follow the same pattern. Compared to the voters, however, the distribution of the MEPs clearly leans to the left. The median and the mean position are 1.0 and 0.8 apart, respectively. This finding is well known from many singlecountry representation studies (Thomassen 1976; Converse and Pierce 1986; Esaiasson and Holmberg 1996). Whatever the cause of this phenomenon may be, it is obviously not related to the special context of European politics. On the issue of an employment programme, the distributions are very similar. Both voters and MEPs lean heavily towards an active employment policy of the Union. On the two remaining issues the situation is totally different: regarding both the abolishment of national borders and a common European currency, we find a wide discrepancy between the preferences of voters and the positions of their representatives in the European Parliament. As one can see in Figure 9.3, almost half of the MEPs have no reservations about continuing to do away with national borders.

Issue Congruence

Flo 9,1. Voters' and MEPs' positions; left-right

FlO 9.2, Voters* and MEPs* positions: employment

193

194

Representation

FIG 9.3. Voters' and MEPs' positions: borders

FIG 9.4. Voters' and MEPs' positions: currency

Issue Congruence

195

There is far less enthusiasm among the mass public. Not more than one in five EU citizens supports open borders without any reservations. The mean score of the voters is 2,5 scale-points away from that of the MEPs, the median 3 points. In the case of the common European currency, the distance between voters and MEPs is even larger. Whereas more than 60% of the MEPs take the most outspoken position in the direction of a new common European currency, less than 20% of the voters do so. Almost 30% of the voters locate themselves at the other extreme of the scale: they want to keep their own national currency without any hesitation. Here, mean scores are some 3 scale-points apart, the medians 4 points. It becomes clear that on the two straightforward Union issues (as opposed to common concern issues) the distances between the members of the European Parliament and the European electorate are enormous. To what extent one might really conclude that voters and MEPs are 'living in different European worlds' depends somewhat on how one should interpret these differences. In the case of a single European currency it is hard to see how else one could understand them than as an enormous discrepancy between elites and mass publics. In the case of open borders the discrepancies are hardly smaller, but at least there is an alternative interpretation that does not necessarily relate to the European project; in the survey question the borders item is related to combating crime. We already noticed that crime is a problem that ordinary people worry about more than representatives do; similar discrepancies between voters and members of parliament were found before. If this interpretation is correct, it would not reduce the distance between voters and MEPs, but at least the deviating opinions of the public would not necessarily have to be interpreted as an anti-European mood.

THE PROCESS OF POLITICAL REPRESENTATION

What we have seen so far about the way the members of the European Parliament reflect the concerns and issue positions of the European people leaves us with a mixed picture. However, limiting our analysis to this general level would do little justice to the political process. People do not necessarily agree, neither on policy positions nor on the rela-

196

Representation

live importance of different problems in society. The very essence of representative democracy is to give people with different political concerns and policy positions the opportunity to vote for a candidate or a political party that best shares their concerns. Therefore, a test of the effectiveness of the system of political representation is its capacity to bring voters and like-minded parties and politicians together. An additional test for the effectiveness of a European system of political representation is the extent to which the opinions of like-minded people across the borders of Europe are reflected by the policies of transnational party groups rather than by those of national delegations. The results of the preceding chapters should enable us to predict with some accuracy the issue congruence between European party groups and their voters. The purpose of the preceding chapters was to test how far the requirements of the Responsible Party Model are met. Five requirements were tested: the distinctiveness and cohesion of party groups, whether people correctly perceive the issue positions of parties, whether people themselves have an opinion on relevant issues, and whether they vote according to these opinions. The better these requirements are met, the closer together we can expect the issue concerns and policy positions of parties and voters to be. How well the five requirements are met differs from issue to issue. In the case of left-right the conclusion is unequivocal. Competition and cohesion are high, people in general know at least where the major parties stand relative to each other, and although fewer people are able or willing to assign themselves a position on the scale, on average 86% still do. The proportion of variance in party choice explained by leftright distance is substantial. Moreover, nationality hardly plays a role here. Therefore, on this dimension the requirements of the Responsible Party Model are satisfied reasonably well (Table 9.3). In the case of the three specific issues the situation is completely different. The desirability of a European employment policy is substantively related to the basic left-right dimension: left-wingers tend to favour an employment programme, while right-wingers trust the market forces. This is clearly apparent on the elite side, where the major party groups take clear and distinct positions very much in line with their positions on the left-right scale.

Issue Congruence

197

TABLE 9,3. The requirements of the Responsible Party Model as they are met in the European Union Left-right

Employment Borders Currency

Distinctive parties'* Cohesive parties'3

3,4 1.3

4.1 2.6

.2 2.3

,9 2.2

Voters perceive parties'1 Voters reveal a preference* Ideology/ issue voting6 "Nation effect on MEP preferencef Nation effect on voter preference*

81 86 13 2 1

60 91 2 17 3

62 94 2 21 5

59 93 2 17

13

Source: European Representation Study, (a) difference of mean position of the members of the two major EP groups; (b) standard deviation of MEP responses on a ten-point scale, averaged across EP groups; (c) average proportion of voters able to locate the most important parties; (d) proportion of voters revealing a preference; (e) per cent variance in vote probabilities explained by ideological or policy distance; (f) per cent variance in MEPs* preferences explained by national affiliation; (g) per cent variance in voters' preferences explained by national affiliation.

However, on the voters' side things look very different: on average, 40% are unable to locate political parties on this issue, and although no less than 91% of the respondents claim to have an opinion, there is little relationship between these positions and party preference. Therefore, in so far as we find any congruence between voters and parties on this issue, it can hardly be due to the mechanism of the Responsible Party Model. For the two remaining issues the conclusion is more or less the same. On the common currency issue the major party groups are flocking together on the pro-European side of the scale. We find much the same phenomenon with the open borders issue. This has few consequences for the cohesiveness of party groups: precisely because the party groups are all concentrated at one extreme of the scale, the internal cohesion of the party groups is no less than on the other issue. The extent to which the requirements of the Responsible Party Model are met on the voters' side can hardly be independent of the supply side of politics. The degree to which political parties take unambiguous and distinct positions is bound to be reflected in the voters' perceptions and behaviour. This became clear in the analyses presented

198

Representation

in Chapters 6 and 7. Just as with the issue of unemployment, the number of voters able to locate the parties3 on one of these two Union issues differs dramatically from the number able to locate the same political parties on the left-right dimension. In neither case is the average percentage of voters who even try to locate political parties much higher than 60%, Still, those who do try appear to have a reasonable idea of where the political parties are located relative to each other, although in general they tend to underestimate the parties' pro-European character. Also, the percentage of respondents claiming to have an opinion on these issues is in, both cases over 90%. Nevertheless, the relationship between opinions on these issues and party preference is very low. The extent to which the electoral process sorts out voters over different parties is hardly related to these issues. On the basis of what we have found out about the preconditions of an effective system of political representation according to the Responsible Party Model, we can only expect that the congruence between European party groups and the voters of their constituent parties on these issues will be low. In the next section we will see whether these expectations are borne out.

POLICY CONGRUENCE BETWEEN P A R T I E S AND VOTERS

In Figures 9.5 to 9.8 the mean scores of members of the European Parliament and their constituents on the four issue scales are plotted party by party. In addition to the EP groups, the constituent national parties are presented. If there were perfect agreement between the representatives of a particular party and their constituents, all parties would be on the diagonal. The further off the diagonal parties are positioned, the less agreement there is between voters and the MEPs of that particular party or party group, A first glance at the figures warrants a few obvious conclusions. Firstly, and this is only a confirmation of what we already knew from Figures 9.1-9.4, there is an enormous difference between, the left-right dimension on the one hand and the three specific issues on the other hand. On the left—right dimension, parties at least follow the diagonal, although most parties are above it, meaning that the political elites in

Issue Congruence

199

general are inclined to position themselves more to the left than the voters. As observed above, this is a well-known pattern found in most studies of elite-mass congruence in West European countries and is not specifically related to the politics of the European Parliament, The three specific issues show a totally different picture. The question of an active European employment programme is an example of an issue where there is hardly any variance among voters, at least none strongly related to party preference. Voters in general, almost independent of their party preference, are in favour of an employment programme, whereas the MEPs are spread over the scale much as one would expect from the basic left—right perspective. As a consequence, the parties on the left of the political spectrum are close to their voters (or at least on the same side of the scale). The further to the right parties are and the more they favour a free market solution to the problem of unemployment, the farther away they are from their voters, The issues of a single European currency and of open European borders, the two issues most closely related to the pro- vs. anti-European dimension, confirm what we found above. The major European party groups and their constitutive national party delegations are all on the pro-European side, whereas the electorate, almost irrespective of their political preference, is grouped together around the middle of the scale. As a consequence, small anti-European parties such as the Europe of Nations group are closer to their voters than the major pro-European catch-all parties are. This leaves us with a somewhat paradoxical conclusion. European political party groups are reasonably representative of their voters as long as we refer to the basic political dividing lines that have nothing to do with the European project. However, on Union issues (i.e. issues related to the development of the European Union as a political system) MEPs appear to be out of touch with those who have elected them. Why is this so? One possible explanation is that these differences in policy preferences occur because members of the European Parliament are members of the European Parliament. Several processes are likely to make MEPs more pro-European than the average citizen. Firstly, the recruitment process of the major political parties with a positive attitude towards Europe will probably attract candidates who sympathize with the European project (in anti-European parties the opposite might

200

Representation

FIG 9.5. Left-right congruence of European parties' voters and MEPs

FIG 9,6, Policy congruence of voters and MEPs: employment

Issue Congruence

FlG 9,7. Policy congruence of voters and MEPs: currency

FlG 9.8. Policy congruence of voters and MEPs: borders

201

202

Representation

happen). Those feelings might be strengthened by the socialization process within the Parliament once they have been elected. In the institutional framework of the European Union it is almost the natural role of the European Parliament to promote an integrationist position (Marsh and Wessels 1997), which is bound to have an impact on the attitudes of individual members. Also, they are literally far away from their constituency, making it more difficult to fee] the pulse of the people. To what extent the MEPs' pro-European stance is due to the specific context of the European Parliament can easily be tested by comparing their attitudes with those of the members of national parliaments. None of the factors just mentioned applies to them. 'Europe' normally plays no rote in their recruitment. As far as the institutional context plays a role it should work in the opposite direction. It is the natural role of national parliaments to protect national sovereignty, and as such, one would expect members of parliament to take a more conservative position on European issues. And the least one can say about their contacts with the voters is that it is hard to think of a reason why they would be less in touch with the mood of their constituency than their European counterparts. In Figure 9.9 the positions of MEPs on the issue of a common currency are plotted against those of members of the national parliament of the same party. The plot for open borders (not shown here) is almost identical. If our expectations are correct, all points connecting MEPs and MPs should be below the diagonal, revealing a subtle proEuropean bias of MEPs. This is mostly borne out, but only just. What catches the eye more than the small differences between the members of the two sorts of parliaments is the cloud of parties in the upper right corner of the diagram meaning that, in most parties, both MEPs and MPs are strong supporters of a single European currency. As a consequence, the figure (not shown here) plotting voters and members of the national parliaments on this issue looks very much like Figure 9.7. Therefore, in so far as one is inclined to see a discrepance on these issues between voters and political elites as a problem, it is not one of the Eurocracy out of contact with the mood of the European people. Rather, the political elite quite generally seems to be marching ahead of the people.

Issue Congruence

203

FIG 9.9. Policy congruence of MEPs and MNPs: currency

NATION AND POLICY CONGRUENCE A truly European system of representation implies that political debates and political alignments cross national borders. If policy positions were defined by nation rather than by party position, we would not speak of a truly European system of political representation, even if in national terms there were a perfect congruence between voters and members of the European Parliament. However, if policy positions are defined by national differences rather than by party (or any other transnational alignment), it is still important to know how well the policy views of the MEPs from a particular country reflect the policy views of their national electorates. If they do, we might conclude that MEPs are obviously in touch with the mood of the electorate in their country, while a truly European system of political representation does not exist. From the data in Table 9.3 we have learned that positions on the

204

Representation

left-right dimension are highly correlated with party affiliation, both at the level of the MEPs and of the electorate. Therefore, it came as no surprise to find in the last section that MEPs are, in terms of party, very representative of their voters on this dimension. At the level of the voters, the issue of employment policy was related neither to party affiliation nor to nationality, which explains the lack of congruence between the party groups in the European Parliament and their voters, There is no reason to expect a different result if one compared the opinions of MEPs and voters by nation. In the case of a European currency and, although to a lesser extent, of open borders, the situation is different. At the level of the voters and among the MEPs of the major party groups, opinions on these issues are more related to nationality than to national party. Therefore, we might expect the differences between the voters of different member-states to be reflected in similar differences among the MEPs. Figure 9.10 shows the mean scores of both MEPs and the electorate of each country on the European currency issue.4 There are clearly huge differences both between countries and between the MEPs and the voters in a large majority of the countries. Whereas the mean scores for the MEPs of seven of the 13 countries in this figure are no more than one scale-point removed from the most extreme position on the pro-European side of the scale, the Swedish and Danish MEPs are about half the length of the scale removed from their colleagues from these seven countries. These differences are a reflection of—or perhaps one should say: are reflected in—the opinions of the mass public in these countries. It is clear that in each country, with the possible exception of Sweden, the MEPs are more in favour of introducing a European currency than the people at home are. However, taking this general tendency into account, one can also see that countries tend to differ from each other according to the same pattern on both levels. The largest difference between MEPs and mass public exists in Germany: whereas German MEPs belong to the strongest supporters of a common currency, their voters prefer the company of the well-known Euro-sceptics from Denmark, Sweden, and Great Britain. Again, we might expect the members of the national parliaments to be somewhat more in line with the opinions of their constituents. However, after what we have seen in Figure 9.9, it should come as no surprise that this is hardly the case. Although the differences between MPs and voters are less than between MEPs and voters, they are still

Issue Congruence

205

FiO 9.10. National congruence of MEPs and voters: currency considerable and follow the same pattern. On the borders issue, the same pattern is repeated (data not shown).

CONCLUSION

This is the last of a series of chapters (Chapters 6-9) testing whether the requirements of the Responsible Party Model are met at the European level, and whether and to what extent, at the end of the political process, parties and voters of similar political outlook find each other. The results of our analyses are mixed and somewhat paradoxical. Firstly, our analyses point to the fact that the European Parliament (with the present membership of the Union) can build upon a part}' system that is not insurmountably different in the several memberstates. In almost all national systems, the dominant dimension of party

206

Representation

competition is the left-right space (Sard and Sartori 1984); this dimension can pretty easily be aggregated to the European level, facilitating efficient communication at the European level as much as at the national level. Political parties compete on this dimension in both the national and the European system; voters recognize the differences between parties and take into account their own distance to the parties on this dimension when voting. As a result, or at least consistent with this, there is not much of a discrepancy on this dimension between voters and members of the European Parliament, Things look very different once we apply the same tests to issues relating to the European Union as such. On these issues there is clearly competition between political parties, as there is a large spread of parties, making it possible for the voters to choose from a wide array of positions. However, that conclusion needs to be nuanced by the fact that the three major party groups, accounting for 62% of the seats in the European Parliament, hardly differ with respect to these issues. Also, on these issues the national element sneaks into the party groups, making it more difficult to maintain party discipline. While roughly a majority of voters still think they know where the parties are on these issues, in many cases they are wrong. We also found that there is hardly a correlation between party preference and voters' positions on these issues. Therefore, it came as no surprise that there is not much systematic congruence between voters and European party groups on these issues. Across the board, voters* attitudes appeared to be less proEuropean than those of the political elites, whether they were members of the European Parliament or of the national parliaments. Therefore, the strange paradox is that the system of political representation at the European level, measured against the criteria of the Responsible Party Model, is feasible as long as it is not concerned with issues that refer to the European Union as such. Of course, this is one of the traditional criticisms of the functioning of the democratic system in Europe, As a result, a number of authors (Andeweg 1995; Bogdanor 1989) have pleaded for a total reshuffle of the European party system, where parties would distinguish themselves on issues referring to the future development of the European Union. In the final chapter we will come back to this discussion and argue that such a reshuffling would be a development for the worse. Whether the discrepancy between the political elites and the mass publics on European issues should be considered a dramatic short-

Issue Congruence

207

coming of EU democracy is a matter of perspective. Firstly, the size of these differences is not unique to the European Parliament. Similar differences on other issues were found in the relationship between national political parties and their electorates as well (Thoraassen 1976; Esaiasson and Holmberg 1996). This, of course, does not change the facts, but it warns us against a simplistic interpretation in terms of a special anomaly of the European Parliament. Also, our observations refer to one point in time, although political representation is a dynamic process. If one accepts the perspective of 'representation from above' (Esaiasson and Holmberg 1996), the evolution of public opinion tends to start from the top, among the political elites. In so far as they try to win over the people to their side, there is nothing wrong with this. As Henry Kissinger observes in his description of Roosevelt's role in preparing the American people: for intervention in World War II: All great leaders walk alone. Their singularity springs from their ability to discern challenges that are not yet apparent to their contemporaries (Kissinger 1994: 370). One might add that it is probably the singularity of a representative democracy that the judgement on leaders is not left to history. Contemporary voters determine how much leeway to give the leaders that march ahead. It is the irony of the situation described in this part of the book that the voters in general seem unaware how far ahead the leaders are. Even worse, as we will see in Chapter I I , is that the political elite does not know either. NOTES 1

2

But there would still be a clear difference from a pre-modern system of political representation. If the European Parliament did not offer a platform for expressing interests and policy views that cut across national interests, it is very unlikely that any of the other institutions, perhaps with the exception of the Commission, would be able to perform this role. Certainly the Council of Ministers or the European Council would hardly be in such a position, representing more than any of the other institutions the interests of the member-states. They were selected by national experts, reflecting their views of what citizens in their countiy were likely to consider important problems. For precise question texts, refer to Chapter 3 above, where this instrument was first introduced.

208 3

4

Representation Of course, in the European elections voters could not vote for any of the European party groups but for one of their constituent national parties. Therefore, they were asked to locate not the European party groups, but the major national parties on various dimensions. At the level of the electorate no data are available for Austria or Finland.

10

Whom to Represent? Role Orientations of Legislators in Europe

BERNHARD WESSELS

INTRODUCTION

A very important aspect of political representation at the European level that has not been dealt with in the previous chapters is whom a member of the European Parliament should represent. The answer to this question very much depends on the kind of Europe one aspires to. Some claim evidence that national interests prevail in the debates and political decision-making processes. As argued in Chapter 1, however, in an integrated Europe one should expect members of the European Parliament to represent the European people-—assuming there is one. Thus, the question we are dealing with here is how the members of the European Parliament actually conceive of their representational role with respect to whom they want to represent. Since roles by definition cannot be thought of without social institutions, the question put forward here is to what degree institutional settings bring about particular role orientations. More precisely, this chapter will analyse to what extent institutional differences between the member-states of the European Union are constraining the attitudes of representatives about one major aspect of the rules of the game, namely whom to represent.

210

Representation

We will explore role orientations of the members of the European Parliament and compare them with those of members of eleven national parliaments. In a first step, the effect of social background characteristics and representational experience on role orientations will be assessed. In a second step, a cross-national approach will be used to explain role orientations by macro-variables, that is, system characteristics of the respective polities. The dependent variable measures a two-dimensional concept of roles which distinguishes the group specificity and the regional scope of the representational focus. This approach clearly goes beyond the scope of previous research. Past studies of role conceptions have mostly been limited to a single legislature. As a result, they focused on differences between individual legislators rather than on those between legislatures (Jewell 1970: 466). With respect to the regional focus, the effect of electoral systems will be tested. Electoral systems shape the competition between candidates and create different degrees of uncertainty. We will test the hypothesis that this has an effect on the representational focus of deputies. With respect to the group specificity of representational foci, we will test whether and to what degree the encompassiveness and inclusiveness of interest group systems affect the role orientation of representatives. Going beyond the nation-state, we will attempt to explore the sources of a European focus of representation. Here we will test a set of hypotheses related to the position and experience of a country in the European Union. C O N C E P T U A L I Z I N G R E P R E S E N T A T I O N A L ROLES 1. The Conception of Roles: General Conclusions from Previous Research Research into representational roles has a long tradition and a classical starting point. In his famous speech to the electors of Bristol in 1774, Edmund Burke called for a legislator who would defend national rather than regional interests and follow not instructions but his own mature judgement. This distinction between whom to represent and how to do it became known as the focus and style dimensions of representational roles.

Whom to Represent?

211

From a sociological perspective, roles can be defined as a set of expectations connected with a particular social position in a society (Dahrendorf 1969). The role of a legislator consists of the rights, duties, and obligations connected with being a representative. His or her role orientations are shaped by expectations towards his or her position, rights, duties, and the like. They can be regarded as individual translations of expectations towards this position. As Wahlke cum suis put it: The chief utility of the role-theory model of the legislative actor is that, unlike other models, it pinpoints those aspects of legislators' behavior which make the legislature an institution (Wahlke, Eulau. Buchanan, and Ferguson 1962: 20),

Two aspects of role analysis are hidden in this argument: the effects of role orientations on the behaviour of legislators, and the sources of legislative roles. These two aspects have been the major topics in research into legislative roles. The starting point of analysis, however, should be to explore the sources of roles: Unless we can explain why legislators differ in their role concepts, role analysis may strengthen our predictive powers without increasing our ability to explain the legislative process. If our investigation should show that some role orientations are simply surrogates for some other variables—education, personality, and legislative tenure—that perhaps could be measured more precisely, then role would appear to be less useful as a technique of analysis (Jewell 1970: 461).

