The Cultural Diversity of European Unity: Findings, Explanations and Reflections from the European Values Study

EVS-6-arts.qxd

14/07/2003

19:46

Page i

the cultural diversity of european unity

EVS-6-arts.qxd

14/07/2003

19:46

Page ii

EUROPEAN VALUES STUDIES

The European Values Studies is a series based on a large-scale, crossnational and longitudinal research program. The program was initiated by the European Value Systems Study Group (EVSSG) in the late 1970s, at that time an informal grouping of academics. Now, it is carried on in the setting of a foundation, using the (abbreviated) name of the group (EVS). The study group surveyed basic social, cultural, political, moral, and religious values held by the populations of ten Western European countries, getting their work into the field by 1981. Researchers from other countries joined the project, which resulted in a 26-nations data set. In 1990, the study was replicated and extended to other countries. By now, all European countries are involved in the study, including those in Central and Eastern Europe. This series is based on the survey data collected in this project.

VOLUME VI

EVS-6-arts.qxd

14/07/2003

19:46

Page iii

THE CULTURAL

DIVERSITY OF

EUROPEAN UNITY

Findings, Explanations and Reflections from the European Values Study EDITED BY

WIL ARTS, JACQUES HAGENAARS & LOEK HALMAN IN COLLABORATION WITH

WIM VAN DE DONK & TON VAN SCHAIK

BRILL LEIDEN • BOSTON 2003

EVS-6-arts.qxd

14/07/2003

19:46

Page iv

This book is printed on acid-free paper. Library of Congress Cataloging-in-Publication Data The cultural diversity of European unity : findings, explanations and reflections from the European values study / edited by Wil Arts, Jacques Hagenaars & Loek Halman ; in collaboration with Wim van de Donk & Ton van Schaik. p. cm. — (European values studies ; v. 6)

Includes bibliographical references and index.

ISBN 90-04-12299-0 (ppk. : alk. paper)

1. European Union. 2. Pluralism (Social sciences)—European Union countries. 3. Social change—European Union countries. 4. Values—Europe. 5. Civil society—Europe. 6. Group identity—Europe. 1. Arts, Wilhelmus Antonius, 1946— II. Hagenaars, Jacques A. III. Halman, Loek. IV. European values studies (Leiden, Netherlands) ; v. 6. JN30.C85 2003 300’94—dc21 2003052251

ISSN 1568-5926 ISBN 90 04 12299 0 © Copyright 2003 by Koninklijke Brill n.v., Leiden, The Netherlands

All rights reserved. No part of this publication may be reproduced, translated, stored in

a retrieval system, or transmitted in any form or by any means, electronic,

mechanical, photocopying, recording or otherwise, without prior written

permission from the publisher.

Authorization to photocopy items for internal or personal use is granted by Brill provided that

the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive,

Suite 910, Danvers, MA 01923, USA.

Fees are subject to change.

printed in the netherlands

CONTENTS Figures and tables ............................................................................... xiii

Dedication ............................................................................................xix

Series editors’ preface ........................................................................ xxi

Acknowledgements ........................................................................... xxv

Chapter One The cultural diversity of European unity:

an introduction

WIL ARTS, JACQUES HAGENAARS & LOEK HALMAN ....................... 1

1 Introduction ................................................................................. 1

2 The European Values Study as a research program .................... 3

3 The plan of this book ................................................................... 7

References ....................................................................................... 11

PART ONE EUROPEAN VALUES IN A BROAD PERSPECTIVE ................... 13

Intermezzo Godfried Danneels on individual autonomy:

‘This extreme individualism and hedonism has no future’

INTERVIEW BY MARGA VAN ZUNDERT ............................................ 15

Chapter Two Exploring Europe’s basic values map

JACQUES HAGENAARS, LOEK HALMAN & GUY MOORS ................. 23

1 Introduction ............................................................................... 23

2 Exploring value dimensions in Europe: The 1999/2000 EVS ... 25

2.1 The number of dimensions ............................................. 26

2.2 Interpretation of the basic value dimensions .................. 30

3 Comparing countries and generations in the 1999/2000 EVS ... 31

3.1 Comparing countries ....................................................... 32

3.2 Value profiles and country characteristics ...................... 36

3.3 Comparing generations ................................................... 37

4 Trends in the two basic value dimensions: 1981-1990-1999 .... 39

4.1 Cohort-period trends ....................................................... 40

4.2 Countries on the move? .................................................. 43

5 Conclusion ................................................................................. 46

References ......................................................................................... 48