Following these considerations we will first look into two main sources of roles: personal characteristics and institutional structures. The concept of representational roles became very prominent in the US literature over the last decades (Miller and Stokes 1963; Miller 1988). However, the research findings proved inconclusive. Kuklinsky and Elling (1977) even raised the question whether continuing such research, would be worthwhile, considering the contradictory evidence. After reviewing a great number of studies, Jewell concluded that the evidence concerning possible sources of legislative roles is fragmentary and sometimes contradictory (Jewell 1970: 483). Much the same was found regarding the effects of role orientations; here too, results are centra-theory and mixed (see Alpert 1979). This dilemma of US research into representational roles could be a result of two shortcomings. The first might result from a faulty specification of research. Evidence from Sweden shows, for example, that the

212

Representation

on-average moderate impact of personal characteristics on representational focus increases the more specifically this focus is defined (Esaiasson and Hoimberg 1996: 73). They also find clear behavioural consequences of role orientations in a theoretically expected direction. Again, the more specific the focus, the higher the effect (ibid.: 67). This seems to suggest that the specificity of the definition and of measurement instruments for representational roles was inadequate in prior research: in particular, the concentration on the trustee/delegate distinction may have been misleading. Secondly, the problem might originate in the limited comparative scope of role research. Jewell suggests that broadening the comparative scope might lead to a better understanding of the sources of roles, because one can expect the norms of a political culture and the structure of the political system to have an impact on legislators' individual orientations (Jewell 1970: 465f). In Western Europe, for example, the study of role orientations proved less useful because the role of political parties, so important in European politics, was inadequately accounted for (Hoimberg 1989; Thomassen 1991). Two general conclusions concerning the definition of role conceptions and two general hypotheses concerning the sources of role orientations can be deduced from the argument above. Firstly, role definitions must include the dimensions specificity and scope. The respective hypothesis is that the more specific the representational focus, the more personal characteristics matter. Secondly, sources of role orientations must be studied in a cross-national perspective. The respective general hypothesis is that political structures, including political culture, have an effect on role orientations.

2. Representational Foci Defined: Group Specificity and Regional Scope It is obvious that role foci must be defined in a way that allows us to differentiate between varying degrees of specificity of the collectivity to be represented. This collectivity, however, can be described on different dimensions. The classic dimension is the territorial one, A second dimension usually referred to is functional, defined in terms of interests of particular collective actors. We will define the latter some-

Whom to Represent?

213

what differently from previous studies (see Eulau and Karps 1978) and refer to it as the group dimension. Group Dimension For the group dimension the focus is specified by referring to the groups in question. Three groups of people have been distinguished: i increasing order of specificity they include party voters, the voters in the constituency, and specific groups. Only for the latter category can one speak of functionally defined groups; the group of party voters is politically defined, and the group of voters in the constituency is regionally defined. Party voters are the least specific group, since they are specified neither in terms of regional nor in terms of functional interests. Constituencies have or at least could have specific regional interests, although they might differ internally with respect to party political orientations and clearly do differ with respect to functional interests. Specific groups are defined by particular social, economic, or cultural interests and are most often homogeneous with respect to party political orientations. Furthermore, their specific interests are not cross-cut by regional considerations (Figure 10.1). Regional Dimension For the territorial dimension, the focus is defined by specifying the regional scope. In most studies, only the constituency and the nation can be differentiated in a sensible way. However, since we are dealing with representation in the European Union, this dimension must be extended to incorporate the European level. Thus, regional scope also distinguishes between three levels of specificity: the local, the national, and the European level (see Figure 10,1). From this scheme five rather than six different foci can be derived, as there is an overlap between the group and the regional dimension regarding the voters in the constituency. These five different foci can

FIG 10.1. Role conceptions: representational foci

Whom to Represent?

215

be studied separately and in relation to each other. The latter point is of importance because legislators might not adopt just one but multiple representational foci (see Wahlke et al. 1962; Jewell 1970). In this case, attention has to be paid to possible 'trade-offs* between representational foci, i.e. to the question which focus is more important than others. In Figure 10.1, analytically meaningful trade-offs are indicated by double-headed arrows between foci. DISTRIBUTION OF REPRESENTATIONAL ROLES ACROSS THE EUROPEAN UNION The distributions of these five foci of representation differ quite strongly between parliamentary levels, i.e. between members of the European Parliament (MEPs) and members of national parliaments (MNPs), For both the focus on 'al! people in the country' (33 percentage points) and the focus on 'party voters' (40 percentage points), differences between MEPs and MNPs are strongest in Ireland, The focus on 'people in the constituency' differs most strongly in Greece (27 percentage points), while the difference in the focus on particular groups is most pronounced in Sweden (14 percentage points) (Table 10.1). With respect to the initial question of which model of Europe might be indicated by the prevailing representational foci, it is obvious that the focus on all people in the country, on the constituency, and even on party voters is more frequently cited than the focus on all people in Europe. But there is a lot of cross-national variation: the proportion of MEPs from the 15 member-countries sharing a European focus of representation ranges from 0 (Luxembourg) to 73 per cent (Greece). Based on a somewhat shaky argument of equivalence, one could assume that the European focus of representation means for MEPs what the national focus means for MNPs. Among MNPs, the difference between a national focus of representation and either a party voter or a constituency focus should be more or less equal to that between the European and the national focus among MEPs. If this were borne out by our data, we could easily speak of an 'elevator effect'. The figures tell a different story for most countries, but not for all. In Belgium, for example, the difference between a European focus and a national focus among MEPs is +5 percentage points, while it is +10 percentage points for MNPs with respect to constituency and national

216

Representation

TABLE 10.1. Distributions of representational foci among MEPs and MNPs (figures are per cent) All people in Europe

A B

DK SF F D

GR IRL I L

NL P E S

UK

MEP MEP MNP MEP MEP MEP MNP MEP MNP MEP MNP MEP MNP MEP MNP MEP MNP MEP MNP MEP MNP MEP MNP MEP MNP MEP

30 39 n.a. 25 18 44 n.a. 63 n.a. 73 n.a. 44 n.a. 65 n.a. 0 n.a. 48 n.a. 31 n.a. 42 n.a. 21 n.a. 29

All people in country

70 44 49 63 91 56 58 57 60 91 75 89 56 74 79 75 75 76 54 81 56 67 75 64 38 54

Constituency

60 72 59 13 73 31 71 59 64 55 82 89 76 41 59 75 57 5 0 56 74 69 66 57 50 79

Party voters

80 67 72 75 73 61 29 50 45 64 58 89 49 50 43 75 57 86 74 38 56 78 70 71 71 50

Special groups

50 11 39 0 9 14 8 17 26 27 38 33 27 20 28 25 32 14 15 19 15 28 29 14 0

14

Source: European Parliamentarians' Study. Figures are percentages of representatives who indicated that the respective focus is of great importance to thern (values 6 and 7 on a 7-point scale, where 1 indicates little importance and 7 great importance), n.a. = not applicable, focus. Similar results can be found for France and Germany, but not for the other eight countries for which comparable survey data from

Whom to Represent?

217

both national and European representatives exist. Thus, a simple exploration of the frequencies does not support the 'elevator argument'. The striking differences between countries rather signify a great diversity in the understanding of political representation which, as we will see, follows different rules. E X P L A I N I N G THE FOCUS OF REPRESENTATION Acting against expectations related to one's social position evokes sanctions. In politics, it might cause the end of a career. Parties in charge of recruitment or the electorate might punish an incumbent for not behaving according to their expectations. However, how does a representative know about the expectations connected to his position? Like anybody else, members of parliament have to learn their role during their political career. They learn how to succeed within the party and how to attract voters. The role a representative defines for her- or himself is thus a product of expectations arising in different contexts. There is the narrow context of the institutions in which a career is made. And there is also the wider context of electoral institutions and political culture, Extensive literature shows that parties and other socializing agencies teach individual actors how to behave according to expectations which are connected to a particular position (Herzog 1975; Jewell 1985). This is the individualistic perspective on explaining role orientations. Comparing results from different countries or legislatures provides evidence that individualistic explanations are of some importance, 1, Personal Factors Legislative experience (Wahlke, Eulau, Buchanan, and Ferguson 1962), socialization during the first term (Bell and Price 1975; Badura and Reese 1976), party career (Herzog 1975), and many other variables affect the legislator's choice of a representational role. While no single variable predominated, a number of them were found to be significant in one or more studies. We will focus here on two dimensions of personal characteristics: social background and political experience. Social background char-

218

Representation

acteristlcs include education and church attendance, which were found to be important in several studies. With respect to political experience we shall look into the impact of political representation experience, interest representation experience, and political seniority in party and parliament. Political representation experience includes work in public offices and governments (local, regional, national). Interest representation experience refers to holding or having held office in the party (local, national), in traditional or new interest groups, and in religious organizations. Two explanatory models have been tested against a simple crossnational model: the political representation experience model and the interest representation experience model. Social background and seniority variables were also included. Multiple classification analyses were applied for the first two models, bivariate analysis of variance in the cross-national model. The models were tested for MEPs and MNPs separately. We find a limited effect of personal characteristics on representational foci. Among MEPs, neither of the two models of representation experience is significant. Neither social background nor representation experience have a real impact on role orientations of MEPs. Among MNPs, the explanatory power of the models is equally poor, although some of the coefficients are somewhat more significant. However, some variables have a limited effect on only some of the foci. Systematic conclusions cannot be drawn from these results, given the heterogeneity of findings. The hypothesis that the more specific the focus, the greater the effect of personal characteristics on role orientations is only very weakly supported by the finding that the interest representation experience model has its strongest effect in the case of specific group foci. Comparing the two multivariate representation experience models with the bivariate cross-national model reveals that 'country' in every case is a better predictor of role foci than the multivariate models (Table 10.2). Overall, we find that individual experiences have only limited effects on individual foci of representation. Differences between individuals can hardly be explained by personal characteristics. The simple cross-national model performs much better. This suggests that it is worthwhile to explore what is behind 'country', i.e. to replace 'country' by characteristics of the respective political systems and to investigate their impact on the distribution of representational foci

Whom to Represent?

219

across countries. TABLE 10.2, Individual-level explanation of representational focus Focus All people in Europe (R2/Eta2) -(mean) eta* 10 - sign, of model - highest eta* 10 All people in country (R2/Eta2) -(mean) eta* 10 - sign, of model -highest eta* 10 Constituency (R2/Eta2) -(mean) eta* 10 - sign, of mode! -highest eta* 10 Party voters

Model 1 MNP

MEP

MNP

.039 6.3 .462

.038 5.3 .574 12,0 .061

Model 2 MEP

Model 3 MEP

MNP

.107 32,7 .002

12.0 ,047

.079

.024

.069

.079

5.2

26.3 .082

28.2 .000

7.1

6,7

8.6

.151 13.0

.000 17.0# .061

.022 16.0 §

.004

.058

.019

.181

.112

42.5 .000

33.5 .000

.07!

9.0

9.3

6.6

6.4

4.2

.067 15.0 §

.000 21.0$ .031

.136 15.0 §

.022

.047

,011

.085

,062

29.1 .020

25.0 .000

.047

,109

21.8 .393

33,0 ,000

.059

7.0

(R2/Eta2) -(mean) eta* 10 - sign, of model -highest eta* 10 Specific group (R2/Eta2) -(mean) eta* 10 - sign, of model -highest eta* 10

5.8

3,5

5.9

2,8

.167 15.0#

.000 16.0# ,068

.288 15.0§

.341 8.0 § .035

.056

,049

5.9

6,6

6,8

5.0

.217 13.0 §

,001 17.0 #

.263 13.0*

.000 12.0 #

Source: European Parliamentarians' Study. Model 1 includes the following variables; university education/other; church attendance once a week/less; held/hold office or not in a) local representative body, b) regional representative body, c) national representative body, d) local government, e) regional government, f) national government; party seniority (0-15 years/16-25/26+); parliament seniority (0-5 years/6-10/11+). Model 2 includes the following variables: university education/other; church attendance once a week/less; held/hold office or not in a) local/regional party, h) national party, c) traditional interest groups (professional, trade union, business), d) new interest groups (women, environment), e) religious organization; party seniority (0-15 years/1625/26+); parliament seniority (0-5 years/6-10/11+). Model 3 includes country dummies for 15 (MEPs) and 11 (MNPs) countries, respectively, § » significant at p=,05; # = significant at p«*.Ql •

220

Representation

2. Institutional Settings and Representational Roles Comparisons between legislatures of different countries, and of different states in the US context, indicate that system and district characteristics have an impact on role orientations, These studies have concentrated mainly on the style dimension of roles, that is, the trusteedelegate continuum. However, insights from this research can help to develop hypotheses with respect to the focus of representational roles as well. Jewell (1970: 465) has emphasized that roles of legislators are defined by norms originating in the political culture of a society and by the structure of a political system. But because cultural norms and political structures do not vary much among legislators from different US states where most of the comparative work on legislatures has been done, the impact of norms and structures has received little attention. However, we know that electoral systems and political cultures have a strong impact on policy representation, i.e. the congruence in political orientations of electors and elected. Elections are obviously the major linkage mechanism between electors and elected. But other factors such as political culture and structures of intermediaries and political opportunity affect their relationship as well (Huber and Powell 1994; Wessels 1996; Marsh and Wessels 1997), Thus the key to understanding differences in role orientations across countries might be found in the characteristics of political systems. In the following, we first turn to the regional dimension of roles and formulate and test hypotheses on the relationship between role orientations and electoral systems. We then go on to the group dimension and develop and test hypotheses on the relationship between characteristics of the systems of interest intermediation and role orientations. Thirdly, we will go beyond characteristics of the national political system and try to formulate and test hypotheses on the relationship between the position of a country in the European Union and the European focus of representation.

Competitiveness, Uncertainty, and the Regional Scope of Representation We first deal with the regional foci on constituency, nation, and Europe and the trade-off between these as indicated in Figure 10.1.

Whom to Represent?

221

Since no systematic cross-national research has been done so far, hypotheses must be generated from a literature that is mainly concerned with variations within a single nation-state. This literature discusses two general dimensions that might affect role-taking: political competition and uncertainty. With respect to political competition, Jewell presents the hypothesis that 'close competition in a district would force a legislator to ... pay close attention to constituency demands' (Jewell 1970: 471). Legislators in smaller districts are more visible and more accessible to their constituents, whereas members of larger delegations may be better able to escape the pressures of constituents (Jewell 1970: 480). A further impact of the electoral system is suggested by the finding that of those legislators elected in a whole county, three-fourths focus on the whole county, whereas of those elected in single-member districts, only onethird do so. With respect to district representation, the distribution is reversed: ten per cent of those elected in a county say mat they represent a district, but 60 per cent elected in a single-member district say so (Jewell 1970:482). Cox found that, for both electors and elites, district magnitude is the most important factor for strategic behaviour (Cox 1997: 228). This is due to the degree to which elections are personalized or parlicized. The basic difference between small and large districts normally refers to a personalized vs. a particized list vote which also coincides with a more majoritarian or more proportional version of an electoral system. In electoral districts in which only one or a very small number of representatives is elected, the individual candidate is more visible and more subject to competition than in large districts. This mechanism explains why Jewell found differences in representational roles according to district character. Thus, the basic hypothesis is that the more personalized an electoral system, i.e. the smaller the district magnitude, the narrower the representational focus representatives will choose. However, in systems where parties define success or failure of a candidature (because their list placement is decisive), the focus will be more broadly defined in regional terms. There, the reference point will be either the party voters or the nation as a whole. This leads directly to the following hypothesis:

222

Representation

the smaller the district magnitude, i.e. the more personalized the electoral competition between candidates, the more the focus of representation will be defined in terms of (he constituency. The average district magnitude is a rough estimate of how strong an electoral competition is personalized or particized. A second aspect that might affect the focus of representation relates to representatives' uncertainty about what to represent. Alpert (1979) has highlighted this dimension in a formal model of role choice as a subjective decision-making process. Like most researchers focusing on legislators' roles, he concentrates on the style of representation. He starts from the basic assumption that the representational role is not necessarily a reflection of any adopted style of representation, but instead reflects a broader concern for the reduction of uncertainty about district opinion (Alpert 1979; 588).

Uncertainty might arise from many factors: social differentiation in the district or complexity of social life, among others, or, maybe most important, just from the size of the constituency. This has been explored elsewhere as the 'number problem of political representation' (Wessels 1999). How can role orientations reduce uncertainty about whose interest to represent? Alpert (1979) discusses two mechanisms. One efficient way to reduce uncertainty about whom to represent from a large constituency is to refer to party labels, i.e. to focus on party voters. Another possibility is to generalize. In turn, a specific regional focus such as constituency or the district is not a very likely role orientation in large districts. Thus the hypothesis is: the larger the number of constituents, the more likely is a role orientation which reduces uncertainty either by a political definition of the representational focus or by generalization of the focus in regional terms. Both the competition and the uncertainty hypothesis will be tested separately for each representational focus. The competition hypothesis suggests a negative relationship between district magnitude and constituency focus as well as a positive correlation with a focus on larger regional entities. A similar pattern can be expected with respect to the size of the constituency, but for different reasons: large districts produce more uncertainty with respect to the question of whom to represent.

Whom to Represent?

223

Results are presented in Table 10,3. The table includes aggregate as well as individual-level correlations. This is mainly to ensure that aggregate correlations, for which significance often is very low due to the small number of cases, are not invalid. With respect to the competition hypothesis, results strongly support the expectation that a small district magnitude goes along with, constituency representation. This is true for representation not only at the national but also at the European level. The smaller the district magnitude, i.e. the more personalized the electoral competition, the more representatives choose to represent the constituency (see Figure 10.2). What these figures do not tell us is what focus legislators choose instead of constituency in case of large district magnitude. Only for the national level do we find some evidence that party voters become more important as a focus if district magnitude is large. The absence of a clear pattern has to do with the variety of multiple role choices between countries. Therefore, the relationship between district magnitude and the trade-off of foci is more conclusive. Individual correlations show that the largest trade-off is between party voter and constituency representation at the national and the European level (Table 10.3). With respect to the uncertainty hypothesis—which rather is a hypothesis on the reduction of uncertainty—results also support the expectation that the constituency is not chosen in electoral systems with large constituencies. This holds true for both the national and the European level. However, correlations with choice of foci do not confirm the expectation that insecurity due to large district representation will be reduced by either using party labels or generalizing the regional focus. Again, the trade-off figures offer more insights: the tradeoff between party voter and constituency focus shows the strongest correlation, the trade-off between the people and the constituency the second strongest. This indicates that representational uncertainty is solved primarily by using political labels and secondarily by generalizing the regional scope. Summing up, the results clearly show a strong co-variation between the distribution of role choices and characteristics of the electoral system. Of particular importance are characteristics that shape the linkage between voters and representatives. Even at the European level, where there are fewer variations between electoral systems than at the national level, district magnitude has a considerable impact on role choice.

TABLE10.3. Competition, uncertainty, and representational focus Competition District magnitude Ail (log) Agg. Ind,

Foci (% important} - people in Europe - people in nation - constituency - party voters - specific group Focus trade-off (a) - people in Europe vs. in nation - people vs. constituency - party voter vs. constituency - constituency vs. group - party voters vs. group

. .11 -.63$ .30 -.01

, ,01 -.24 $ .19$ -.03$

.65$ .63$ -.66$ .26 26

.20$ ,35$ -.18$ .18$

MEP (log) Agg. Ind. .33 ,05 .12 .06 -.55 § -.22$ .0? .09 .00 .03

.20 .56 § .50 § -.60* .06 15 1681

Uncertainty Size of constituency

MNP (log) Agg. Ind. .11 -.73 # .56 ,02

. -.01 .23$ .80 # .28$ .76 # -.19$ -.75 # .02 A3 314 11

-.02 -.24$ .20$

-.04 .17$ .35$ -.17$ .20$ 1367

AH (log) Agg. Ind. . .22 -.55 # .28 -.15 .62$ .56* -.50 # .33 26

.09$ -.20$ .14$ .04 .22$ .27$ -.19$ .09

MEP (log) Agg. Ind. .53 § -.02 -.52 § -.12 -.06 ,45 § .46 §

.38 -.54 § -.06

.11 .03 -.22$

.04 .02 .07 .20$ .23$

-.19$ .02 314

MNP (log) Agg. Ind. .12 -.81 # .45 -.07 .87$ .77 # -.78*

.05 § -.22$ .16$

,06

.21 $ .30$ -.22$ .09$ 1367

.41 11 Source: European Representation Study, Agg,=aggregate correlations; lnd.=individual-level correlations; §=significant at .05; ^significant at ,01; $=§ignifieaitt at .001. (a) Trade-offs arc defined as a priority choice between two possible foci of representation. For example: a representative might take the view that she or he represents party voters as well as voters in the constituency. In order to scale the priorities, answers to the first focus mentioned in Table 10.2 have been coded T when a respondent gave high priority (values 6 and 7 on a 7-point scale) to this focus, They have been coded '-I' when high priority {same definition) was given to the opposite focus mentioned second in Table 10.2. Lower values for importance have been coded *0'. The two values of choice of foci have then been added together. The result is a classification of respondents on a 3-point scale: *+!' where the first pole is I and the second 0, "0" where both poles have the value 0 or the one -1 and the other +1; *-F where the second pole was given priority and the first was not. -: not applicable.

N

1681

15

Whom to Represent?

225

aggregate correlation r =-,63, p =.001, n =25; individual-level correlation r —.24, p =,000, n =1681.

FlG 10,2, Electoral competition and constituency focus of representatives in Europe

Encompassiveness, Inclusiveness, and the Group Focus of Representation Turning to the group dimension of representational roles, other factors than the electoral system have to be taken into account in order to explain variations in their choice of focus between representatives of different nations, A focus on particular interest groups rather than on interest groups in general might be an indication of larger divisions in society because it implies representing one particular interest against others. Olson, for example, argues that political life is more divisive in political systems in which special-interest organizations play a major role (Olson 1982: 47), In turn, one might argue that representatives in more inclusive and consensual societies rarely adopt a special-interest group focus.

226

Representation

This general expectation can be further specified. If a society produces 'encompassing organizations' (Olson) rather than specialinterest organizations, these organizations will care less about rather particularistic interests. Since they represent the interest of larger parts of the society in a particular field, they are more concerned with the production of public goods than with the pursuit of very specific particularistic interests. Thus, the existence of encompassing interest groups in a society might affect representational role choice according to the following hypothesis: the focus of special group representation is less frequent in societies in which interest groups are encompassing, i.e. organizing large parts of the population. Another measure indicating the integration of interest groups in the production of public goods is the degree of corporatism in a society. Corporatism refers to the inclusion of interest groups in the bargaining and policy-making structures of the state. Such inclusion necessitates organizations which promote not only their particularistic interests but also common goals. In addition, corporatism goes along with the consensual type of democracy (Lijphart and Crepaz 1991: 235), i.e. the opposite of a divisive society. Both aspects, the participation of interest groups in the production of public goods and the degree of consensus indicated by corporatism, lead to the hypothesis that the more corporatistic interest intermediation in a country, the less emphasis political representatives will put on special group representation. However, this might only be part of the story. It has been argued that large interest groups with a monopoly on representation embedded in corporatist arrangements can co-exist with a blocking of interests other than those included in the established structures. This hypothesis has been put forward for example with respect to new politics interests (Brand 1982: 58; Kitschelt 1988: 212ff.). It has been used to explain the high mobilization in the new movement sector and the success of Green parties. Thus, encompassiveness and corporatism might lead to a particular kind of selectivity against interests that are not so well established. Following Olson (1982: 48), one can argue that, in addition to encompassiveness and corporatism, it is important how inclusive the organizational life of a society is, The less inclusive, the more likely it

Whom to Represent?

22?

might be that particularistic interests try to gain recognition and representation. Inclusiveness can be measured on two dimensions: how many are included and who is included (or excluded). The first measure of inclusiveness therefore is the overall membership density of a society. The second measure considers the degree of social inequality in participation in groups. These considerations lead to two additional hypotheses: the more inclusive an interest group system is, the less legislators will concentrate on special group representation; the less social inequality in participation in the intermediary system exists, the less emphasis will be put on special group representation. The first hypothesis brings up the problem of how to measure eneompassiveness of an organizational system. Since empirical information is hard to come by, we concentrate here on trade unions as a surrogate measure. Union density, or the share of the working population that belongs to a union, serves as a yardstick for the encompassiveness of one of the most central sectors of interests in European democracies. Another indicator of encompassiveness is the degree of corporatism as defined by Lijphart and Crepaz (1991). We find that the portion of legislators claiming special group representation as part of their role is greater the less encompassing an intermediary system is. This result holds irrespective of whether encompassiveness is measured in terms of union density or degree of corporatism. Correlations for the trade-off between foci of representation show that there is no significant trade-off between constituency and special group representation, though there is between party voter and group representation. But this pattern exists only with respect to members of national parliaments. For the members of the European Parliament no such evidence can be found (Table 10.4). This indicates that the choice of EP members to focus on special groups is rather independent of the national setting of interest groups. With respect to the inclusiveness hypothesis, the same pattern exists. It holds only with respect to members of national parliaments. And again, there is a clear relationship with the choice of the group focus and the trade-off between party voters and group representation, but not between constituency and group representation (Table 10.5).

TABLE 10.4. Encompassiveness and special group focus of representation All

Agg.

Ind.

Union density MEP Agg. Ind.

MNP Ind. Agg.

All

Agg.

Ind.

Corporatism MEP Ind. Agg.

MNP Agg. Ind.

Foci (% important) - specific group

-.21

-.14$

-.23

-.04

-.40

-.16$ -.08

-.14$

.20

.03

-.68

-.18$

Focus trade-off - constituency vs. group - party voters vs. group

.15 .39 §

.03 .22$

.27 .34

,09 .04

.20 .35

.02 .26$

-.03 .26$

-33 .26

-.05 .10

-.24 .87 §

-.04 .32$

N

26

1681

15

314

10

1367

-.30 .46 17

1342

11

247

6

1095

Source: European Representation Study. Agg.=aggregate correlations; Ind.=individual-level correlations; ^^significant at .05; #=significant at .01; {^significant at .001. Union density operational ized as membership share of working population, data are from the macro databank, research w\t Institutions and Social Change, WZB. Corporatism data are from Lijphart and Crcpaz, 1991: 239. They are factor scores; high values indicate high degrees of corporatism, low values low degrees.

TABLE 10,5. Inclusiveness of interest group systems and special group focus of representation Inclusiveness

All

MEP Ind. Agg.

Agg.

Ind.

Foci (% important) - specific group

-,42§

-.15$ -,44

Focus trade-off • constituency vs. group • party voters vs. group N

-.28 -.04 .64$ .18$ 22 1578

-.13 .66 # 13

MNP Agg. Ind.

All Ind. Agg.

-.09

-.54

-.17$

.27

-.02 • 12§ 299

-.45 .61 8

.09 -.05 -.21$ -.49 § 20 1279

Organizational inequality MEP MNP Ind. Agg. Ind. Agg.

.15$

.10

-.01 .08 -.21$ -.41 1532 12

.03

.56

.17$

-.00 -,14§

,11 -.66 §

.00 -.24$

253

8

1279

Source: European Representation Study. Agg.=aggregate correlations; lnd,=indtvidual-ievel correlations; §=significant at .05; ^significant at .01; $=signifieant at .001. Inclusiveness operationalized according to Wessels, 1997. Membership share in a society based on World Values Surveys 1990. Organizational inequality according to Wessels, 1997; 213-15. Inequality in membership rates between high- and low-income groups in a society with respect to traditional interest groups {for example trade unions, professional associations).

230

Representation

Taking the results on both sets of hypotheses together, we note that special group representation is more important for representatives in countries with a lower degree of organizational density and organizational integration of citizens. This however only applies to the national context of political representation. Country Size, Membership Experience, and the European Focus of Representation The prospects of political representation in the European Union depend, among other things, upon the conditions under which the members of the European Parliament extend the scope of their representational focus beyond the nation-state. As elaborated in the introduction, the choice of focus in this regard may be conceived of as an indication for the kind of Europe that representatives have in mind. We have seen above that roughly two-fifths of the members of the European Parliament consider the representation of all people in Europe as very important. However, there is considerable variation between country delegations. What accounts for this variation? Under which conditions do MEPs adopt a European focus of representation? It is obviously hard to formulate hypotheses in a virgin field of research. Two complementary strategies have been chosen. Although popular support for European integration is different from a European focus of political representation among representatives, we will try to utilize our knowledge from analyses of mass support for European integration. In addition, we will refer to the relative power of a country within the EU. Until recently, popular support for European integration was highest in the original six member-states (Niedermayer 1995a). This has led to the hypothesis of societal learning arising from exposure to the European institutions (Inglehart and Rabier 1978). To simplify the hypothesis: duration of membership breeds familiarity. The hypothesis arising from these considerations with respect to the choice of representational focus is thus: the longer one's country has been an EU member, the more MEPs adopt a focus of representation transcending the nation-state. Thus, the longer the membership, the more European the representational focus.

Whom to Represent?

231

A complementary hypothesis refers to the relative power of a memberstate within the EU. One might argue that 'weak' countries may be more concerned than strong countries with realizing national interests with the help of the EC, while at the same time protecting themselves against EU dominance. One possible indicator of a nation's strength is its population size. This figure also translates into political terms, for example with regard to the national share of EP seats. Thus, the expectation is that representatives of more powerful countries are more willing to adopt a European focus of representation than those of smaller ones. Testing these two simple and straightforward hypotheses yields the following results: the length of EC/EU membership indeed increases the chance that MEPs adopt a European focus of representation. This holds true for the endorsement of the European focus as well as for the trade-off between the European and the national focus of representation. Thus, familiarity indeed breeds 'Europeanness* (Table 10.6). TABLE 10.6. Country size, membership experience, and European focus of representation among MEPs Experience years of membership (tog) Agg, Ind.

Size population (log) Agg. Ind.

Size + experience population x years of membership (log) Agg. Ind.

Focus {% important) - all people in Europe .37

.20$

.65 #

.20 $

.66 #

.23 $

Focus trade-off - all people in Europe vs. in nation .49

.28$

.80$

.28$

.84$

.31$

314

15

314

15

314

N

15

Source: European Representation Study. Agg=aggregate correlations; Ind,=individuatlevel correlations; Insignificant at .05; ^significant at .01; Significant at .001.

The association between the population size of a country and the adoption of a European focus of representation is equally strong or, at the

232

Representation

aggregate level, even stronger. In particular, the correlation with the trade-off between the European and the national focus indicates that representatives of smaller countries tend to stick to the national focus of representation, whereas those of larger countries also take the European focus into account. However, the best predictor of a European focus of representation is a combination of both measures, i.e. the interaction between membership experience and country size. It accounts for 71 per cent of the variance in the mean value of role orientation trading-off Europe vs. nation at the aggregate level (Figure 10.3). This indicates that both hypotheses are indeed complementary.

familiarity=length of EC membership; relative power=population size; aggregate correlation: r = .84, p=,000, n=15; individual-level correlation; r = .31, p=.000, n= 314.

FIG 10.3. Impact of membership experience and relative power in the EU on European focus of representation

Whom to Represent?

233

It is obviously hard for members of the European Parliament from smaller countries, in particular if their countries are newer members of the Union, to conceive of political representation in terms of representing the people of Europe. They seem to feel safer with the traditional nation-state-bound conception of political representation. Maybe they even sense a need to stress the national interest just because they are representatives at the European level, facing the danger of being overruled by larger and more powerful countries with more experience in European policy-making.

SUMMARY

This chapter provides evidence that electoral mechanisms and national settings indeed have an influence on representatives* understanding of whom they should represent. The more personalized political competition, and the smaller the constituency units, the narrower the representative focus. The more representatives are the objects of voters' evaluation, the more they will concentrate on serving the constituency. This applies to members both of national parliaments and the European Parliament. Evidently, this finding has implications for political representation at the European level. With respect to group representation, results suggest that members of national parliaments are affected by their country's systems of interest intermediation while members of the European Parliament are not. The more encompassing and the more inclusive intermediary systems are, the less national representatives seem to feel a need to represent specific group interests. However, since these factors account only for variations between members of national parliaments, this finding has no direct implications for political representation at the European level MEPs from large countries with long membership experience more often share a European focus of representation than those who come from smaller countries with little membership experience. Representatives from smaller and newer members of the Union feel the need to represent their nation rather than the European people. The implications for political representation at the European level are obvious. Differences in national electoral systems and settings have quite an impact on the distribution of representational foci in a country.

234

Representation

As long as these national patterns persist, no homogeneous ideal of political representation at the European level can arise, Furthermore, the different incentives of electoral systems as well as differences between larger and smaller, and older and newer, member-states indicate that a perspective of representation which would fit a more or less unitary model of a federal Europe is still a long way off.

11 Wishful Thinking Among European Parliamentarians

SOREN HOLMBERG

INTRODUCTION

IN the 1940s, the psychologist R, M. W, Travers discovered—or maybe rediscovered, since most important human thoughts have been thought before—-what must be considered one of the fundamental laws of human relationships: that when people perceive the attitudes and opinions of others, they have a strong tendency to believe that others hold the same views as themselves. People have a tendency to project their own points of view onto others. The explanation for this tendency can be found in cognitive balance theory (Heider 1946 and 1958), which states that people strive for consistency in their cognitions of reality, or in the related theory of social judgement (Sherif and Hovland 1961; Eiser 1992), which notes that people tend to assimilate (pull towards them) stimulus they like and contrast (push further away) things they dislike. Expressed with less social science jargon, this means that people prefer to be in accord with what they like and in discord with what they do not like. People tend to believe their friends think as they do and that their enemies think differently. Applied to politics, this tendency to wishful thinking—to see what one wants to see—has profound implications. It works as a cohesive cement in a political system, promoting stability and harmony. Among

236

Representation

voters, where wishful thinking means that voters tend to think their own party or candidate has the same opinion they do, impulses to change voting habits, and maybe also inclinations to examine issues more closely, are weakened (Berelson, Lazarsfeld, and McPhee 1954; van der Brag 1996). The result might be stability instead of rational mobility. Among representatives, projection leads to an underestimation of the true differences in opinions vis-a-vis voters, producing less incentive for altering behaviour in parliament or trying to change the opinions of the public (Holmberg 1989). Once more the result might be inactivity and no change instead of more public debate and opinionformation from above or roll-call voting in accordance with what people want. To the degree that stability and harmonious relations are valuable to a political system, the tendency to wishful thinking is a positive phenomenon. It helps parties stay together and it presumably increases the level of trust between voters and representatives. An obvious problem arises if people's tendency to project becomes so strong that everybody only sees himself and no one sees reality. In that case rational politics and democracy are in serious trouble. Meaningful public discourse will not function. Politicians and voters will live in their own subjective worlds with only dim notions of each other's realities. Such a fool's paradise is an extreme case, but it highlights a dilemma: some wishful thinking might be good for stability, too much wishful thinking is clearly detrimental to the workings of all political systems, especially democratic systems. It would mean that even if politicians have a role conception of being responsive to the political views of their voters (however 'the voters' might be conceived; see Chapter 10), this would not be very helpful because they have a wrong idea of what these views are.

PREVIOUS RESEARCH

Previous studies on how elected representatives perceive the opinions of their voters have proved that the projection tendency is pronounced, especially for issues with lower degrees of saliency (Clausen 1977; Granberg 1993). The degree of ambiguity of the stimulus object plays a role. It is easier to see what you want to see if what you look at is a

Wishful Thinking

23?

bit fuzzy and far away. Hence, salient issues closer to home tend to produce less biased perceptions. For example, perceptual accuracy among members of parliament tends to be better on left-right issues than on other less politicized issues (Converse and Pierce 1986; Esaiasson and Holmberg 1996). Changing perspective from the stimulus object to the observer, political research on perceptual accuracy among MPs has indicated that such individual factors as educational level, political experience, contact network, media exposure, and power position are only weakly related to the members' knowledge of voter sentiments (Converse and Pierce 1986; Esaiasson and Holmberg 1996). Apparently, rational knowledge formation processes are not working very effectively in this area. The main reason for this failing of normal information processes is the strong impact of another observer-related factor: the tendency among elected members to project their own opinions onto their voters. Among members with the same opinion as their voters, wishful thinking of this kind leads to artificially high levels of perceptual accuracy. On the other hand, among members with opinions at odds with their voters, projecting their own point of view results in mistaken perceptions. Elite perceptions of the political attitudes of the general public as a research area does not have a particularly long history. The first, and still the best-known study, is Miller's and Stokes' American Representation Study from 1958. Congressmen's perceptions of district opinions are one of the four basic variables in their diamond model of representation (Miller and Stokes 1963; Thomassen 1994). Furthermore, few studies of elite perception have been published, which might be surprising given the importance of elite knowledge of voter sentiments, but may be understandable given the difficulties and costs involved in collecting relevant data from among elected officials as well as the public.

THE DATA

Hence, the data on elite perceptions of voter attitudes collected in the context of the European Representation Study are unique. As we have seen earlier, we have information on the opinions and perceptions of

238

Representation

the members of the European Parliament (MEPs) and of the members often national parliaments (MNPs) as well as information on what the general public in member-states thinks about four issue dimensions. Three of them are related to the European Union; for or against a common European currency, for or against a massive employment policy within the EU, and for or against removing national borders; the fourth issue is the left-right dimension. Opinions among members as well as voters were measured on 10-point rating scales. The perception question to the members asked them to place their own party voters* position on the four dimensions.

THE CASE OF A COMMON EUROPEAN CURRENCY

The data open up the possibility to study a whole range of interesting topics related to political representation. In addition to the difference between elite and mass attitudes on these four issue dimensions (which has been dealt with in Chapter 9), the extent to which elites are knowledgeable about those mass attitudes is of course of great interest if we study representational processes in the European Union. Figures II.I and 11.2 present a first impression of the relevant data for one of the issue dimensions, the common European currency. From Chapter 9 we already know that members (MEPs), on average, tended to be more positive about a common European currency than the citizenry. Here we can add to this the equally striking finding that MEPs, on average, were aware that citizens tended to be less enthusiastic about the EMU project than they were.1 Thus, elite perceptions were correct at least in the sense that MEPs knew that citizens, compared to themselves, tended to be more, not less, critical of the idea of a common European currency. However, if we want to study perceptions in more detail, the kind of analysis represented in Figure 11.1 is not focused enough. Results aggregated to the national level have an obvious interest, but since the perception questions put to the members dealt with attitudes among the members' own party voters, the analysis should ideally be performed broken down by party.

Comments: attitudes of MEPs and perceptions of MEPs of voters' attitudes were measured on a 10-point scale where 1 was labelled "independent national currency* and 10 was labelled 'new common European currency'. The means in the figure have been multiplied by 10 in order to get a scale running from 10 (national currency) to 100 (common EU currency). FIG 1 1 . 1 . A common European currency? Positions of MEPs compared to their perceptions of the positions of their party's voters (by nation)

Comments: attitudes of MEPs and of citizens were measured on a iO-point scale where 1 was labelled 'independent national currency' and 10 'new common European currency'. The means in the figure have been multiplied by 10 in order to get a scale running from 10 (national currency) to 100 (common EU currency).

FlO 11.2. A common European currency? Attitudes of MEPs compared to those of the mass public (by nation)

Wishful Thinking

241

THREE HYPOTHESES Focusing on perceptions, the data make it possible to test a whole series of hypotheses derived from perceptual theory and previous research on elite perception of mass attitudes. I will concentrate on analysing three hypotheses, one of a more general and two of a more specific character. The more general hypothesis deals with the core of perceptual theory, i.e. that a strong tendency indeed exists among elected representatives to project their own opinions onto their voters. This means that we expect to find more cases of assimilation than contrast when members perceive the positions of their party's voters, and we expect members in attitude agreement with their voters to exhibit a higher level of perceptual accuracy than members at odds with their voters. Furthermore, when testing the strength of the projection tendency, we expect members' opinions to have a strong effect on their perceptions, maybe an even stronger effect than reality (= voters' opinions). The two more specific hypotheses deal with the propositions that fuzzy objects are more susceptible to perceptual distortion than less fuzzy objects and that more distant observers are more prone to faulty perceptions than less distant observers. In terms of our research problem, this means that we expect members' perceptions to be more accurate on the old and highly politicized left-right super issue than on the new and less salient European issues. And we expect members of the European Parliament (MEPs)—being in many ways more distant to their voters back home than the national MPs—to be less accurate in their voter perceptions than the members of the national parliaments (MNPs). Our hypotheses can be tested in a number of different ways.2 In this chapter, I have chosen the political parties as the unit of analysis and shall test the hypotheses on the party level. After all, the interview question put to the members asked what the members thought about the opinions of their own party's voters. Despite the very limited number of interviewed members for many parties, the number of parties included in my analysis is fairly decent. With a limit of at least five interviewed members among MEPs, we get 19 parties represented. Among MNPs with a limit of ten or more interviewed members, we can include 34 parties.3

242

Representation THREE MODELS OF PERCEPTUAL ACCURACY

A lucid and easy-to-grasp way of arranging the data is to compute the party means for our three relevant variables on all four issue dimensions. The three relevant variables are the members* own positions on the issues (MPatt)» the members' perceptions of their party voters' position (MPp,,rc), and the voters' true position (VOT8tt). The means for VOTatt have been calculated based on information on how the respondents voted in the 1994 European Parliament elections (for the MEP analysis) and on information on how the respondents voted in the last national election (for the MNP analysis). In practice, however, the definition of party voters is not essential to our results. No matter how we define party voters—with the help of vote choice in the 1994 EU elections, present party sympathy, vote choice in the last national election, or vote intention in the next national election—the results are very similar. A set of three means without ties can be ordered in six different ways. Disregarding the mirror images we end up with three different and theoretically important ways of ordering the phenomena MPatt, MPpere, and VOTat,. We can talk about three models of perceptual accuracy. One of the models could be called the assimilation ordering model since it is characterized by members perceiving the position of their voters closer to themselves than is actually true. The results for the German Green Party in Figure 11.3 illustrate such an assimilation ordering on the issue of a common European currency. Green MNPs are on the average more positive towards a common currency (mean 9.3) than their voters (mean 5.1) and they believe on average that the voters' opinions are closer to them (mean 7.4) than they really are. The order of the means reveals an assimilation ordering, MPm, MPpere, VOT.tt. Another possible model might be called a contrast ordering model, characterized by members believing their voters* opinions to be further away than they are in reality. The contrast ordering could be illustrated by the results for the British Conservatives on the issue of border controls. On average, Tory MEPs tend to be more positive towards removing national borders (mean 4.6) than their voters (mean 6.6). However, when asked about their perceptions of what position the voters have, Tory MEPs on average exaggerate the distance to their voters.

Wishful Thinking

243

Assimilation ordering: the German Greens (MNPs)

Contrast ordering: the British Conservatives (MEPs)

Wrong direction ordering: the Italian Forza Italia (MEPs)

A = tree opinion distance between MPs and voters; B = MPs' perceived distance to voters; C = MPs' mispereeived distance to voters.