Appendix The items included in the analyses ................................. 49

VI

CONTENTS

Intermezzo Ruud Lubbers on European pride and presidents: ‘A united Europe is an antidote to globalization’

INTERVIEW BY MARGA VAN ZUNDERT ............................................ 59

Chapter Three Europe and its values in an historical perspective

WIL ARTS, ARNOUD-JAN BIJSTERVELD & KAREL VERAGHTERT .... 67

1 Introduction ............................................................................... 67

2 The unity of European diversity ................................................ 70

3 The diversity of European unity ................................................ 78

4 Conclusion and discussion ........................................................ 82

References ....................................................................................... 85

Intermezzo Herman Wijffels on the vocation of Europe:

‘We are giving birth to the organically ordered society’

INTERVIEW BY MARGA VAN ZUNDERT ............................................ 87

PART TWO EUROPEAN VALUES IN DEPTH ............................................. 93

Chapter Four Opening the black-box of regional culture.

Entrepreneurial attitude and economic growth in 54 European

regions

SJOERD BEUGELSDIJK & NIELS NOORDERHAVEN ........................... 95

1 Introduction ............................................................................... 95

2 Why would entrepreneurial culture matter? .............................. 97

3 Entrepreneurial characteristics .................................................. 99

3.1 Data and method ...........................................................101

3.2 Dependent variables ......................................................101

3.3 Independent variables ...................................................102

3.4 Control variables ...........................................................104

3.5 Method ..........................................................................105

3.6 Findings ........................................................................105

4 Empirical test ...........................................................................109

5 Implications and limitations ....................................................113

6 Conclusion ...............................................................................114

References .....................................................................................115

CONTENTS

VII

Chapter Five Participation in civil society and European regional economic growth SJOERD BEUGELSDIJK & TON VAN SCHAIK ...................................119

1 Introduction .............................................................................119

2 Trust ........................................................................................121

3 Group membership ..................................................................123

4 Empirical test ...........................................................................125

5 Data .........................................................................................126

6 Trust ........................................................................................127

7 Group membership ..................................................................129

8 Economic data and testing .......................................................134

9 Implications, limitations and suggestions for further research 141

References .....................................................................................143

Chapter Six Bridging and bonding social capital: Which type is good for economic growth? SJOERD BEUGELSDIJK & SJAK SMULDERS .....................................147

1 Introduction .............................................................................147

2 Background .............................................................................149

3 The model ................................................................................152

3.1 Individuals’ static decision problem .............................152

3.2 Static equilibrium under symmetry ...............................157

3.3 A dynamic version of the model ..................................161

4 The hypotheses ........................................................................166

5 Measurement ...........................................................................166

5.1 Economic growth ..........................................................167

5.2 Bridging social capital ..................................................169

5.3 Bonding social capital and family ties ..........................169

5.4 Materialism ...................................................................170

6 Testing the model ....................................................................172

7 Conclusion and discussion ......................................................176

References .....................................................................................178

Appendix A static and a dynamic model .....................................181

VIII

CONTENTS

Chapter Seven Shall the twain ever meet? Differences and changes

in socio-economic justice norms and beliefs in Eastern and

Western Europe at the turn of the millennium

WIL ARTS, JOHN GELISSEN & RUUD LUIJKX .................................185

1 Introduction .............................................................................185

2 Research questions ..................................................................187

3 Theoretical framework ............................................................190

4 Hypotheses ..............................................................................193

5 Data and operationalization .....................................................197

6 Models .....................................................................................201

7 Findings ...................................................................................201

8 Conclusions and discussion .....................................................212

References .....................................................................................215

Chapter Eight Civil society, social trust and democratic

involvement

PAUL DEKKER, PETER ESTER & HENK VINKEN ............................217

1 More civil society, better citizens? ..........................................217

2 Measuring civil society, trust and democratic involvement ....221

2.1 Civil society ..................................................................221

2.2 Trust and democratic involvement ...............................224

3 Nations .....................................................................................226

4 Individuals ...............................................................................234

5 Participation in civil society and generational change ............238

6 Discussion ...............................................................................243

References .....................................................................................246

Appendix Indicators (%) in 20 countries .....................................249

Chapter Nine Missing links? Consensus democracy, voluntary

association and European values

WIM VAN DE DONK, FRANK HENDRIKS & RUUD LUIJKX ..............255

1 Introduction .............................................................................255

2 Patterns of democracy and social capital: Lijphart, Putnam

and the European Values Study ...............................................258

3 Missing Link I: Consensus democracy and voluntary

association ...............................................................................263

4 Missing Link II: Consensus democracies, voluntary

associations and cultural orientations ......................................265

CONTENTS

IX

5 Discussion and conclusion: The values of European

democracy ...............................................................................272

References .....................................................................................274

Chapter Ten The welfare state: Villain or hero of the piece?