FIG 11.3. Three orderings of members' attitudes, members* perceptions of voter attitudes, and voters' attitudes on an issue dimension (means)

244

Representation

The third possible model might be called the wrong direction ordering model. Members look in the wrong direction when trying to locate where their voters are situated on the issue dimensions. For example, members might think their voters are to their left, while in reality they are to the right. In comparison, the assimilation and contrast orderings are both characterized by members looking in the right direction when perceiving the position of the voters. In political terms, looking the wrong way, creating a wrong direction ordering, could be considered a more serious mistake with potentially graver consequences than the other two possible mistakes of assimilation and contrast. In the former case, members will change their position in the wrong direction if they want to accommodate the will of their voters, while in the case of assimilation and contrast mistakes members will at least change their position in the right direction. The results for the party Forza Italia demonstrate a case of a wrong direction ordering. On average, Forza MEPs turn out to be more in favour of an employment programme within the EU (mean 3.9) than their voters (mean 5.5). However, when Forza Italia MEPs perceive the position of their voters they think the voters tend to be more positive toward an employment programme than they are themselves, while in reality their voters on average were more negative (mean 2.5). When the empirical occurrences of the three ways of ordering the means of members* attitudes, members' perceptions, and voters' attitudes are calculated among the MEP and MNP parties across our four issue dimensions, an unequivocal result emerges that strongly supports our general projection hypothesis (see Figure 11.3). As it rams out, a clear majority of the orderings are of the assimilation kind, 77 per cent among MNP parties and 72 per cent among MEP parties. Contrast orderings and wrong direction orderings are much rarer. We found only about 10—15 per cent of each kind in our data among MEP as well as MNP parties. Of course, the number of test cases is limited, but not in any absolute sense too small. After all, the data encompass 75 cases among the MEP parties (19 party means measured on four issue dimensions) and 127 cases among the MNP parties (34 parties, four issue dimensions).4

Wishful Thinking

245

N A T I O N A L VERSUS EUROPEAN PARLIAMENTARIANS Another way of testing our general projection hypothesis as well as our two more specific hypotheses (members of national parliaments are more aware of the positions of their voters than members of the European Parliament, and perceptions of voter positions on the left-right super issue are more accurate than perceptions of voter positions on the EU issues) is illustrated in Table 11.1, A set of average distances has been calculated across the investigated parties: first, the true distance between members' own attitudes on the issues and the positions of their voters on the same issues (AiMPjtt-VOTgtt); second, the members' perceived distance to their own voters (fiMP^-MPpj.^ and third, the distance between where members think the voters are and where they really are (C:MPpereVOTatt). The last distance (C) is a useful measure of perceptual accuracy. If all members correctly perceive the issue positions of their voters, the C distance of misperception would be zero. In actuality it is on average slightly more than half the length of the members' true issue distance to the voters, and that cannot be interpreted as anything but inaccuracy. From the analysis in Figure 11.2 we know that the main cause behind this strong tendency to rnisperceive voter positions is wishful thinking. In almost 80 per cent of our cases members projected their own attitudes onto their voters, thereby creating assimilated perceptions and low levels of perceptual accuracy. The results in Table 11.1 also give strong support to our two specific hypotheses. First, overall and on all four issues, MNPs tend to exhibit a higher degree of perceptual accuracy than MEPs. Thus, our results support the conclusion that European-level politicians tend to be less knowledgeable about their voters than national politicians. Second, the old and highly politicized left—right super issue produced more accurate perceptions among both MNPs and MEPs than the more recently emerging EU issues of border controls and a common currency. On average, the misperceived distances to the voters tend to be about twice as long for the EU issues as for the left-right super issue.

246

Representation

TABLE 11.1. Distance to voters: true, perceived, and misperceived (differences between party means) A: true B: perceived C: misperceived distance to voters distance to voters distance to voters Issue area

MEP

MTMP

MEP

MNP

MEP

MNP

Currency Employment Borders Left-right

3.1 1.7 2,5 1.2

2.7 1.5 1.9 0,9

1.8 0.9 1.3 0.6

1.6 0.5 1.2 0.5

1.7 1.6 1.3 0.8

1.3 1.3 1.0 0.5

4-issue average

2.1

1.8

1.2

1.0

1.4

1.0

Source: European Representation Study. The distances are defined in Figure 11.2. A (true distance) equals the average difference between MPan and Vot,,, B (perceived distance) is defined as the average difference between MPm and MPp,K, while C (misperceived distance) is the average difference between MPperc and Votm. The distances can vary between 0.0 and 9.0. Parties (19 among MEPs and 34 among MNPs) are the units of analysis.

THE EFFECT OF ISSUE A G R E E M E N T

One of the main reasons why MNPs outperformed MEPs when it came to voter knowledge is disclosed in the two first columns of Table 1 1 . 1 . There one can see that the MNPs on average, and on all four issues, had a shorter true distance to their voters than the MEPs. This means that the degree of issue agreement between members and voters was higher among MNPs than among MEPs. This is a crucial bit of information since we know that members' own attitudes and opinion agreement with their voters play an important role when it comes to perceptions and degrees of perceptual accuracy. The results in Table 11.2 clearly demonstrate the importance of issue agreement in this circumstance. Almost without exception, members of parties in opinion agreement with their voters tend to perceive the positions of their voters accurately—100 per cent among MEP parties and 97 per cent among MNP parties.3

Wishful Thinking

247

TABLE 11.2. Issue agreement and perceptual accuracy (per cent) Issue agreement between MPs and voters

MPs* perception of voters' opinions Sum Number of Accurate Wrong per cent eases

MEPs

Yes No

100 38

0 62

100 100

54 21

Yes No

97 41

3 59

100 100

90 39

MNPs

Source: European Representation Study. Parties are the units of analysis, four issue dimensions are included. Issue agreement is coded 'yes' for a party if mean MPm and mean VoUi both are less or more than 5.5 (the mean position on the dimensions). If the MPm and Vot,n means are located on different sides of the 5.5 mean, a "no* is coded. MPs* perceptions of voters' opinions are classified as accurate if MPpcK and Votal, are located on the same side of the overall mean 5.5. If they are located on opposite sides of the 5.5 mean the perception is coded as wrong.

The results are quite different among members of parties in opinion disagreement with their own voters. Here we find a majority misperceiving the positions of their voters - 62 per cent among MEPs and 59 per cent among MNPs. The similarity between the results among MNPs and MEPs is striking. It increases our credence in the universality of the process of wishful thinking. On average, perceptual accuracy is somewhat better among MNPs than among MEPs, the main reason being that the level of issue agreement between members and voters tends to be higher among MNPs than among MEPs. However, among MNP and MEP parties in opinion disagreement with their own voters, the tendency to wishful thinking is equally strong. In both cases we find a majority of parties with members locating their voters' opinions incorrectly.

A PURE W I S H F U L - T H I N K I N G MODEL

The result could be formulated more drastically, since it is a drastic result. When members with different views than those held by the voters they represent judge an important piece of reality, i.e. voter opinion.

248

Representation

their own attitudes play a greater role than reality does. Wishful thinking has a stronger impact on parliamentarians' perceptions than does actual voter opinion. The statement may seem a bit far-fetched, yet it has found support in previous research on members of the Swedish Riksdag (Esaiasson and Holmberg 1996)7, and it is supported by our results among MNP and MEP parties, as can be seen in Table 11.3. TABLE 11.3. MPs' perceptions of voter opinion: two models tested among MEPs and MNPs (entries are per cent members with pro-perceptions) Voters' opinions Pro Con

Effect of voters' opinions

Pure knowledge model MPs' opinion Pro Con Effect of MPs' opinions

100 100 0

0 0 0

+100 +100

Pure wishful-thinking model MPs' opinion Pro Con Effect of MPs* opinions

100 0 + 100

100 0

0 0

+ 100

Empirical test among MEPs MEPs* opinion Pro Con Effect of MEPs* opinions

100 38 +62

62 0 +62

+38 +38

Empirical test among MNPs MNPs' opinion Pro Con Effect of MNPs' opinions

94 57 +37

78 2 +76

+ 16 +55

Source: European Representation Study. The results arc based on 129 cases (34 parties, four issue dimensions) for Ihe MNPs and on 75 cases (19 parties, four issue dimensions) for the MEPs. "Pro' and 'Con' are defined as below or above the overall 5.5 mean on the four issue dimensions.

Wishful Thinking

249

The results show the outcome of an effect analysis with the members' attitudes and the true opinions of their voters as the independent variables, and the members' perceptions of the opinions of their voters as the dependent variable, measured as the average per cent of perceptions of pro-opinions. Note that perceptual accuracy is not the dependent variable. Perceptions of pro-opinions among voters is the dependent variable. The resolt of the effect analysis will indicate what is most significant when members try to ascertain what voters think—projection or reality. The outcome of the analysis is a bit disconcerting. Both explanatory factors have an independent effect, but MPs' attitudes mean more than voter reality. The results fit a pure wishful-thinking model better (in which members' opinions govern their perceptions) than a pure knowledge model (in which reality governs their perceptions). Wishful thinking overshadows voter reality.

A DISCONCERTING PARADOX

The answer to the question of what explains elected representatives' knowledge of voter opinion should be quite obvious by now. Our general projection hypothesis has received strong support. Members' own opinions, together with their degree of opinion agreement with their voters, are without doubt the most important explanatory factors. The motor driving the process is the strong urge members have to see what they want to see: that they enjoy the support of their voters. The unpleasant truth is that a largely irrational tendency toward wishful thinking is more significant for members' knowledge of voter opinion than other more rational processes of knowledge acquisition. Some comfort can be found in the fact that the parties that made our analysis possible, that is, those with opinions differing from those of their voters, constituted a minority among MNPs (30 per cent) as well as MEPs (28 per cent). Most parties exhibited the same majority opinion among members and voters. When members of these parties ascribed their opinions to their voters they were most often correct, for they hold the same opinion as their voters. Consequently, the problem in terms of perceptual accuracy rests with members and parties with opinions at odds with their own voters. It is only among them that a

250

Representation

majority perceives voter opinions wrongly. And that is somewhat paradoxical. The information channels between members and voters function badly when they should work best, that is, when members and voters disagree, and they function well when they are less needed, i.e. when members and voters agree anyway. And the villain in the piece is the strong tendency towards wishful thinking among parliamentarians—on the national level as well as in the European Parliament. If Travers was right that wishful thinking is a fundamental human law, the villain will be tough to beat. The only possible alleviation would be more communication and less distance between elected representatives and voters. This is best achieved by politicizing important issues.

NOTES

1

2

3

Since voter attitudes were measured in 1994 and MEP perceptions of voter attitudes two years later in 1996, opinion changes could have occurred in the meantime among citizens, making them more positive towards the EMU project. However, according to the results of the Eurobarometer (EB 47) there seem to have been no major opinion shifts on the common currency issue among EU citizens between 1994 and 1996. If we have to discern a trend, it is in the opposite direction, making people somewhat more negative, A possible analysis would be to test the hypotheses on the individual member level using a correlational approach, although correlations are not a perfect instrument in this circumstance (Saris 1988). However, disregarding this, the individual-level correlations between the members* own opinions and their perceptions of the positions of their party's voters are impressive indeed. Among all MNPs the unweighted correlations (r) are .69 for the common currency issue, .85 for the employment issue, .74 for the border issue, and ,82 for the left-right issue. Controlling for party, the average correlations decrease somewhat, but not very much. The average with-in party correlations were .54 for the common currency issue, .75 for the employment issue, .68 for the border issue, and .56 for the left-right issue. The following parties have been included in the MNP analysis; from Germany, CDU/CSU, Greens (Die Gninen), SPD, FDP, PDS; from Greece, ND, PASOK; from Ireland, FF, FG; from The Netherlands, CDA, D66, PvDA; from Portugal, PS, PDS; from Belgium, CVP, PRL, PS, PSC, SP, VLB; from Italy, AN, Forza Italia, PDS, PPI; from Luxembourg, CSV;

Wishful Thinking

4

5

6

7

251

from Spain, PP, PSOE; from Sweden, CP, FP, KD, M, MP, SP, VP. The following parties have been included in the MEP analysis: from France, FN, PS; from Germany, CDU/CSU, Greens (Die Griinen), SPD; from Greece, ND; from Ireland, FF; from Italy, Forza Italia, PDS, PPI; from The Netherlands, CDA, VVD; from Portugal, PS, PDS; from Spain, PP, PSOE; from Sweden, SP; from Great Britain, Cons, Lab. The employment question was not included in the Swedish voter and MNP studies, which explains one missing case in the MEP analysis and seven missing cases in the MNP analysis. Overall, the proportion of parties where members accurately perceived the position of their voters was 82 per cent among EP parties and 80 per cent among NP parties. Note that accurate in this instance means that both the perceived and the true location of the voters* opinions are on the same side of the 5.5 mean on the issue dimensions. Operationally, issue agreement has been coded 'Yes' for a party if the MPstt and VOTslt means are on the same side of the overall mean of the issue dimension ( 5.5). If the MPatt and VOTatt means are situated on different sides of the 5.5 mean, a 'No' was coded. MPs perceptions of voters* opinions were classified as accurate if MPpwc and VOTalt were located on the same side of the overall mean of 5.5. If they were situated on opposite sides of the 5.5 mean, the perceptions were coded as wrong. The Swedish studies were performed in 1968/69, 1985. and 1988 covering 20, 9, and 12 issues, respectively. In all three studies, the results were very similar. Among members with the same opinion as their party's voters, about 90 per cent perceived the majority positions of their voters accurately, while among members with opinions at odds with their voters, only about 30 per cent perceived their voters' majority positions correctly (and about 70 per cent perceived them wrongly). The analyses were performed using individual members as units of analysis.

This page intentionally left blank

CONCLUSION

This page intentionally left blank

12

In Conclusion: Political Representation and Legitimacy in the European Union

JACQUES THOMASSEN AND HERMANN SCHMITT

THE purpose of this book was to expand our knowledge of political representation in the European Union and the legitimacy of its political order. In this last chapter we will summarize what we have learned on these two subjects and what it tells us about the European Union as a developing democratic political system. The hypothesis that a wellfunctioning system of political representation is a precondition for a legitimate democratic political system will serve as the leading theme. Therefore, we will first try to evaluate the quality of the system of political representation of the EU, and from there continue with what we have learned about the legitimacy of the Union.

POLITICAL REPRESENTATION IN THE EUROPEAN UNION

One needs criteria to evaluate the quality of the system of political representation in the European Union. Such criteria can only be deduced from a normative view, not only on political representation, but on the European Union as well. In order to put the results of this study in the proper perspective, we should first summarize the view on political representation in the European Union we took as a starting point.

256

InConclusion

In Chapter 1, we argued that political control and political representation should take place at the same level where decisions are taken, The European Union is best conceived as a multi-tiered system of government where some decisions are taken at the sub-national level, some at the national level, and some at the European level. In the latter case, decisions are taken both in an intergovernmental and a supranational mode. EU decisions under the intergovernmental regime are not taken on behalf of the European people, but on behalf of the peoples of Europe. The peoples of Europe are represented by (ministers of) their governments, who—at least in principle—are subject to the political scrutiny and control of the national parliament representing the people in a particular member-state. The proper level of political control and political representation is the national level, and a European Parliament, in any traditional meaning of the term, is an anomaly. However, once EU decisions are taken under its supranational regime—i.e. when individual member-states have no veto power—indirect political control and representation via the national parliaments cannot be effective anymore. If decisions are taken at the supranational level, both political control and political representation must happen at that level as well. It is in this respect that the European Parliament has a unique function that cannot be fulfilled by national parliaments. It is a basic assumption of any modern concept of political representation that the interest of the one nation should prevail over local or regional interests. Translated to the European Union this requires that European interests should prevail over national interests.1 If the process of political representation in the European Union is no more than a mechanism to put national interests on the European agenda, a truly European system of political representation does not exist. An important criterion to assess the quality of representative democracy in the European Union is therefore the extent to which supranational or at least cross-national political views and interests take precedence over purely national interests. A first indication of the extent to which a truly European system of political representation exists (or is at least developing) can be found in the role conceptions of MEPs. In a truly European system of political representation we would expect them to give precedence to a common European interest rather than to national or other partial interests. Chapter 10 established that this focus of representation is certainly not

In Conclusion

257

the most conspicuous trademark of most MEPs. Only a minority considers it important to represent 'all people in Europe', The MEPs from all countries but one think it more important to represent the people in their country. For most MEPs, representing the national interest is more important than representing the general European interest. However, these role conceptions are not evenly distributed among the representatives of the different member-states. Representatives from the newer member-countries, and those from smaller and less powerful member-countries, are inclined to see themselves representing national interests rather than a European interest. This seems to suggest that a European focus of representation among MEPs takes time to develop. However, what is perhaps more important is the extent to which a truly European party system is developing. Not individual members of parliament but political parties are the key actors on the elite side of the process of political representation in Europe, and parly government has become the dominant doctrine of democratic government. According to this doctrine, cohesive political parties with different policy platforms compete for the support of the voters. As long as voters vote for the party that is closest to their own policy preferences, they will elect a parliament that more or less accurately reflects their policy preferences. An additional requirement of party government is that the government is formed and controlled by a majority in parliament. It is hard to think of a democratic political system that would not meet these basic principles. Therefore, an important test of the development of a truly European system of political representation is the extent to which a European party system is developing that links the policy views of the European people to those expressed in the European Parliament. As an analytical tool for this test we introduced the responsible party model. This model enumerates a number of conditions—distinct and cohesive political parties and knowledgeable and rational voters—that need to be met in order to connect the will of the people to public policy.2 The extent to which the requirements of the responsible party model are met at the European level is summarized in Chapter 9. The essential finding is that a truly European system of political representation involving transnational political parties competing for the votes of a single European electorate might be more feasible than is often suggested—although such a system does not now exist. We have shown

258

In Conclusion

that the basic left-right ideology shared by most European countries is, at the European level, aggregated into party groups that are no less distinct and cohesive than their constituent national parties. Equally, as Chapters 7 and 8 have demonstrated, voters throughout Europe seem to be aware of the main differences between parties and to consider them when making up their minds at election time.

A CRISIS OF POLITICAL REPRESENTATION?

Despite all its formal deficiencies, the European system of political representation is not in such bad shape as long as we apply it to the basic left-right dimension capable of reflecting the cleavage structure of most European countries. However, none of the above is true once we come to issues related to the European project as such. On these issues the major party groups are hardly distinct, differences of opinion between members of these groups are more related to their national background than to their party affiliation, voters have a problem recognizing what the views of the party groups are, and hardly take these issues into account when deciding what party to vote for. Therefore, it is anything but surprising that we found little congruence between voters and representatives on these issues. Of course, these findings are grist to the mill of those who perceive a failure of the European system of political representation. They tend to argue for reshuffling the European party system into a system that would allow the voters to express their opinion on the future development of the European Union, This would be possible if different parties represented different positions on the simple continuum of anti-Europe—status quo—further integration (Bogdanor 1989; 209; van der Eijk and Franklin 1996: 377; Andeweg 1995: 69). The development of such a party system is most unlikely (Hix 1998). Moreover, the validity of the argument is disputable. The basic problem is that there are at least three different kinds of European issues, none of which lends itself easily as a basis for reshuffling the European party system (Thomassen 1998), The first kind of European issues refers to the delicate balance of national sovereignty and European integration. According to many observers, this is the dimension on which a truly European party system

In Conclusion

259

should be based. However, if political representation should take place at the same level where decisions are taken, it makes little sense to base a system of representation, and therefore a party system, on issues that are decided at a different level. Formal decisions on a further transfer of sovereignty from the national to the European level are subject to the intergovernmental regime of European decision-making. They need the consent of national governments and are, at least in principle, under the control of national parliaments and national electorates. Therefore, the interesting paradox is that what usually are called European issues are basically national issues. As far as the existing party system fails to offer a meaningful choice to the voters, this is a problem at the national rather than the European level/ A second kind of European issues refers to the division of power between different institutions of the European Union, given a certain competence of the Union. The basic issue for the European Parliament, and therefore a possible issue in European Parliament elections, is the accountability of the Commission and the Council of Ministers to the European Parliament. However, this is hardly an issue that lends itself to politicizing within the European Parliament. Efforts to strengthen the European Parliament are unproblematic within the Parliament, but produce conflicts between the Parliament as an institution and other institutions of the European Union. On this issue differences between party groups will be secondary to the institutional interests of the Parliament as such. And an issue on which political parties do not compete and do not want to can hardly be a basis for the formation of a party system. A third kind of European issues refers to substantive policy issues decided at the European level. Once the Union has the competence to pursue its own policies in a particular area, the question remains what the content of that policy should be. Suddenly, the relevant conflict dimensions do not necessarily differ from those at the national level. In a healthy and stable democracy, the political debate and political conflicts will usually refer to substantive policy issues within the constraints of a constitutional order (Schmitt 1994; Thomassen and Schmitt 1998). The constitutional order itself should not be a matter of permanent dispute. If the unification of Europe is at all viable, major political disputes must gradually shift from constitutional to substantive policy issues. In that process the European issue dimension might gradually be absorbed by the left-right dimension as other conflict di-

260

In Conclusion

mensions have been in the past (Fuchs and Klingemann 1992; Marks and Wilson 1998), The more the debate concentrates on European policy rather than on European integration in abstracto, the more the two dimensions will be intertwined.4 From this perspective, the relative consensus among political elites about the future of Europe is not a problem. It is rather a condition for further development of the European Union as a democratic political system. We see no reason why, at least in the long run, the basic structure of the party system at the European level should differ from that at the national level. Quite the contrary: one of the main reasons to consider the development of a European system of political representation feasible at all is the existence of common roots in the party systems of the member-states. Therefore, it is our contention that many of the arguments on which the notion of a crisis of political representation in the European Union is based are disputable. That an effective system of political representation in Europe is feasible does not imply that such a system already exists. We will discuss the democratic quality of the EU political system, and strategies for its reform, at the end of this conclusion. But first we will turn to our findings on the legitimacy of the European Union.

LEGITIMACY Throughout this book we have defined legitimacy in terms of the legitimacy beliefs of the people. To what extent the incomplete system of political representation at the European level translates into legitimacy beliefs about the European Union is the question to which we now turn. In Part II we analysed three aspects of the legitimacy of the European Union: the political community (Chapter 2), the division of labour between the several levels of government (Chapter 3), and European institutions (Chapter 4) in relation to support for EU policies (Chapter 5). One of the most basic questions with respect to the legitimacy of the European Union is whether people are willing to allocate governmental responsibilities to the European Union (Chapter 3). The answer to this question is very clear. People think that the policy responsibilities of the European Union should be extended, compared to what they per-

In Conclusion

261

ceive to be the present situation. These preferences seem to be well considered. They are highest in those Southern European countries that have profited most from their membership in the Union and for those policy areas in which, objectively, national policies can hardly be effective because of the intrinsically international nature of the problem, These findings seem to confirm an instrumental orientation to European politics: whenever a policy responsibility at the European level is thought to be more effective, people will support it; if not, they will prefer responsibility at the national or regional level of government. If people's attitudes towards the Union are mainly instrumental, we can expect their feelings towards the institutions of the Union to be strongly influenced by their evaluation of the Union's performance. This expectation was borne out. We found a spillover effect of people's perceptions of the economic performance of the Union on their evaluations of European institutions and the functioning of democracy in the Union. Variations in appreciation seem to depend on what people have to lose. In the prosperous and established liberal democracies of Northern Europe, people are more critical about the institutions of the European Union than in Southern European countries with less prosperous economies and a shorter democratic tradition. The most important variable explaining variations in the assessments of EU democracy was its economic performance, confirming the interpretation that support for the European political system is mainly of an instrumental nature. However, there is a general understanding that the effectiveness of a political system is a very vulnerable and therefore insufficient basis for a stable political system (Easton 1965; Lipset 1960). An important aspect of a more diffuse and durable form of legitimacy is the development of a sense of community among the people. A minimal sense of community, or at least the acceptance of fellow Europeans as members of a commonwealth of mutual interests, might be considered a basic condition for the development of a permissive consensus about European integration, Successful EU policies in turn might be expected to strengthen feelings of identification with Europe. In Chapter 2 we examined to what extent such feelings have developed among the people of the European Union. The results indicate that a majority of EU citizens identify wife the new political community. A map of European identifications in the mid-1990s reveals a centre-periphery picture, marking highest levels of identifications for

262

In Conclusion

the six original member-countries. Elites generally identify more strongly than ordinary citizens. As far as they define the public agenda, this might point in the direction of a further increase in European identification among the mass public as well. Majorities of EU citizens trust the people of member-countries. A rather clear-cut mental map of the EU finally includes all of Western Europe and some central European countries, while the countries of Eastern Europe and nonEuropean countries are not included. The latter finding points to what is probably the most serious challenge to the development of a European political community: the continuous extension of the Union. A slow and gradual extension of the Union, in addition to an effective economic policy, seem to be important preconditions for the development of an ever closer Union. POLITICAL REPRESENTATION AND R E P R E S E N T A T I V E DEMOCRACY In various chapters of this book we have used the responsible party model as a benchmark to assess the quality of the system of political representation in the European Union. However, if elections are decided according to the rules of the responsible party model but the outcome of the elections would have no effect on the composition of government or on public policy, we would at least hesitate to speak of an effective system of political representation. It is hardly a matter of dispute that in a representative democracy elections decide, directly or indirectly, which parliamentary majority should form a government and what kind of policies the government should pursue. In the European Union none of this applies. A European government with any similarity to national governments does not exist. The composition of the European Commission is independent of the outcome of European Parliament elections but depends on the nominations of the national governments. Nor does the composition of the Council depend on the outcome of European Parliament elections, but on that of national elections. Because there is no European government based on the consent of a majority of members of the European Parliament, there is no clear pattern of government and opposition parties. Therefore, the European political system does not meet the minimal defmi-

In Conclusion

263

tion of a representative democracy, offering the voters the possibility to throw the rascals out. 'Thus the elections cannot fulfil either of the two main functions of general elections—the choice of a government or the formation of public policy' (Bogdanor 1986; 168). How serious a problem this is is a matter of perspective. From the federal perspective on the European Union the argument is pretty straightforward. If the European Union can be considered a separate political entity, it should—cetera pari6tts—meet the democratic standards that apply to a state. From this view the democratic deficit is obvious: elections are not clearly translated into political power, the EP is too weak, the Commission is not elected, and the Council of Ministers and European Council are not properly controlled or accountable (Newman 1996: 173). However, if one does not share this view and wants to see the Union as no more than a co-operation of sovereign states, one would not identify a failure of democracy at the European level. From this perspective a possible democratic deficit is located at the national level and consists of a lack of domestic control over the EU (Newman 1996: 173). But even if one refuses to recognize the European Union as a separate polity, it is not so easy to brush aside complaints about the failure of democracy at the European level. Nobody can get around the fact that there is a directly-elected European parliament. And the least these elections suggest is that the votes of the people make a difference. If they do not, or if people think they do not, this might be detrimental to the legitimacy, not only of the European Parliament, but of the whole European project. Van der Eijk and Franklin (1996: 331) argue that a clear relevance of European elections for the allocation of power and thereby for the direction of European policy-making would not leave the voters unaffected. Essentially the same argument was used by Bogdanor to plead for a radical change in the European political system (Bogdanor 1986; Bogdanor and Woodcock 1991). He discussed two possible alternatives to strengthen the relationship between the outcome of the elections and the allocation of power in the European Union. The first possibility is the introduction of a parliamentary system. This implies that the Commission should be based on the confidence of (the majority of members of) the European Parliament. The Commission would become responsible to the Parliament, and the Union would come to resemble a parliamentary system of government. European elections

264

In Conclusion

would gain in significance since they would help determine the political colour of the Union executive. The members of the Commission would share a similar political tendency, allowing electors who disliked its orientation over the last five years to vote through the European Parliament elections for a Commission of a different political orientation (Bogdanor and Woodcock 1991: 483). This, he expects, would lead to an increase in interest and voter turnout. An alternative proposal is the introduction of a presidential system. It entails directly electing the European Commission as a whole, using a two-ballot system. According to Bogdanor and Woodcock (1991: 490), the introduction of a presidential system 'would probably serve to encourage bipolarity in a multiparty system, a realignment of the Community party system which would clarify the basis of electoral choice and hence make democratic accountability easier'. Which problems would implementing these proposals solve? Would they help increase the involvement of European citizens, manifesting itself in higher turnout? Such a result is not unthinkable. In both proposals the power question is at issue, which might arouse the interest of European citizens more than does the present situation. Are these proposals practicable? This is most unlikely. The practicability of the proposal to introduce a kind of presidential system in the European Union can be disputed, mainly because it would imply a shift of power from the member-states and the Council of Ministers to the Commission, in particular the President. As Bogdanor (1986: 174) argues, 'the Commission would genuinely become what it was originally intended to be, the powerhouse of European integration*. Being directly elected by the people, the Commission's democratic legitimacy compared to the Council of Ministers would increase enormously. And democratic legitimacy is a potential source of power. It is most unlikely that such a radical shift of power would be accepted by the member-states. 'There is therefore no question of the European Council accepting the idea of a directly elected Commission which could then present itself as a real EU executive. Nor, of course, would the EP support the proposal since a directly elected Commission would also rival its democratic legitimacy. Indeed, since the Commission's povvers are so much greater, it could even eclipse the EP' (Newman 1996: 187). More or less the same argument applies to the proposal to introduce a parliamentary system.

In Conclusion

265

Are these proposals desirable? It all depends on what one wants to achieve. If the purpose is to increase the turnout at European elections, the answer is probably yes. However, turnout is not the only factor to take into account. The introduction of either a presidential system or a full-fledged parliamentary system where the most relevant dividing line in Parliament would be the distinction between government and opposition would completely change the institutional context of the European Parliament. And it is most likely that such changes would influence the functioning of the European Parliament and its members. One of the reasons why party groups within the EP and Parliament as a whole usually act unanimously is the institutional weakness of the European Parliament. In order to carry any weight at all, Parliament is forced to unite. However, in a situation where the field of force is dominated by the distinction between government and opposition, all this might change. The election of the Commission (either directly or by a parliamentary majority) might enhance its legitimacy and its powers; in a more general sense, it might also strengthen the legitimacy of supranational institutions. At the same time, this might be the death blow to consensual politics in Parliament, A powerful European-level government, supported by a political majority in Parliament, would evoke its own resistance. The development towards a clear-cut distinction between government and opposition that at first might be considered progress towards a more democratic Union might turn out to be progress with a vengeance. Rather than strengthening and democratizing the Union, it might undermine the position of Parliament as the motor of European integration. Therefore, even without referring to the predictable unwillingness of national governments—and national parliaments, for that matter—to agree with a development that would undermine their own power, it is very disputable that democracy at the European level would be enhanced by the development of a majoritarian system of government. As Lijphart (1984) convincingly argued, a majoritarian model of government requires specific conditions to function properly. One is a homogeneous society. Lijphart argued that several of the present memberstates of the Union do not, or at least did not at the time, meet this requirement. It is a challenge to imagine how Europe as a whole would meet it. Democracy at the European level cannot and should not be based on anything other than a consensual form of government which

266

In Conclusion

would maximize consensus both within Parliament and among the member-states. Such a system of government does not offer a clear-cut choice to the voters, but that is the unfortunate fate of any consensual system of government. A European system of government where the voters themselves decide with their vote on the political colour of a European government—the Commission—still seems far off. However, that is not to say that one must resign oneself to the status quo of European elections in which national political parties with national candidates compete on national issues. The basic idea of consensual politics at the European level would not be threatened if European political parties competed for the votes of a European electorate within the constraints of a shared electoral system. It stili would not render European elections the most exciting thing on earth, but it would at least direct the attention to the proper level of government.

NOTES

1 2

3

This is not to say that national or regional interests should be kept off the agenda, as little as territorial interests are or should be kept off the national political agenda, The relevance of these conditions can easily be demonstrated by phrasing the underlying democratic premises negatively: how democratic is a representative system if people do not have the choice between different political parties with different policy programmes, or if political parties cannot live up to their pledges because of a lack of cohesion; how well is a system of political representation functioning if the voters have no idea where political parties stand on important policy issues or—if they do— do not take those stands into account when deciding how to vote? Of course, to some extent this is begging the issue, National political parties do not compete on these issues either, and certainly not in national elections. However, this is not unique for this kind of issues. It is only one of the many pieces of evidence that the populist assumption on which the Responsible Party Model is based is untenable. Elections can never guarantee an electoral mandate on any particular issue (Riker 1984; Thomassen 1994). Only a referendum can solve this problem, and even then the outcome cart be contaminated by considerations that have nothing to do with the subject of the referendum (Franklin, van der Eijk, and Marsh 1995),

In Conclusion 4

267

The attitudes of people and political parties on European integration were never totally independent from their perception of the kind of policies 'Brussels' would pursue. This is nicety illustrated by the following cry from the heart, made by Mrs Thatcher when she was still prime minister, 'for a free, enterprising and open Europe—not a socialist Europe which would suffer from detailed controls, bureaucracy and protectionism. A socialist majority in the European Parliament would work in that direction, seeking to turn the clock back to policies associated in Britain with failure and decline. We must elect as many Conservative members of the European Parliament as possible to buttress the present Centre-Right majority' (as quoted in Bogdanor 1989: 210).

This page intentionally left blank

Appendix The European Representation Study

THE European Representation Study consists of four distinct survey operations: the European Election Study 1994, the European Candidates Study 1994, the Study of Members of the European Parliament 1996, and the European Members of Parliament Study. This appendix reports technical details of each of these surveys.

THE E U R O P E A N ELECTION STUDY 1994 (EES94)

The European Election Study 1994 (EES94) is a survey study of the electorates of the then twelve member-states of the European Union, It essentially consists of a survey conducted immediately after the European Parliament election of June 1994 as a part of Eurobarometer 41.1. The study was prepared by an international group of scholars involving Pilar del Castillo (Madrid), Roland Cayrol (Paris), Cees van der Eijk (Amsterdam), Mark Franklin (Houston), Renato Mannheimer (Genoa), Michael Marsh (Dublin), Karlheinz Reif (Brussels), and Colette Ysmal (Paris). This group was co-ordinated by Hermann Schmitt and supplemented by a number of country specialists: Mario Baccalhau (Lisbon), Michael Guilljam (Goteborg), Ilias Nikolakopuolos (Athens), Lieven De Winter (Louvain), and Torben Worre (Copenhagen). The survey was made possible through grants from the Deutsche Forschungsgemeimchaft (DFG) and the Nederlandse Stickling voor Wetenschappelijk Onderzoek (NWO). Additional support was given by the University of Amsterdam and the University of Mannheim. Fieldwork was carried out by INRA Europe (International Research Associates), a European network of market- and public opinion research agencies located in Brussels.

270

Appendix

Like standard Eurobarometers, the EES94 was based on national multi-stage probability samples. The universe in each country was the national resident population aged 15 years and older. Sampling points were drawn proportional to national population density. Face-to-face interviews were carried out in the respective language in respondents* homes. Fieldwork dates, sample sizes and return rates are given in Table A.I The data are available from the Zentralarchiv fur Empirische Sozialforschung, Cologne, Germany (Study Number 2865). TABLE A.I. EES94: fieldwork period and number of interviews and return rates Fieldwork period

Belgium Denmark Germany East Germany West Greece Spain France Ireland Italy Luxembourg The Netherlands Portugal Great Britain Northern Ireland

13.6. 11.6. 13.6. 13.6. 13.6. 13.6. 13.6. 13.6. 14.6. 13.6, 11.6. 15.6. 13.6. 14.6.

- 30. 6. - 28. 6. - 05. 7. - 02. 7. - 30. 6. - 27. 6. - 27. 6. - 08. 7. - 28. 6. - 08. 7. - 02. 7. - 04. 7. - 07. 7. - 30. 6.

1994 1994 1994 1994 1994 1994 1994 1994 1994 1994 1994 1994 1994 1994

Number of interviews

Return rate*

1003 1000 1052 1082 1002 1000 1000 1000 1067 502 1005 1000 1078 305

47 37 48 45 50 43 47

53 40 43 39 51 50 32

a = completed interviews in per cent of effective contacts.

The three elite components of the study to which we now turn were jointly prepared by a research group consisting of Richard Katz (Baltimore, MD), Pippa Morris (Cambridge, MA), Jacques Thomassen (Enschede), and Bernhard Wessels (Berlin). Herman Schmitt participated in most meetings of the group to ensure comparability with the voters' study.

The European Representation Study

271

THE 1994 EUROPEAN CANDIDATES STUDY (ECS94)

The European Candidates Study 1994 was carried out by mail questionnaire, in the weeks ahead of the 1994 European Parliament election, in ten of the then twelve member-countries of the EU, It was coordinated by Jacques Thomassen and carried out by himself in The Netherlands, and by Tom Bryder in Denmark, Cecile Chavel in France, Irene Delgado and Lourdes Lopez Nieto in Spain, Louisa Gardella in Italy, Peter Linch in Great Britain, Michael Marsh in Ireland, Maria Jose Stock in Portugal, Bernhard Wessels and Achitn Kielhorn in Germany, and Lieven De Winter in Belgium and Luxembourg, In the weeks preceding the 1995 European Parliament election in Sweden, the study was carried out there by Martin Brothen, Peter Esaiasson, and SOren Holmberg. Overhead funds were provided by the University of Twente. Table A.2 shows for each country the number of interviews and the return rate. The data are available from the Zentralarchiv fur Empirische Sozialforschung, Cologne, Germany (Study Number 3077), TABLE A.2, ECS94: number of interviews and return rates Number of interviews Return rate Belgium Britain Denmark France Germany Ireland Italy Luxembourg Netherlands Portugal Spain Sweden Overall

117 134 105 104 395 12 137 33 125 24 74 514

35 38 57 13 34 23 10 55 51 8 4 86

1726

35

272

Appendix THE 1996 MEP STUDY AND THE MNP STUDIES

The study among the members of the European Parliament was coordinated by Bemhard Wessels and funded by the Deutsche Forschungsgemeinschaft. Face-to-face interviews were conducted by INFRA TEST BURKE, The study among the members of the national parliaments was co-ordinated by Bernhard Wessels and Jacques Thomassen. National study directors were Irene Delgado and Lourdes Lopez Nieto in Spain, Louisa Gardella in Italy, Soree Holmberg, Martin Brothen, and Peter Esaiasson in Sweden, Manina Kakepati in Greece, Christina Leston Bandeira in Portugal, Michael Marsh in Ireland, Bernhard Wessels and Achim Kielhorn in Germany, Lieven De Winter and Patrick Dumont in Belgium and Luxembourg, and Colette Ysmal in France. The MNP studies were administered as mail surveys. Overhead funds for this part of the study were provided by the universities of Mannheim and Twente and the Wissenschaftszentrum Berlin. Table A,3 informs about fieldwork period, number of interviews, return rates, and representativity of both the MEP Study and the MNP Studies. Given the great variation in return rates, the question of representativity of these studies is of major importance. Representativity has been tested for party and gender and, where data were available, for age (and for the MEP study also for nation). The Duncan index of dissimilarity, measuring deviations between the original and the sample distributions, shows that altogether sample deviations from the real distributions are acceptably low in most studies. For both the MEP Study and the MNP Studies, the deviation of party composition in the samples from their universes is about eight percentage points; it is about one for gender compositions and about three for age composition. For individual national studies the Duncan index is below ten in most cases except for party and age for some countries. A methodological caveat may remain for those countries with extremely low return rates. However, since most analyses reported in this book describe relations between variables rather than univariate distributions, this potential problem should be minimized. The data of both studies are available from the Zentralarchiv /fir Etnpirische Sozialforschung, Cologne, Germany (Study Numbers 3078 [MEP] and 3079 [MNP]).

TABLE A3, MEP Study and MNP Studies: fieldwork period, number of interviews, return rates, and representativity Number of Duncan index of dissimilarity 8 regarding Fieldwork period interviews Return rate Party Gender Age Nation 20.05,- 2 1.6. 1996

314

50

8

0

4

9

MNP Studies Belgium b France Germany Greece Ireland Italy Luxembourg Netherlands Portugal Spain Sweden

3.6. 1996-28.4. 1997 2.7. 1996- 10.7. 1997 21 .5. -6. 8. 1996 15. 4. -31.7. 1996 9. 7. -25. 10. 1996 June -October 1996 12.6.1996-28.2. 1997 29. 4. - 19. 7. 1996 13. 6. -4. 8. 1996 Autumn 1996 5, 3, - 10. 12. 1996

87 146 317 60 71 94 28 65 54 130 315

58 25 47 20 43 15 47 43 24 37 90

7 7 16 3 18 14 11 9 5 2

i 0 2 1 0 2 3 4 1 3 1

4 7 6 11 . I

itIt # * # # # # #

MNP overall

29.4. 1996- 10.7. 1997

1367

38

7

1

3

MEP Study

8

#

if i

Notes: (a) The Duncan index measures the percentage point differences between the respective distribution in the universe and the sample. It ranges from 0 (no difference) to 100 (maximum difference), (b) House of Representatives. - = not available. $ = not applicable.

This page intentionally left blank

References ALMOND, GABRIEL A. and VERSA, SIDNEY, The Civic Culture: Political Attitudes and Democracy in Five Nations (Princeton: Princeton University Press, 1963). ALPERT, EUGENE J,, 'A Reconcepturalization of Representational Role

Theory', Legislative Studies Quarterly, 4:4 (1979), 587-603. AMERICAN POLITICAL SCIENCE ASSOCIATION, 'Towards a More Responsible

Two-Party

System', American

Political

Science

Review, 22 (1950), 475-510. ANDERSON, CHRISTOPHER J. and KALTENTHALER, KARL, 'The Dynamics of Public Opinion toward European Integration, 19731993', European Journal of International Relations, 2:2 (1995),

175-99. ANDERSON, SVEIN S. and ELIASSEN, KJELL A. (eds.), The European Union: How Democratic Is It? (London: Sage, 1996). ANDEWEO, RUDY, 'The Reshaping of National Party Systems', West European Politics, 18 (1995), 58-78. ARDY, BRIAN, 'The National Incidence of the European Community Budget', Journal of Common Market Studies, 26 (1988), 401-29. BADURA, BERNHARD and REESE, JURGEN, Jungparlamentarier in Bonn; Ihre Sozialisation im Dentschen Bundestag (Stuttgart and Bad Canstatt: Frommann-Holzboog, 1976). BARDI, LUCIANO, // Parlamento delta Comunita Europea: Legittimita e riforma (Bologna: Mulino, 1989). BARRY, BRIAN M., Sociologists, Economists and Democracy (London:

Collier-Macmillan, 1970). BECK, NATHANIEL and KATZ, JONATHAN N., 'What to Do (and Not to Do) with Time-Series-Cross-Section Data in Comparative Polities', American Political Science Review, 89 (1995), 634-47. BEETHAM, DAVID and LORD, CHRISTOPHER, Legitimacy and the EU (London and New York: Longman, 1998), BELL, CHARLES G. and PRICE, CHARLES M., The First Term: A Study of Legislative Socialization (Beverly Hills: Sage, 1975).

276

References

BERELSON, BERNARD, LAZARSFELD, PAUL F., and MCPHEE, WILLIAM N., Voting: A Study of Opinion Formation in a Presidential Campaign (Chicago: University of Chicago Press, 1954), BOGDANOR, VERNON, 'The Future of the European Community: Two Models of Democracy*, Government and Opposition, 21 (1986), 161-76. BOGDANOR, VERNON, 'Direct Elections, Representative Democracy and European Integration', Electoral Studies, 8 (1989), 205-16. BOGDANOR, VERNON and WOODCOCK, GEOFFREY, 'The European Community and Sovereignty', Parliamentary Affairs, 44 (1991), 481-92, BOSCH, AGUSTI and NEWTON, KENNETH, 'Economic Calculus or Familiarity Breeds Content?', in Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and Internationalized Governance (Oxford: Oxford University Press, 1995), 73-104. BRAND, KARL-WERNER, Neve soziale Bewegungen: Entstehung, Fiinktion und Perspektive neiter Protestpotentiale: Eine Zwischenhilanz (Opladen: Westdeutscher Verlag, 1982). BROWN, M.C. and HALABY, C.N., 'Pooling Time Series and CrossSectional Data: Estimating the "Cost" of Political Machines', paper prepared for the Annual Meeting of the American Political Science Association, Denver, CO, 1982, BUDGE, IAN and FARLIE, DENNIS J., Voting and Party Competition: A Theoretical Critique and Synthesis Applied to Surveys from Ten Democracies (London: Wiley, 1977). BUDGE, IAN and FARLIE, DENNIS 1, Explaining and Predicting Elections: Issue Effects and Party Strategies in 23 Democracies

(London: Allen and Unwin, 1983). BULLETIN DE COMMUNITE EUROPEENNE, 12-1992. BUNDESVERFASSUNGSGERJCHT, 'Urteil vom 12. Oktober 1993', Europaische Grundrechte Zeitschrift, 20:17 (1993), 429-46. CAMPBELL, ANGUS, CONVERSE, PHILIP E,, MILLER, WARREN E., and STOKES, DONALD E. (eds,), The American Voter (New York and London: Wiley, 1960). CEPR, Making Sense of Subsidiarity: How Much Centralization for Europe? A CEPR Annual Report (London: Centre for Economic Policy Research, 1993), CINNIRELLA, MARCO, 'Towards a European Identity? Interactions Between the National and the European Social Identities Manifested

References

277

by University Students in Britain and Italy", British Journal of Social Psychology, 36:1 (1997), 19-31. ClTRlN, JACK, 'Comment: The Political Relevance of Trust in Government', American Political Science Review, 68 (1974), 973-

88. CLARKE, HAROLD D., DUTT, NITISH, and KOKNBERG, ALLAN, 'The Political Economy of Attitudes Toward Polity and Society in Western European Democracies', Journal of Politics, 50:4 (1993), 998-1021. CLAUSEN, AAOE R., 'The Accuracy of Leader Perceptions of Constituency Views', Legislative Studies Quarterly, 2:4 (1977),

361-84. CONOVER, PAUL J. and FELDMAN, STANLEY, 'How People Organize the Political World: A Schematic Model', American Journal of Political Science, 28 (1984), 95-126. CONVERSE, PHILIP E., 'Public Opinion and Voting Behavior', in Fred I. Greenstein and Nelson W. Polsby (eds.), Handbook of Political Science, vol. 4 (Reading: Addison-Wesley, 1975), 75-169. CONVERSE, PHILIP E. and PIERCE, ROY, Political Representation In France (Cambridge, MA: Belknap Press of Harvard University Press, 1986). CORBEY, DOROTHEA J.M., Stilstand is vooruitgang, De dialectiek van het Ewopese integratieproces (Assen and Maastricht: Van Gorcum, 1993). COX, GARY W., Making Votes Count: Strategic Coordination in the World's Electoral Systems (Cambridge: Cambridge University

Press, 1997). CROZIER, MICHEL, HUNTINGTON. SAMUEL P., and WATANUKJ, JGJI, The Crisis of Democracy: Report on the Governability of Democracies to the Trilateral Commission (New York: New York University Press, 1975), DAALDER, HANS, 'A Crisis of Party?', Scandinavian Political Studies, 15:4 (1992), 269-88. DAHL, ROBERT A., Democracy and its Critics (New Haven: Yale

University Press, 1989). DAHL, ROBERT A. and TUFTE, EDWARD R., Size and Democracy (Stanford : Stanford University Press, 1973).

278

References

DAHRENDORF, RALF, Homo sociologicus: Ein Versuch zur Geschichte, Bedeutung and Kritik der Kategorie der sozialen Rolle (KOln and Opladen: Westdeutscher Verlag, 1969), 8* edition. DALTON, RUSSELL J., Citizen Politics: Public Opinion and Political Parties in Advanced Industrial Democracies (Chatham, NJ: Chatham House, 1996), 2nd edition.

DALTON, RUSSELL J., 'Political Support in Advanced Industrial Democracies', in Pippa Morris (ed.), Critical Citizens: Global Support for Democratic Development (Oxford: Oxford University Press, 1999, forthcoming). DALTON, RUSSELL J, AND EICHENBERG, RICHARD C., 'Economic Evaluations and Citizen Support for the European Integration', paper presented at the Annual Meeting of the American Political Science Association, Washington DC, 1991. DALTON, RUSSELL J. and EICHENBERO, RICHARD C., 'A People's Europe: Citizen Support for the 1992 Project and Beyond', in Dale L. Smith and James Lee Ray (eds.), The 1992 Project and the Future of Integration in Europe (Armonk: M.E. Sharpe, 1992), 73-91. DALTON, RUSSELL J. and EICHENBERG, RICHARD C,, 'Economic Perceptions and Citizen Support for European Integration', unpublished paper, 1994.

DALTON, RUSSEL J., FLANAGAN, SCOTT C., and BECK, PAUL ALLEN (eds.), Electoral Change in Advanced Industrial Democracies:

Realignment or Dealignment? (Princeton: Princeton University Press: 1984). DE BRUYCKER, PHILIPPE, 'Les garanties europeennes territoriales', in Jacques Lemaire and Andre Miroir (eds.), La Belgique et ses nations dans la nottvelle Europe, La Pensee et les Hommes (Bruxelles:

Universite de Bruxelles, 1997) 145-59. DEFLEM, MATHIEU and PAMPEL, FRED C., 'The Myth of Post-National Identity: Popular Support for European Unification', Social Forces, 75:1(1996), 119-44. DEUTSCH, KARL W. et al., Political Community and the North Atlantic

Area (Princeton, NJ: Princeton University Press, 1957). DE WINTER, LIEVEN and FROGNIER, ANDRE-PAUL, 'L'evolution des identitfe politiques territoriales en Belgique durant la periode 19751995', in Serge Jaumain (ed,), La reforms de I'Etat... et apres? L'impact des debats imtitutionnels en Belgique et an Canada

(Bruxelles: Editions de 1'Universite de Bruxelles, 1997), 161-76.

References

279

DE WINTER, LIEVEN and TURSAN, HURJ, Regionalisi Parties in Western Europe (London, New York: Routledge, 1998). DINAN, DESMOND, Ever Closer Union? An Introduction to the European Community (London: MacMillan, 1994). DONY, MARIANNE, *Les regions face au principe de subsidiarite', in Jacques Lemaire and Andre" Miroir (eds.), La Belgique et ses nations dans la nouvelle Europe, La Pensee et les Hommes (Bruxelles: Universite de Bruxelles, 1997), 161-66. DOWNS, ANTHONY, An Economic Theory of Democracy (New York: Harper & Row, 1957). DUCHESNE, SOPHIE and FROGNIER, ANDRE-PAUL, 'Is there a European Identity?', in: Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and Internationalized Governance (Oxford and New York: Oxford University Press, 1995), 193-226. DUFF, ANDREW (ed.), Subsidiarity within the European Community, A Federal. Trust Report (London: Federal Trust for Education and

Research, 1993). DUFF, ANDREW, PINDER, JOHN, and PRYCE, ROY (eds.), Maastricht and Beyond: Building the European Union (London: Routledge, 1994). Du GRANRUT, CLAUDE, Europe: Le temps des regions (Paris: Librairie Generate de Droit et de Jurisprudence, 1994). Du GRANRUT, CLAUDE, La citoyennete europeenne: Une application du principe de subsidiarite (Paris: Librairie G6nerale de Droit et de Jurisprudence, 1997). EASTON, DAVID A., A Systems Analysis of Politcal Life (New York: John Wiley, 1965, repr. 1967 and 1979). EASTON, DAVID A., 'A Re-Assessment of the Concept of Political Support', British Journal of Political Science, 5 (1975), 435-57. EICHENBERG, RICHARD C, 'Measurement Matters: Cumulation in the Study of Public Opinion and European Integration', paper presented

at the Annual APSA Convention, Boston, September 1998. EICHENBERG, RICHARD C, and DALTON, RUSSELL J., 'Europeans and the European Community: The Dynamics of Public Support for European Integration', International Organization 47:4 (1993), 507— 34. EIPA, Subsidiarity: The Challenge of Change: Proceedings of the Jacques Deters Colloquium: Subsidiarity—Guiding Principle for Future EC Policy Responsibility? Organized by the European

280

References

Institute of Public Administration, Maastricht, 21—22 March 1991 (Maastricht: European Institute of Public Administration, 1991), ElSER, J.R., 'Attitudinal Judgment: The To and Fro of AssimilationContrast', in Donald Granberg and Gian Samp (eds.), Social Judgment and Intergroup Relations: Essays in Honor of Muzafer S/K?r/y(New York: Springer-Verlag, 1992). ELCHARDUS, MARK, Jongeren en Cultuur in Beweging (Leuven:

Davidsfonds, 1997). ENELOW, JAMES M. and HINICH, MELVIN J. The Spatial Theory of Voting: An Introduction (Cambridge and New York: Cambridge University Press, 1984), ESAIASSON, PETER and HOLMBERG, SOREN, Representation from Above: Members of Parliament and Representative Democracy in Sweden (Aldershot: Dartmouth, 1996). ESTELLA DE NORIEGA, ANTONIO, 'The Principle of Subsidiarity and its Critique: A "Contextual" Analysis of the Principle of Subsidiarity', unpublished Ph.D. Thesis, Florence: European University Institute, 1997. EULAU, HEINZ et al, 'Some Empirical Observations on the Theory of Edmund Burke', in Heinz Eulau and John C. Wahlke (eds.).. The Politics of Representation: Continuities in Theory and Research (Beverly Hills and London: Sage, 1978), 111-26, EULAU, HEINZ and KARPS, PAUL D., 'The Puzzle of Representation', in Heinz Eulau and John C. Wahlke (eds.), The Politics of Representation: Continuities in Theory and Research (Beverly Hills and London: Sage, 1978), 55-72. EUROBAROMETER TRENDS, Brussels, Commission of the European Community, 1994. FIDE, The Principle of Subsidiarity: 16lh Congress of the International Federation for European Law (Rome: International Federation for European Law, 1994). FIORINA, MORRIS P., Retrospective Voting in American National Elections (New Haven: Yale University Press, 1981). FRANKLIN, MARK N., MACKIE, TOM, and VALEN, HENRY (eds.), Electoral Change: Responses to Evolving Social and Attitudinal Structures in Western Countries (Cambridge: Cambridge University

Press, 1992). FRANKLIN, MARK N., MARSH, MICHAEL, and MCLAREN, LAUREN, 'Uncorking the Bottle: Popular Opposition to European Unification

References

281

in the Wake of Maastricht', Journal of Common Market Studies, 32:4(1994), 455-72. FRANKLIN, MARK N., VAN DER EIJK, CEES, and MARSH, MICHAEL, 'Referendum Outcomes and Trust in Government: Public Support for Europe in the Wake of Maastricht', West European Politics, 18:3 (1995), 101-17. FRANKLIN, MARK N., VAN DER EIJK, CEES, and OPPENHUIS, ERIK, 'Turnout Decline and the Motivational Basis of British Participation in the European Elections 1994', in Colin Railings, David Farrell, David Broughton, and David Denver (eds.), British Elections and Parties Yearbook 1995 (London: Harvester Wheatsheaf, 1996). FRANKLIN, MARK N. and WLEZIEN, CHRISTOPHER, 'The Responsive Public: Issue Salience, Policy Change, and Preferences for European Unification', Journal of Theoretical Politics, 9:3 (1997), 347-63. FUCHS, DIETER, GUIDOROSSI, GIOVANNA, and SVENSSON, PALLE, 'Support for the Democratic System', in Hans-Dieter Klingemann and Dieter Fuchs (eds,), Citizens and the State (Oxford and New York: Oxford University Press, 1995), 323-53. FUCHS, DIETER and KLINGEMANN, HANS-DIETER, The Left-Right Schema: Theoretical Framework', in M. Kent Jennings, Jan W. van Deth, et al, Continuities in Political Action: A Longitudinal Study of Political Orientations in Three Western Democracies (Berlin and New York: Walter de Gruyter, 1990), 203-34. GABEL, MATTHEW, 'Public Support for European Integration: An Empirical Test of Five Theories', Journal of Politics, 60:2 (1998), 333-54. GABEL, MATTHEW and PALMER, HARVEY D., 'Understanding Variation in Public Support for European Integration', European Journal of Political Research, 27:1 (1995) 3-19. GABEL, MATTHEW and WHITTEN, GUY D., 'Economic Conditions, Economic Perceptions, and Public Support for European Integration', Political Behaviour, 19:1 (1997), 81-96. GABRIEL, OSCAR W., 'Politische Einstellungen und politische Kultur', in Oscar W. Gabriel and Frank Brettschneider (eds.), Die EUStaaten in Vergleich: Stnikturen, Prozesse, Politikmhalte (Opladen: Westdeutscher Verlag, 1994), 2Btl edition., 96-133. GAFFNEY, JOHN (ed.)> Political Parties and the European Union (London and New York: Routledge, 1996).

282

References

GAMBLE, ANDREW, 'Recession and Disenchantment with Europe', in Jack E.S, Hayward (ed.), The Crisis of Representation in Europe (London: Frank Cass, 1995), 158-74, GlSCARD D'ESTAING, VALERIE, 'Rapport de la Commission institutionnelle sur le principe de subsidiarite', Parlement European, Doc A3-267/90, 1990. GOLUB, JONATHAN, 'Sovereignty and Subsidiarity in EU Environmental Policy*, EUI working paper Robert Schuman Center, No, 96/2, 1996. GRANBERO, DONALD, "An Anomaly in Political Perception', Public Opinion Quarterly, 49:4 (1985), 504-16, GRANBERG, DONALD, 'Political Perception', in Shanto lyengar and William S, McGuire (eds.), Explorations in Political Psychology (Durham: Duke University Press, 1993). GRANBERO, DONALD and HOLMBERG, SOREN, The Political System Matters: Social Psychology and Voting Behavior in Sweden and the United States (Cambridge: Cambridge University Press, 1988). GRANBERG, DONALD, KASMER, JEFF, and NANNEMAN, TIM, 'An Empirical Examination of Two Theories of Political Perception*, Western Political Quarterly, 41:1 (1988), 29-46. GREEN, DONALD P. and SHAPIRO, IAN, Pathologies of Rational Choice Theory: A Critique of Applications in Political Science (New Haven and London: Yale University Press, 1994). GUJARATI, DAMODAR N., Basic Econometrics (New York and London: McGraw-Hill, 1995), 3rd edition.. GUNTHER, RICHARD, DIAMANDOUROS, P. NIKIFOROS, and PUHLE, HANS-JORGEN (eds.), The Politics of Democratic Consolidation: Southern Europe in Comparative Perspective (Baltimore: The Johns Hopkins University Press, 1995). HAAS, ERNST B, and SCHMJTTER, PHILIPPE C,, 'Economics and Differential Patterns of Political Integration: Projections about Unity in Latin America', International Organization, 18:4 (1964), 705-37. HANDLEY, DAVID H., 'Public Opinion and European Integration: The Crisis of the 1970s', European Journal of Political Research, 9:4 (1981), 335-64. HAYWARD, JACK E.S, (ed.), The Crisis of Representation in Europe (London: Frank Cass, 1995). HAYWARD, JACK E.S. (ed.), Elitism, Populism and European Politics (Oxford: Clarendon Press, 1996).

References

283

HEIDER, FRITZ, 'Attitudes and Cognitive Organization', Journal of Psychology,21 (1946), 107-12. HEIDER, FRITZ, The Psychology of Interpersonal Relations (New York: Wiley, 1958). HERZOO, DIETRICH, Politische Karrieren: Selektion and Profess ionalisierung politischer Fuhrungsgruppen (Opladen:

Westdeutscher Verlag, 1975). HESSE, JOACHIM JENS and SHARPE, LAURENCE ]., 'Local Government in International Perspective: Some Comparative Observations', in Joachim Jens Hesse (ed.), Local Government and Urban Affairs in International Perspective: Analysis of Twenty Western Industrialized Countries (Baden-Baden: Notnos Verlagsgesellschaft, 1991), 603-21. HlX, SIMON, 'Parties at the European Level and the Legitimacy of EU Socio-Economic Policy', Journal of Common Market Studies, 33 (1995), 527-54. HlX, SIMON, 'Dimensions and Alignments in European Union Politics: Cognitive Constraints and Partisan Responses', paper prepared for the Annual Meeting of the American Political Science Association, Boston, MA, 1998. HlX, SIMON and LORD, CHRISTOPHER, Political Parties in the European (/won (New York: St Martin's Press, 1997). HOFFMANN, STANLEY, 'The European Process at Atlantic CrossPurposes', Journal of Common Market Studies, 3 (1965), 85—101. HOFFMANN-LANGE, URSULA, Eliten, Macht und Konflikt in der Bitndesrepublik (Opladen: Leske + Budrich, 1992). HOLLAND, MARTW, European Community Integration (London: Pinter, 1993). HOLMBERG, SOREN, 'Political Representation in Sweden*, Scandinavian Political Studies, 12:1 (1989), 1-36. HOLMBERG, SQREN, 'Collective Policy Congruence Compared', Department of Political Science, University of Gothenburg, September 1996 (unpublished manuscript). HOOGHE, LlESBETH and MARKS, GARY, 'Restructuration territoriale au sein de PUnion Europeenne: Les pressions regionales', in Vincent Wright and Sabino Cassese, La recomposition de I'Etat en Europe (Paris: La Decouverte, 1996), 207-26.

284

References

HRBEK, RUDOLF (ed.). Die Anwendung des Subsidiaritatsprinzips in der Evropaischen Union: Erfahrungen und Perspektiven (BadenBaden: Nomos Verlagsgesellschaft, 1995). HUBER, JOHN D. and INQLEHART, RONALD, 'Expert Interpretations of Party Space and Party Locations in 42 Societies', Parly Politics, 1:1 (1995), 73-111. HUBER, JOHN D. and POWELL, G. BINGHAM, JR., 'Congruence Between Citizens and Policymakers in Two Visions of Liberal Democracy', World Politics, 46:3 (1994), 291-326, HUNTrNGTON, SAMUEL P., American Politics: The Promise of Disharmony (Cambridge, MA, and London: Harvard University Press, 1981). HUNTINGTON, SAMUEL P., The Third Wave: Democratization in the Late Twentieth Century (Norman: University of Oklahoma Press, 1991). INGLEHART, RONALD, The Silent Revolution: Changing Values and Political Styles Among Western Publics (Princeton, NJ: Princeton University Press, 1977). INGLEHART, RONALD, 'The Changing Struture of Political Cleavages in Western Society', in Russel J. Dalton, Scott C. Flanagan, and Paul Allen Beck (eds.), Electoral Change in Advanced Industrial Democracies: Realignment or Dealignment? (Princeton: Princeton University Press: 1984), 25-69. INGLEHART, RONALD, Culture Shift in Advanced Industrial Society (Princeton, NJ: Princeton University Press, 1990). INGLEHART, RONALD, Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies (Princeton, NJ: Princeton University Press, 1997). INGLEHART, RONALD, 'Posttnodemization, Authority, and Democracy', in Pippa Norris (ed.), Critical Citizens: Global Support for Democratic Development (Oxford: Oxford University Press,

1999, forthcoming). INGLEHART, RONALD and RABIER, JACQUES-RENE, 'Economic Uncertainty and European Solidarity: Public Opinion Trends', Annals of the American Academy of Political and Social Science,

440(1978), 66-97. IYENGAR, SHANTO and KINDER, DONALD R., News That Matters: Television and American Opinion (Chicago: University of Chicago Press, 1987).

References

285

JACOBS, FRANCIS, CORBETT, RICHARD, and SHACKLETON, MICHAEL, The European Parliament (London: Cartermill Pub,, 1995), 3rf edition, JANSSEN, JOSEPH I.H., 'Postmaterialism, Cognitive Mobilization, and Public Support for European Integration', British Journal of Political Science, 21:4 (1991), 443-68. JEWELL, MALCOLM E., 'Attitudinal Determinants of Legislative Behaviour: The Utility of Role Analysis', in Allan Kornberg and Lloyd D. Musolf (eds.), Legislatures in Developmental Perspective (Durham: Duke University Press, 1970), 460-500. JEWELL, MALCOLM E., 'Legislators and Constituents in the Representative Process', in Gerhard Loewenberg, Samuel C. Patterson, and Malcolm E. Jewell (eds.), Handbook of Legislative

Research (Cambridge: Harvard University Press, 1985), 96—131. JONES, BRADFORD and STEENBERGEN, MARCO, 'Modelling Multilevel Data Structures', paper prepared for the 14th meeting of the Political

Methodology Society, Columbus, OH, July 1997, KAASE, MAX and NEWTON, KENNETH, Beliefs in Government (Oxford and New York: Oxford University Press, 1995). KATZ, RICHARD S. and MAIR, PETER, How Parties Organize: Change and Adaptation in Party Organizations in Western Democracies

(London and Thousand Oaks, CA: Sage Publications, 1994). KATZ, RICHARD S. AND WESSELS, BERNHARD (eds.), The European Parliament, the National Parliaments, and European Integration

(Oxford: Oxford University Press, 1999a). KATZ, RICHARD S. and WESSELS, BERNHARD, 'Parliaments and Democracy in Europe in the Era of the Euro', in Richard S. Katz and Bernhard Wessels (eds.), The European Parliament, the National Parliaments, and European Integration (Oxford: Oxford University Press, 1999b). KEATING, MICHAEL, Nations Against the State: The New Politics of Nationalism in Quebec, Catalonia and Scotland (London:

Macmillan, 1995). KEOHANE, ROBERT O. and HOFFMANN, STANLEY (eds.), The New European Community: Decision Making and Institutional Change (Boulder: Westview Press, 1991). KERLlNGER, FRED N,, Liberalism and Conservatism (Hillsdale: Lawrence Erlbaum. 1984).

286

References

KERREMANS, BART, Internationale Organisaties (Brussel: Katbolieke Universiteit Brussel, 1996), KlELMANNSEGG, PETER GRAF, 'Llsst sich die Europaiscfae Gemeinschaft demokratisch verfassen?', in Werner Weidenfeld (ed.), Reform der Europaischen Union: Materialien zur Revision des Maastrichter Vertrages (Giltersloh: Bertelsmann, 1995), 229^*2. KIRKPATRICK, EVRON M, 'Toward a More Responsible Two-Party System', American Political Science Review, 72 (1971), 165-77. KISSINGER, HENRY A,, Diplomacy (London: Simon & Schuster, 1994). KITSCHELT, HERBERT P., 'Left-Libertarian Parties: Explaining Innovation in Competitive Party Systems', World Politics, 40:2 (1988), 194-234, KLINGEMANN, HANS-DIETER, 'Mapping Political Support in the 1990s: A Global Analysis', in Pippa Morris (ed,), Critical Citizens: Global Support for Democratic Governance (Oxford: Oxford University Press, 1999, forthcoming). KLINGEMANN, HANS-DIETER and FUCHS, DIETER, Citizens and the State (Oxford and New York: Oxford University Press, 1995). ICONIC, THOMAS, RIEGER, ELMAR, and SCHMITT, HERMANN (eds.), Das Europaische Mekrebenen-System: Konzeptionen und Perspektiven, Mannheimer Jahrbuch fur Europaische Sozialforschung, vol. 1 (Frankfurt am Main: Campus Verlag, 1996), K.UHNLE, STEIN, 'Citizenship', in Vemon Bogdanor (ed.), The Blackwell Encyclopaedia of Political Science (Oxford and Cambridge: Basil Blackwell Ltd., 1987), 94-5, KUKLINSKI, JAMES H. and ELLING, RICHARD C., 'Representational Role, Constituency Opinion, and Legislative Roll-Call Behaviour', American Journal of 'Political Science, 21:1 (1977), 135-47, LADRECH, ROBERT, 'Political Parties in the European Parliament', in John Gaffney (ed.), Political Parties and the European Union (London and New York: Routledge, 1996), 291-307. LAZARSFELD, PAUL FELIX, BERELSON, BERNARD, and GAUDET, HAZEL, The People's Choice: How the Voter Makes Up His Mind in a Presidential Campaign (New York: Duell, Sloan, and Pearce, 1944). LEHOCQ, FABRICE and WALL, DAVID L., 'Electoral Incentives under Proportional Representation', 56'1' Annual Meeting of the Midwest Political Science Association, Chicago, IL, April 25-28, 1998,

References

287

LEIBFRIED, STEPHAN and PIERSON, PAUL (eds.), European Social Policy: Between Fragmentation and Integration (Washington, DC: Breakings Institution, 1995), LlJPHART, AREND, Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries (New Haven: Yale University Press, 1984). LIJPHART, AREND and CREPAZ, MARKUS Ml., 'Corporatism and Consensus Democracy in Eighteen Countries: Conceptual and Empirical Linkages', British Journal of Political Science, 21:4 (1991), 23 5^*6. LILLI, WALDEMAR, 'Europaische Identitat: Chance und Risiken ihrer Verwirklichung aus einer sozialpsychologischen Grundlagenperspektive', in Thomas K6nig, Elmar Rieger, and Hermann Schmitt (eds.), Europa der Burger? Voraussetzungen, Alternative^ Konsequenzen, Mannheimer Jahrbuch fur Europaische Sozialforschung, vol. 3 (Frankfurt am Main: Campus, 1998), 13958. LlNDBERG, LEON N. and SCHEINGOLD, STUART A. (eds.), Europe's Would-Be Polity: Patterns of Change in the European Community (Englewood Cliffs, NJ: Prentice Hall, 1970). UPSET, SEYMOUR MARTIN, Political Man (Garden City: Doubleday, 1960). LIPSET, SEYMOUR MARTIN and SCHNEIDER, WILLIAM, The Confidence Gap: Business, Labor, and Government in the Public Mind (Baltimore: Johns Hopkins University Press, 1987), rev. edition. LlSTHAUG, OLA and WlBERG, MATTI, 'Confidence in Political and Private Institutions', in Hans-Dieter Klingemann and Dieter Fuchs (eds.). Citizens and the State (Oxford and New York: Oxford University Press, 1995), 298-322. LOCKERBIE, BRAD, 'Economic Dissatisfaction and Political Alienation in Western Europe*, European Journal of Political Research, 23:3 (1993), 281-93. LONG, J. SCOTT, Regression Models for Categorical and Limited Dependent Variables (Thousand Oaks: Sage Publications, 1997), LOURAU, RENE, Le principe de subsidiarite contre I'Europe (Paris: Presses Universitaires de France, 1997), LOBBE, HERMANN, Abschied vom Superstaat: Vereinigte Staaten von Europa wirdes nieht geben (Berlin: Siedler Verlag, 1994).

288

References

LUTTBEG, NORMAN R. (ed.), Public Opinion and Public Policy: Models of Political Linkage (Homewood: Dorsey Press, 1974), rev, edition. MAILLET, PIERRE, Trots Defis de Maastricht: Convergence, cohesion, siibsidiarite (Paris: L'Harmattan, 1993). MARKS, GARY and WILSON, CAROLE J., The Past in the Present: A Cleavage Theory of Party Response To European Integration (rnimeo, University of North Carolina. 1998). MARSH, MICHAEL and NORMS, PIPPA (eds.), Political Representation in the European Parliament (Dodrecht: Kluwer, 1997), MARSH, MICHAEL and WESSELS, BERNHARD, 'Territorial Representation*, European Journal for Political Research, 32:2 (1997), Special Issue ed. by Michael Marsh and Pippa Norris, 22741. MARTINOTTI, GuiDO and STEFANIZZI, SONIA, 'Europeans and the Nation State', in Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and Internationalized Governance (Oxford and New York: Oxford University Press, 1995), 163-89. MASSART-PIERARD, FRANCOISE, 'L'identite institutionrtelle des regions au sein de 1'Union Europeenne', Revue Internationale de Politique Comparee, 5:1 (1998), 35-51. MILLER, ARTHUR H., 'Political Issues and Trust in Government: 19641970*, American Political Science Review, 68 (1974), 951-72. MILLER, ARTHUR H. and BORRELLI, STEPHEN A., 'Confidence in Government During the 1980s', American Politics Quarterly, 19:2 (1991), 147-73. MILLER, ARTHUR H. and LISTHAUG, OLA, 'Political Parties and Confidence in Government: A Comparison of Norway, Sweden and the United States', British Journal of Political Science, 20:3 (1990), 357-86. MILLER, WARREN E., Without Consent: Mass-Elite Linkages in Presidential Politics (Lexington, KY: Kentucky University Press, 1988). MILLER, WARREN E. et al.. Political Representation in Western Democracies (Oxford: Oxford University Press, forthcoming). MILLER, WARREN E. and STOKES, DONALD E., 'Constituency Influence in Congress', American Political Science Review, 57 (1963), 45-56.

References

289

MlLLON-DELSOL, CHANTAL, L'Etat subsidiaire: Inference et noninference: Le principe de suhsidiarite aux fondements de I 'histoire eurapeenne (Paris: Presses Universitaires de France, 1992). MlLLON-DELSOL, CHANTAL, Le principe de subsidiairite (Paris; Presses Universitaires de France, 1993). MlNTZEL, ALF, Multikuhurelle Gesellschaften in Eitropa und Nordamerika: Konzepte, Streitfragen, Analysen, Befimde (Passau: Wissenschaftsverlag Rothe, 1997). MlTRANY, DAVID, A Working Peace System (Chicago: Quadrangle Books, 1966). MORAVCSIK, ANDREW, 'Negotiating the Single European Act', in Robert O. Keohane and Stanley Hoffmann (eds.), The New European Community: Decision Making and Institutional Change (Boulder: Westview Press, 1991), 41-84. MORAVSCIK, ANDREW, 'Why the European Community Strengthens the Nation State', Centre for European Studies, Harvard University Working Papers Series No. 52, 1994. MORGAN, ROGER and TAME, CLARE (eds.), Parliaments and Parties: The European Parliament in the Political Life of Europe (New York: St. Martin's Press, 1996). MORLINO, LEONARDO and MONTERO, JOSE R., 'Legitimacy and Democracy in Southern Europe', in Richard Gimther, P. Nikiforos Diamandouros, and Hans-Jilrgen Puhle (eds.), The Politics of Democratic Consolidation: Southern Europe in Comparative Perspective (Baltimore, MA: The Johns Hopkins University Press, 1995), 231-60. MORLINO, LEONARDO and TARCHI, MARCO, 'The Dissatisfied Society: The Roots of Political Change in Italy*, European Journal of Political Research, 30:1 (1996), 41-63. NANNESTAD, P. and PALDAM, MARTIN, 'The v-p Function: A Survey of the Literature on Vote and Popularity Functions', Public Choice, 79(1995), 213-45. NEWMAN, MICHAEL, Democracy, Sovereignty, and the European Union (New York: St. Martin's Press, 1996). NlEDERMAYER, OSKAR, 'Trends and Contrasts', in Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and Internationalized Governance (Oxford and New York: Oxford University Press, 1995a), 53-72.

290

References

NlEDERMEYER, OSKAR, 'Trust and Sense of Community', in Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and Internationalized Governance (Oxford and New York: Oxford University Press, 1995b), 227-45, NIEDERMAYER, OSKAR, 'Die Entwicklung der offentlichen Meinung zu Europa', Beitrag zum Forschungsprojekt 'Der Wandel europapolitischer GrandverstSndnisse', Berlin, July 1997, NIEDERMAYER, OSKAR and SINNOTT, RICHARD (eds,), Public Opinion and Internationalized Governance (Oxford and New York: Oxford University Press, 1995). NIEDERMAYER, OSKAR and WESTLE, BETTINA, 'A Typology of Orientations', in Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and Internationalized Governance (Oxford and New York: Oxford University Press, 1995), 33-50. MORRIS, PlPPA, (ed.), Critical Citizens: Global Support for Democratic Governance (Oxford: Oxford University Press, 1999, forthcoming), MORRIS, PIPPA, 'The Growth of Critical Citizens?', in Pippa Norris (ed.), Critical Citizens: Global Support for Democratic Development (Oxford; Oxford University Press, 1999a, forthcoming). NUGENT, NEILL, The Government and Politics of the European Union (London: Macmillan, 1994), 3rd edition. OECD, Main Economic Indicators: July 1996 (Paris: OECD Statistics Directorate, 1996). OFFICIAL JOURNAL OF THE EC, C 303, November, 14* 1995. OLSON, MANCUR, The Rise and Decline of Nations: Economic Growth, Stagflation and Social Rigidities (New Haven: Yale University Press, 1982). OPPENHUIS, ERIK, Voting Behavior in Europe: A Comparative Analysis of Electoral Participation and Party Choice (Amsterdam: Het Spinhuis, 1995). PAGE, BENAJMIN I., Choices and Echoes in Presidential Elections: Rational Man and Electoral Democracy (Chicago: University of Chicago Press, 1978). PALMER, HARVEY D. and GABEL, MATTHEW, 'National Interest, Public Opinion, and European Integration: A Political-Economy Model of National Level Public Support', unpublished manuscript, nd, PFETSCH, FRANK R., Die Europaiscbe Union: Geschichte, Institutionen, Prozesse: Eine Einfiihrung (Mttnchen: Fink. 1997).

References

291

PIERCE, ROY, 'Mass-Elite Linkages and the Responsible Party Model of Representation', in Warren E, Miller et al., Political Representation in Western Democracies (Oxford: Oxford University Press, 1999, forthcoming), PIERSON, PAUL and LEIBFRIED, STEPHAN, 'Multitiered Institutions and the Making of Social Policy', in Stephan Leibfried and Paul Pierson (eds.), European Social Policy: Between Fragmentation and Integration (Washington DC: Brookings Institution, 1995), 1—40. PITK1N, HANNA FENICHEL, The Concept of Representation (Berkeley: University of California Press, 1967). POMPER, GERALD M., 'From Confusion to Clarity: Issues and American Voters, 1956—1968', American Political Science Review, 66 (1972), 415-28.

PRZEWORSKI, ADAM and TEUNE, HENRY, The Logic of Comparative Social Inquiry (New York: Wiley, 1970). PUTNAM, ROBERT D., 'Comment on Martin Slater, Political Elites, Popular Indifference, and Community Building*, Journal of Common Market Studies, 21 (1981), 87-9. REIF, KARLHEINZ, 'Ten Second-Order National Elections', in Karlheinz Reif (ed,), Ten European Elections: Campaigns and Results of the 1979/81 First Direct Elections to the European Parliament (Aldershot: Gower, 1985), 1-36. REIF, KARLHEINZ and SCHMITT, HERMANN, 'Nine National SecondOrder Elections: A Systematic Framework for Analysis of European Elections Results', European Journal of Political Research, 8 (1980), 3-44. REIMER, B., 'Post-Modern Structures of Feeling: Values and Life Style in the Post-Modern Age', in John R. Gibbins (ed.), Contemporary Political Culture: Politics in a Post-Modern Age (London: Sage, 1992), 110-26. RlKER, WILLIAM H., Liberalism Against Populism: A Confrontation Between the Theory of Democracy and the Theory of Social Choice (San Francisco, CA: W.H. Freedman and Company, 1982). SANI, GIACOMO and SARTORI, GIOVANNI, 'Polarization, Fragmentation, and Competition in Western Democracies', in Hans Daalder and Peter Mair (eds.), Western European Party Systems (London: Sage, 1983), 307-40.

292

References

SARIS, WlLLEM E,, Variations in Response Functions: A Source of Measurement Error in Attitude Research (Amsterdam: Sociometric Research Foundation, 1988). SARTOR!, GIOVANNI. Parties and Party Systems: A Framework for Analysis (Cambridge and New York: Cambridge University Press, 1976). SCHATTSCHNEIDER, ELMER E,, Party Government (New York: Farrar and Rinehart, 1942).

SCHILLING, THEODOR, 'Subsidiarity as a Rule and a Principle, or; Taking Subsidiarity Seriously', Harvard Jean Monnet Working Paper, 10/95, Cambridge, 1995. SCHMITT, HERMANN, 'Zwischen Homogenitat mid Polarisierung', in Jtirgen W. Falter, Christian Fenner, and Michael Th, Greven (eds.), Politische Witlensbildung und Interessenvennittlung (Opladen: Westdeutscher Verlag, 19S4), 205-14. SCHMITT, HERMANN, 'Was war "europaisch" am Europawahlverhalten der Deutschen? Eine Analyse der Europawahl 1989 in der Bundesrepublik*, in Oskar Niedermayer and Hermann Schmitt (eds.), Wahlen und Ettropaische Einigung (Opladen: Westdeutscher Verlag, 1994), 63-83. SCHMITT, HERMANN, 'National Party Systems and the Policies of the European Union. First Results from the 1994 European Elections Study', II Politico, 90 (1995), 93^107. SCHMITT, HERMANN, 'Issue-Kompetenz oder Policy-Distanz? Zwei Modelle des Einflusses politischer Streitfragen auf das Wahfverhalten und die einpirische Evidenz aus drei Nachwahlumfragen', in Max Kaase and Hans-Dieter Klingemann (eds.), Wahlen und Wdhler: Anafysen aus Anlafl der Bundestagswahl 1994 (Opladen and Wiesbaden: Westdeutscher Verlag, 1998), 145-72. SCHMITT, HERMANN and SCHEUER, ANGELIKA, 'Region—Nation— Europa: Drei Ebenen politischer Steuerung und Gestaltung in der Warhnehmung der Burger', in Thomas Konig, Elmar Rieger, and Hermann Schmitt (eds.), Das europdische Mehrebenensystem, Mannheimer Jahrbuch fur Europaische Sozialforschung, vol. 1 (Frankfurt am Main: Campus Verlag, 1996), 160-79. SCHMITT, HERMANN and TREIBER-REIF, HELGA, 'Structure in European Attitudes', Report on behalf of the Cellule de Prospective

References

293

of the Coinmsssion of the European Communities, ZEUS, Mannheim, 1990. SCHMITT, HERMANN, VAN DER EUK, GEES, PAPPI, FRANZ URBAN and SCHOLZ, ELVIRA, The European Elections Study 1994: Design, Implementation and Results, computer file and codebook (Cologne: Zentralarchiv fur Empirische Sozialforschung, 1997). SCHMlTTER, PHILIPPE C., 'Examining the Present Euro-Polity with the Help of Past Theories', in Gar)' Marks et al., Governance in the European Union (London, Thousand Oaks, CA: Sage, 1996), 1-14. SHEPHERD, ROBERT J., Public Opinion and European Integration (Lexington, MA: Lexington Books, 1975).

SHERIF, MUZAFER and HOVLAND, CARL I., Social Judgement: Assimilation and Contrast Effects in Communication and Attitude Change (New Haven and London: Yale University Press, 1961). SlNNOTT, RICHARD, 'Policy, Subsidiarity, and Legitimacy', in Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and Internationalized Governance (Oxford and New York: Oxford University Press, 1995), 246-76. STIMSON, JAMES A., 'Regression in Space and Time; A Statistical Essay', American Journal of Political Science, 29 (1985), 914^47. STOKES, DONALD E., 'Spatial Models of Party Competition', American Political Science Review, 57 (1963), 369-77. STOKES, DONALD E., 'Valence Polities', in Dennis Kavanagh (ed.), Electoral Politics (Oxford: Clarendon Press, 1992), 141-164. THE EUROPEAN COMMISSION, Eurobarometer 47, Spring 1997 (Brussels, 1997). THOMASSEN, JACQUES, Kiezers en gekozenen in een representative demokratie (Alphen aan den Rijn: Samsom, 1976). THOMASSEN, JACQUES, 'Empirical Research into Political Representation: A Critical Reappraisal', in Hans-Dieter Klingemann, Richard StoS, and Bernhard Wessels (eds.), Politische Klasse und Politische Institution-en; Probleme tind Perspektiven der Elitenforschung: Dietrich Herzog zum sechzigsten Geburtstag (Opladen: Westdeutscher Verlag, 1991), 259-74. THOMASSEN, JACQUES, 'Empirical Research into Political Representation: Failing Democracy or Failing Models?*, in M. Kent Jennings and Thomas E. Mann (eds.), Elections at Home and Abroad: Essays in Honor of Warren E, Miller (Ann Arbor: University of Michigan Press, 1994).

294

References

THOMASSEN, JACQUES, 'Support for Democratic Values', in HansDieter Klingemann and Dieter Fuehs (eds,), Citizens and the State (Oxford and New York: Oxford University Press, 1995), 383-416. THOMASSEN, JACQUES, 'Political Representation in the European Union", paper prepared for the Annual Meeting of the American Political Science Association, Boston, 1998. THOMASSEN, JACQUES and SCHMITT, HERMANN, 'Policy Representation', European Journal for Political Research, 32 (1997), 165-84,

THOMASSEN, JACQUES and SCHMITT, HERMANN, 'Partisan Structures in the European Parliament", in Richard S. Katz and Bemhard Wessels (eds.), The European Parliament, National Parliaments, and European Integration (Oxford: Oxford University Press, 1999), 129-48. TlLLIE, JEAN, Party Utility and Voting Behavior (Amsterdam: Het Spinhuis, 1995). TRAVERS, ROBERT .M.W., 'A Study in Judging the Opinions of Groups', Archives of Psychology, 37 (1941), 5-73. VAN DER BRUO, WOUTER , Where '$ the Party? Voters' Perceptions of Party Positions (Amsterdam: University of Amsterdam, 1996). VAN DER BRUG, WOUTER, 'The Informed Electorate: Political Perceptions and Party Behavior', Acta Politico, 33 (1998). 20-55. VAN DER EJJK, CEES, 'Measuring Agreement in Ordered Rating Scales', in Meindert Fennema, Gees van der Eijk and Huibert Schijf (eds.), In Search of Structure: Essays in Social Science and Methodology (Amsterdam: Het Spinhuis, 1998). VAN DER EIJK, CEES and FRANKLIN, MARK.N., 'European Community Politics and Electoral Representation: Evidence from the 1989 European Elections Study', European Journal of Political Research,

19:1 (1991), 105-27. VAN DER EIJK, CEES, FRANKLIN, MARK N., et al., Choosing Europe? The European Electorate and National Politics in the Face of Union (Ann Arbor: Michigan University Press, 1996.). VAN DER ElJK, CEES and NIEMOLLER, BROER, Electoral Change in the Netherlands: Empirical Research and Methods of Measurement (Amsterdam: CT Press, 1983). VAN DER EIJK, CEES and NIEMOLLER, BROER, 'The Netherlands', in Mark N. Franklin, Tom Mackie, and Henry Valen (eds.), Electoral Change: Responses to Evolving Social and Attitudinal Structures in

References

295

Western Countries (Cambridge: Cambridge University Press, 1992), 255^83. VAN DER KOLK, HENK, SCHMITT, HERMANN, SCHOLZ, ELVIRA, and THOMASSEN, JACQUES, 'The European Election Study 1994', European Journal for Political Research, 32 (1997), 283-9, VANDERSANDEN, GEORGES (ed.), L'Europe et les regions: Aspects juridiqites (Bruxelles: Editions de TUniversite de Bruxeltes, 1997). VAN DETH, JAN W., 'Interest in Polities', in M. Kent Jennings, Jan W. van Deth, et si,. Continuities in Political Action: A Longitudinal Study of Political Orientations in Three Western Democracies (Berlin and New York: De Gruyter, 1990), 275-312, WAHLKE, JOHN c., EULAU, HEINZ, BUCHANAN, WILLIAM and FERGUSON, LORNA C., The Legislative System: Explorations in Legislative Behavior (New York: Wiley & Sons, 1962), WARE, ALAN, Political Parties and Party Systems (Oxford: Oxford University Press, 1996). WEIL, FREDERICK D., 'The Sources and Structure of Democratic Legitimation in Western Democracies: A Consolidated Model Tested with Time Series Data in Six Countries Since World War II', American Sociological Review, 54 (1989), 682-706. WEILER, JOSEPH H.H., HALTERN, ULRICH R. and MAYER, FRANZ C., 'European Democracy and its Critique', in Jack E.S. Hayward (ed.), The Crisis of Representation in Europe (London: Frank Cass, 1995), 4-39. WEISBERG, HERBERT ¥,, The Political Psychology of Party Identification", paper prepared for the Annual Meeting of the American Political Science Association, Boston, 1998. WESSELS, BERNHARD, 'Development of Support: Diffusion or Demographic Replacement?*, in Oskar Niedermayer and Richard Sinnott (eds.), Public Opinion and Internationalized Governance (Oxford: Oxford University Press, 1995), 105-36. WESSELS, BERNHARD, 'System Characteristics Matter: Empirical Evidence from Ten Representation Studies', WZB-Discussion Paper FS III 96-206, Wissenschaftszentram Berlin, 1996. WESSELS, BERNHARD, 'Organizing Capacity of Societies and Modernity*, in Jan W. van Deth (ed.), Private Groups and Public Life: Social Participation, Voluntary Associations, and Political Involvement in Representative Democracies (London and New York: Routledge, 1997), 198-219,

296

References

WESSELS, BERNHARD, 'European Parliament and Interest Groups', in Richard S. Katz and Bernhard Wessels (eds.), The European Parliament, National Parliaments, and European Integration (Oxford; Oxford University Press, 1999a), 105-28. WESSELS, BERNHARD et al, European Political Representation Project: The Surveys Among Members of the European Parliament and Among Members of Twelve National Parliaments (codebook anddatafile) (Berlin: Wissenschaftszentrum, 1997).

WESSELS, BERNHARD and KIELHORN, ACHIM, 'Which Political Competencies for Which Political Level?', in Richard S. Katz and Bernhard Wessels (eds.), The European Parliament, National Parliaments, and European Integration (Oxford: Oxford University Press, 1999), 174-96. WESTLE, BETTINA, Politische Legitimitat: Theorien, Konzepte, empirische Befunde (Baden-Baden: Nomos Verlagsgesellschaft, 1989). WILKE, MARC and WALLACE, HELEN, 'Subsidiarity: Approaches to Power-Sharing in the European Union', Royal Institute of International Affairs, discussion paper 27, London, 1990. WILSON, GRAHAM K., Interest Groups (Oxford: Basil Blackwell, 1990). ZORN, MICHAEL, 'Ober den Staat und die Demokratie im Europaischen Mehrebenensystem', Politische Vierteljahresschrift, 37:1 (1996), 27-55.

Index 'probability to vote' question 167, 172, 182-3 Almond, Gabriel A, 12, 275 Alpert, Eugene Jf, 211, 222, 275 Althusius 69 Andersen, Svein S, 12,275 Anderson, Christopher J. 93, 94, 100,275 Andeweg, Rudy 206, 258,275 Aquinas, Thomas of 69 Aristotle 69 Austria 28,33,35,37,41,70, 77,79,81-3,98-9,114,208 Badura, Bernhard 217,275 Bardi, Luciano 35,275 Barry, Brian M. 90,275 Beck, Paul Allen 127, 278, 284 Belgium xv, 33-4, 37, 39-40, 45, 49, 68-72, 79-82, 99, 114, 127, 137, 158, 190, 215,250,270-3 Bell, Charles G, 217,275 Berelson, Bernhard 85, 236, 276,286 Bogdanor, Vernon 128,206, 258,263,264, 267, 276,286 Bosch, Agusti 32, 72, 85, 94, 276 Brand, Karl -Werner 226,276 Brown, M.C. 169,276

Buchanan, W, 211,217,295 Budge, Ian 116, 130, 154, 162,276 Bundesverfassungsgericht 26, 276 Burke, Edmund 14, 16, 210, 280 Campbell, Angus 182,276 centre—periphery 35,44, 569,69,261 CEPR, 70,276 citizenship 11, 25-6, 32, 286 Citrin, Jack 84,277 Clarke, Harold D. 84, 85, 277 Clausen, Aage R. 236,277 cleavages 28,91,284 cognitive balance theory 235 Committee of the Regions 50 Common Agricultural Policy 6,61 agricultural surpluses 68, 70,72,128,179, 184, 190-1 common currency 117,121, 124,148,197,202,204, 242, 245, 250 common market see Single European Market Conover, Paul J, 185,277 constituency xi, 202,213-6, 219-25,227-9,233,277,' 286,288

298

Converse, Philip E. 129, 161, 192, 237,276,277 Corbey, Dorothea J,M. 5, 277 Council of Ministers 4-6, 18, 75-6, 80, 82, 89,207, 259, 263-4 Cox, Gary W. 221,277 Crepaz, Markus M.L. 226, 227, 228, 287 crime 68,70,72, 117, 122-8, 179, 190-2, 195 Crozier, Michel 84,277 Cyprus 40,41,43 Dahl, Robert A. 9, 10, 61, 277 Dahrendorf, Ralf 211,278 Dalton, Russel J. 32, 74, 77, 93,94, 103, 127,278,279, 284 deBruycker, Philippe 50,278 De Winter, Lieven vi, vii, ix, xv, 47, 49, 70, 269, 271, 272,278,279 decision-making viii, 5, 7-8, 10-1, 13,26,47,50,52-9, 61-3,65-7,72,74,79-80, 83,85-6,91-2, 175,209, 222, 259,285,289 Deflem, Mathieu 91,278 Delors, Jacques 69, 127,279 democratic deficit 4, 11-12, 47, 52, 63, 66, 263 democratization 3, 65, 69, 72, 83, 284 demos 9-11,25 Denmark 32-33, 36-37, 3940,65,68-69,71,77,7982,87,97, 102, 114, 133, 134, 138, 141, 157, 164,

Index

168,174, 176, 190,204, 270-1 Deutsch, Karl W. 29, 30, 85, 278 Diamandouros, P. Nikiforos 83, 282,289 Dinan, Desmond 5,279 Dony, Marianne 50,279 Downs, Anthony 130,279 du Granrut, C, 50, 69, 279 Duff, Andrew 69,279 Dutt, Nittish 85,277 Easton, David A. 12, 13, 29, 30,61,74,75,90,261,279 econometric analysis 93 economy economic well-being 93, 101, 103 Eichenberg, Richard C. 32, 93,94, 103, 107,278,279 EIPA 70,279 Eiser, J.R, 235,280 electoral systems majoritarian 221,265, 287 proportional 221,286 electorate see voters Eliassen, Kjeli 12,275 Elling, Richard C. 211,286 employment xi, 56-9, 92, 117, 122, 125, 127, 133-4, 140, 142, 145, 149-50, 152, 160, 162, 174, 179, 192-3, 196-7, 199,201,204,238, 243-4, 246, 250-1 Endow, James M. 130,280 environment 56-9, 67-8, 70, 72,92, 117,122-3,125-6, 128, 179, 190-1,219

Index Esaiasson, Peter vii, 113,192, 207,212,237,248,271, 272,280 Estella de Noriega, Antonio 70,71,280 Estonia 40 Eulau, Heinz 14,211,213, 217,280,295 Eurobarometer v-vi, 20, 313,35,38-9,41,70,75-82, 85-8,93-4,96-8, 101, 1034-5, 107,250,269-70,280, 293 Europe Central 42,45,262 Eastern 38,40-2,45,69, 72, 83, 262 Northern 32-4,80,87, 112, 261 Southern 32-4, 44, 83, 86, 261,282,289 European Commission 6, 76, 80, 82, 89, 127, 262, 264, 293 European Community for Coal and Steel 63 European Council 6, 18,61, 71,207,263-4 European Court of Justice 6, 80,89 European Election Study v, 20-1,39,52,64, 131, 135, 139, 163-4, 166-8, 173-4, 176, 179-81, 190,269,294 European identity 11-2, 27, 276, 279 European Monetary Union see common currency

299 European Representation Study x-xi,20-l,31,37, 55,58-9,87, 119-21, 145, 152, 191, 197,224,228-9, 231,237,246-8,269 European Union benefits of membership vi, 6, 49, 62-4,66-7, 73, 85, 88,91,93-9, 101, 106-7 expansion 32-4,40, 69 length of membership 65, 101-2, 104 member states see individual countries relative power of members 230-2 Euro-scepticism 37, 50, 54, 67, 79, 204 failure of democracy see democratic deficit Farlie, Dennis J. 116, 130, 154,162,276 federal system 113, 117 Feldman, Stanley 185,277 Ferguson, L.C. 211,217,295 Finland 33,35,37,41,79-82, 98-9, 102,114,208 Fiorina, Morris P. 182,280 Flanagan, Scott C. 127, 278, 284 Flanders 128, 134, 137-8, 141, 149, 157, 164, 168, 174-6, 181 foreign policy 49, 60, 86 France 28, 33, 37, 39-40,49, 68-72, 79-83, 87, 99, 103, 106,114,128, 134,138, 141, 149, 155, 157, 164,

300 168,174,176, 190,216, 251,270-3,277,287-9 Franklin, Mark N. v, vi, x, xv, 4,20,92,94, 127, 128, 130, 137, 141, 151, 153, 158, 161,162, 169, 172, 173, 175, 178, 183, 184, 185, 258, 263, 266, 269, 280, 281,294 Fuchs, Dieter 77, 130,163 260,281,286,287,293 functionalism functionalist theory 49, 90, 92 Gabel, Matthew 93,94,281, 290 Gabriel, Oskar W. 63,281 Gaffney, John 111,281,286 Gamble, Andrew 94,282 Germany East 69,71,83, 133, 149, 155, 157, 164, 168, 1746, 178,270 reunification 34, 127 West 33,69,71, 147, 155, 157, 164, 168, 174-6, 178,270 Giscard D'Estaing, Valerie 70, 282 Golub, Jonathan 72,282 Granberg, Donald 130,160, 182,236,280,282 Great Britain see United Kingdom Greece 32-3, 35, 37, 39-40, 45,65,68-9,71-2,79-83, 86,98-9, 102-3, 114, 128, 133-4, 138,140-1, 157,

Index

164, 168, 174, 176, 178, 182,190,215,250,268, 270-1 Green, Donald P. 116,282 Gunther, Richard 83, 282,289 Haas, Ernst B. 50,282 Halaby,C.N. 169,276 Handley, David H. 84, 100, 282 Hayward, Jack E.S. 282, 295 Hegel, Georg W. F. 69 Heider, Fritz 235, 283 Herzog, Dietrich 217,283, 293 Hesse, Jens Joachim 49, 283 Hinich, Melvin J. 280 Mix, Simon 111,258,283 Hoffmann, Stanley 5,92,283, 285,289 Holland, Martin 3,283 Holmberg, Soren vii, x, xv, 11.3, 130, 182, 192,207, 212,235,236, 237,248, 271,272,280,282,283 Hooghe, Liesbeth 50,283 house effects 158, 164 Hovland, Carl L 235,293 Hrbek, Rudolf 70,284 Huber, John D. 63, 220,284 Huntington, Samuel P. 83, 84, 277,284 immigration 6, 56-9, 68, 70, 72, 128, 179,190-1 inflation 60,93, 101-2, 104, 107 Inglehart, Ronald 63, 75, 77, 84,93,162,185,230,284

Index interest group xiv, 210, 2189, 225-7,229,295-6 Ireland 32-5, 37, 39-40, 44, 65, 68-72, 77, 79-83, 86, 97,101-2,104, 114, 130, 136, 140, 143, 159, 166, 170, 176, 178, 180, 192, 217,252,270-3 issue salience 116, 142,281 issues issue agreement xiv, 246-7, 251 issue congruence x, 186, 188, 196 issue emphasis 118,120, 124, 126-7 issue-ownership 143, 162, 170, 183 position issues 117-8, 126, 130, 132-4, 136-8, 140, 143-4, 150, 162-4, 1701, 173, 175, 177, 185 valence issues 117,126, 130, 134, 136-8, 140, 143, 162-3, 170-1, 173-4, 176, 177, Italy 33, 35, 37, 39-40, 46, 49, 68-9,71-2,77,79,80-3,86, 99, 101-2, 104, 114, 128, 134, 138, 140, 157, 164, 168,174, 176, 190,250251,270-3,277,289 lyengar, Shanto 116, 282, 284 Jacobs, Francis 111,285 Jewell, Malcolm E, 210,211, 212,215,217,220,221,285 Kaltenthaler, Karl 93,94,100, 275

301

Karps, PaulD, 213,280 Katz, Richard S. vi, 12, 26, 270, 285, 294, 295, 296 Keating, Michael 49, 285 Keohane, Robert 5,285,289 Kerlinger.FredN. 185,285 Kerremans, Bart 72,285 Kielhom, Aehim vii, 71, 72, 271,272,296 Kielmannsegg, Peter Graf 12, 286 Kinder, Donald R. 116, 284 Kissinger, Henry A. 207,286 Kitschelt, Herbert 226, 286 Klingemann, Hans-Dieter 77, 130, 163,260,281,286, 287, 292, 293 Konig, Thomas 7, 286, 287, 292 Kuhnle, Stein 25, 286 Ladrech, Robert 112,286 Lazarsfeld, Paul Felix 85, 236, 276, 286 left-right dimension 119.121, 126, 130, 132, 136-7, 141, 144, 150-1, 162, 181, 196, 198,204,238,258-9 Leibfried, Stephan 6, 7,286, 291 ievel of governance 48, 51, 66 Lijphart, Arend 226, 227, 228, 265, 287 Lilli, Waldemar 34,287 Lindberg, Leon N. 31, 36, 61, 287 Lipset, Seymour Martin 84, 261,287 LISREL analysis 176

302 Listhaug, Ola 77, 287,288 Lockerbie, Brad §5,287 Long, J.Scott 73, 287 Lord, Christopher 111, 275, 283 Lourau, Rene 70,287 Lttbbe, Hermann 27,287 Luttbeg, Norman R. 19,287 Luxembourg 33-4, 36-7, 3940, 68-72, 79-82, 87, 99, 107, 122-3, 125, 128, 134, 138, 141, 147, 157, 164, " 168, 174, 176, 182, 190, 215,250,270-3 Maastricht Treaty v, 4, 17,26, 34,50,92,99, 141,188 Mackie, Tom 127, 151, 175, 183,280,294 Maillet, Pierre 70, 28§ Marks, Gary 50,260,283, 288,293 Marsh, Michael vi, vii, ix, xv, 90,94, 141,202,220,266, 269,271,272,280,281,288 Martinotti, Guide 72, 73,288 Massart-Pierard, Francoise 70, 288 McLaren, Lauren 94,280 McPhee, William M. 236,276 media 48,86, 136, 142,237 Miller, Arthur H. 84,288 Miller, Warren E, 211,237, 276, 288,290, 293 Millon-Delsol, Chantal 69, 288, 289 Mintzei, Alf 27,28,289 Mitrany, David 90,289 modernization 83,284

Index Monnet, Jean 3,292 Montero, Jose R, 83,289 Montesquieu, Charles de 69 Moravscik, Andrew 91,289 Morgan, Roger 111,289 Morlino, Leonardo 83, 289 Morocco 41,42,43 national sovereignty 60, 63, 152,202,258 nationality 45,91, 196,204 nation-building 25 Netherlands, The xv, 10,28, 33, 36-7, 39-40,44, 68-9, 71,79-82,87,99, 103, 128, 133-4, 138, 141-2, 155, 157, 164, 168, 174, 176, 182, 250-1,270-1,273,294 Newman, Michael 263, 264, 2 §9 Newton, Kenneth 32, 72, 85, 94, 276,285 Niedenmayer, Oskar 13,34, 38,61,62,63,65,70,74, 75, 77, 230,276, 279, 288, 289,290, 292, 293, 295 Niemdller, Broer 130,183, 185,294 Morris, Pippa vi, ix, xv, 74, 76, 88, 270, 278,284, 286, 288, 290 Northern Ireland 69, 72, 127, 134, 138, 157-8,164, 168, 174-6, 178,181,270 Nugent, Neil 60, 290 OECD 72, 101, 104, 107,290 Olson, Mancur 225,226, 290 opinion leaders 85-7

Index Oppenhuis, Erik 151,172, 175, 178, 183, 184, 185, 281,290 Page, Benjamin I. 142, 290 Palmer, Harvey D, 93,94, 281,290 Pampel, Fred C. 91,278 perceptions xi, xiii, 26-7, 30, 56,58,63, 132 of elites xiv, 132, 152, 155, 159,237-9,241-51,277 of voters xiv, 29, 32, 38, 62, 66, 84-5, 93, 95, 99100, 103-4, 106-7, 127, 129, 131-3, 136-7, 141-9, 151-5, 158-9, 162, 170, 181,185, 197,241,261, 267,278,281-2,294 perceptual accuracy xiv, 237,241-2,245-7,249 permissive consensus 4,31, 36,61,261 Pfetsch, Frank R. 27,28, 290 Pierson, Paul 6,7,286,291 Pitkin, Hannah 187,291 policy policy congruence xi, 198, 200-1,203,283 policy output 62,92-3, 100, 106 policy performance ix, 85, 87-90 policy preferences x, 16, 19, 113, 116, 131, 136-7, 161, 163-5, 169, 173, 179,180-2, 188, 199,257 policy satisfaction xi, xiii, 94-5, 100-6

303

policy voting 151,161,163 policy-making 4,11,13, 26, 34, 58, 60, 73, 92, 106, 112, 132,226,233, 263 political parties party choice x, xiii, 151, 161-3, 165-171, 173-6, 179-184,196,290 party cohesiveness ix, xiii, 15,29, 111, 113, 116-8, 120,124-6, 196-7,2578,266 party competence 130,132, 135-7,143,153-4,1589, 162, 173, 176-7, 179 party distinctiveness ix, xiii, 29, 111, 117-24, 126-7, 196-7,257-8 party preference xiii, 158, 163, 165-80, 182-5, 1979,206 party voters 116, 200,2I35,219, 221-4, 227-9, 238-9, 241-2, 250-1, Pomper, Gerald M. 142,291 Portugal 32-3, 35, 37, 39-40, 46, 65,68-72, 79-83, 86, 98-9, 102-3, 115, 128, 133-4, 138, 157, 164, 168, 174, 176,182, 190,250-1, 270-3 post-materialism 84,91,285 Powell, G, Bingham Jr. 220, 284 Price, Charles M, 217,275 Przeworski, Adam 92,291 Puhle, Hans-Jttrgen 83, 282, 289

304 Putnam, Robert D, 91,291 Rabier, Jacques-Rene 93, 230, 284 Reese, JOrgen 217.275 referendum 141,266,281 regions 6,17,55,57,66,69, 71, 149, 181 see also Committee of the Regions regression analysis xiii, xiv, 64,86-7, 101, 104, 144-6, 169-70, 176-9, 183-4,287, 293 Reif, Karlheinz ii, 269, 291 Reimer, B. 63,291 religion 128, 167, 175-8, 181, 185-6 responsible party model 15, 17, 19, 111, 113, 116-8, 128-9, 131, 145, 153, 155-6, 163, 165, 167, 179, 182-3, 188, 198-200,207-8,268 Riker, William H. 16,268, 291 role, representational xi, 14, 19,209-12,214-5,217, 220-2,225-7, 236, 256-7, 275,285-6 role orientation x, 209-12, 217-18,220,222,232 Roosevelt, Franklin D. 207 Sani, Giacomo 206,291 Saris, Willera E, 250,291 Sartori, Giovanni 130,206, 291 Scheingold, Stuart A, 31, 36, 61,287 Schengen Agreement 29, 70

Index Scheuer, Angelika viii, ix, xv, 25,62,65,70,71,75,292 Schilling, Theodor 50, 292 Schrnitt, Hermann vi, viii, ix, x, xv, 3, 7, 26, 32,47, 62, 70,71, 111,116, 127, 129, 136, 158, 161, 182, 186, 255,259,269,270,286, 287,291,292,294 Sehmitter, Philippe C. 50, 69, 92, 282, 293 Schneider, William 84, 287 Shapiro, Ian 116,282 Sharpe, Laurence J. 49, 283 Shepherd, Robert J. 93, 293 Sherif, Muzafer 235, 280, 293 Single European Act (SEA) 3, 5,34,92,99, 105,289 Single European Market 26, 127, 133, 140, 142, 149, 160, 184 Sinnott, Richard 12, 48, 62, 63, 276,279, 288, 289, 290, 293,295 social class 84,87, 119, 127, 165, 171, 173-6, 178-9 social judgement theory 235, 293 Soviet Empire 38-40 Spain 28, 32-3, 35, 37, 39-40, 49,65,68-9,71-2,79-83, 86,98, 102, 115, 128, 1334, 138, 140, 157, 164, 168, 174, 176, 190,251,270-3 stacked data 169-172,184 Stefanizzi, Sonia 72,288 Stimson, James A. 169,293 Stokes, Donald E. 116, 130, 211,237,276,288,293

Index subsidiarity 47-8, 50, 60-1, 66, 69-70, 276,279-80,282, 292-3, 296 surveys see European Election Study, European Representation Study Svensson, PaJIe 77,281 Sweden xv, 33, 35, 37, 41, 79, 81-2,98-9, 102, 115, 122-3, 125, 128,204,211,215, 251,271-3,280,282-3,288 Tame, Clare 111,289 Tarchi, Marco 83,289 Teune, Henry 92, 291 Thatcher, Margaret 267 Thomassen, Jacques vi, vii, viii, ix, x, xv, 3, 14, 16,26, 47,77, 111, 113, 116, 127, 129,151,161, 163,180, 186, 192,207,212,237, 255,258,259, 266,270, 271,272,293,294 Tillie, Jean 172, 182, 183, 185,294 time series 32, 36, 39, 77,93, 275-6,295 Toqueville, Alexis de 69 Travers, R.M.W, 235,250, 294 Treiber-Reif, Helga 32,292 Turkey 40-3,69,72, 184 Tiirsan, Huri 49,279 United Kingdom xi, 32-3, 35, 37,40,49,65,68-9,71-2, 79-82, 97, 101-2, 104, 190, 216 see also Northern Ireland

305

United States of America xv, 84, 282, 288 Valen, Henry 127, 151, 175, 280,294 Van der Brug, Wouter x, xv, 129, 130, 142, 154,158, 159, 161, 163, 180,185, 236, 294 Van der Eijk, Cees v, vi, x, xv, 4, 20, 124, 128, 129, 130, 137, 141, 151,153, 158, 161, 162, 169, 172, 173, 175, 178, 181,183, 184, 185,258,263,266, 269,281,292,294 Van der Kolk, Henk vi, viii, 294 Van Deth, Jan W. 70,281, 295 Vandersanden, Georges 50, 294 Verba, Sidney 12,275 voter turnout 264-5, 281, 290 voter studies v, 45, 155 see also European Election Study voters voter perceptions see perceptions of voters voter preferences 199 see also policy preferences Wahlke, John C. 211,215, 217,280,295 Wallace, Helen 70,296 Wallonia 128, 133-4, 138, 140, 157, 164, 168, 174-6, 181

306 Watanuki, Joji 84,277 Weil, Frederick D. 91,295 Weiler, Joseph H.H. 6, 7, 295 welfare state 34, 56-9, 68-9, 71-2,85,92,286,291 Wessels, Bemhard vi, vii, x, xv, 12,26,71,72,84,86, 202,209, 220, 222,229, 270,271,272,285,288, 293,294, 295, 296 Westle, Bettina 13,70,74, 290, 296 Wiberg, Matti 77,287 Wilke, Marc 70, 296 Wilson, Carole J. 260, 288 Wlezien, Christopher 92,281 Woodcock, Geoffrey 263, 264,276 World Values Study 76

Index