WIL ARTS, LOEK HALMAN & WIM VAN OORSCHOT .....................275

1 2 3 4

Introduction .............................................................................275

The ‘moral evil effects of the welfare state’ debate ................276

Hypotheses ..............................................................................279

Data and measurements ...........................................................282

4.1 Aggregate level analyses ..............................................283

4.2 Individual level analyses ...............................................285

5 Analysis and results .................................................................286

5.1 Aggregate level analysis ...............................................286

5.2 Individual level analyses ...............................................297

6 Conclusions .............................................................................306

References .....................................................................................308

Chapter Eleven Country differences in sex-role attitudes: Cultural

and economic explanations

MATTHIJS KALMIJN .......................................................................311

1 Background and research questions ........................................311

2 Hypotheses ..............................................................................314

3 Data and measurement ............................................................316

3.1 Dependent variables ......................................................317

3.2 Independent variables ...................................................319

4 Results .....................................................................................320

4.1 Gross country differences .............................................321

4.2 Associations at the macro level ....................................324

4.3 Individual effects ..........................................................326

4.4 Contextual effects .........................................................330

4.5 Residuals .......................................................................333

5 Conclusions ............................................................................334

References .....................................................................................336

X

CONTENTS

Chapter Twelve Cross-national differences in public consent to divorce: Effects of cultural, structural and compositional factors JOHN GELISSEN ..............................................................................339

1 Introduction .............................................................................339

2 Theory and hypotheses ............................................................341

2.1 Contextual-level characteristics ....................................342

2.2 Effects of individual-level variables .............................347

3 Data, measurements, and method ............................................349

3.1 Data ...............................................................................349

3.2 Measurements ...............................................................350

3.3 Method ..........................................................................353

4 Results .....................................................................................353

5 Conclusion and discussion ......................................................366

References .....................................................................................369

Chapter Thirteen Religious and moral pluralism in contemporary Europe VEERLE DRAULANS & LOEK HALMAN ..........................................371

1 Introduction .............................................................................371

2 Theoretical notions ..................................................................375

3 Measurements and analytical strategy .....................................379

3.1 The dependent variables ...............................................379

3.2 Independent variables ...................................................380

3.3 Analytical strategy ........................................................384

4 Results .....................................................................................384

4.1 Europe’s pattern of religious and moral pluralism .......384

4.2 Understanding religious and moral pluralism in Europe .386

4.3 Generational differences ...............................................390

4.4 Trends ...........................................................................393

5 Conclusions .............................................................................396

References .....................................................................................399

CONTENTS

XI

PART THREE EUROPEAN VALUES IN REFLECTION ...............................401

Chapter Fourteen Unity produces diversity: The economics of Europe’s social capital

LANS BOVENBERG ...........................................................................403

1 Introduction ...............................................................................403

2 Economics and social capital....................................................404

2.1 The relevance of social capital........................................404

2.2 Social capital: Extrinsic or intrinsic motivation?............405

3 European economics and values................................................411

3.1 Cultural base for European economic integration ..........411

3.2 European integration as driver of cultural change ..........413

4 Methodology .............................................................................415

4.1 Measuring social capital .................................................415

4.2 Measuring the impact of social capital ...........................416

5 Conclusions ...............................................................................417

References .......................................................................................418

Chapter Fifteen European identity and interreligious dialogue

ERNST HIRSCH BALLIN ..................................................................421

1 Introduction .............................................................................421

2 A complicated relation ............................................................424

3 The contribution of Islam to European culture ........................425

4 Respect as a condition for interreligious dialogue ...................426

5 An in-depth dialogue ...............................................................429

6 Diversity requires unity ...........................................................431

References .....................................................................................431

Chapter Sixteen One European world of welfare or many: The role

of values, behavior and institutions

RUUD MUFFELS ..............................................................................433

1 Introduction .............................................................................433

2 Comparative research into the welfare state ............................436

2.1 Value research ..............................................................436

2.2 Socio-economic research ..............................................437

3 A conceptual framework .........................................................439

4 Results from socio-economic research ....................................441

5 Summary and conclusions .......................................................446

References .....................................................................................447

XII

CONTENTS

Chapter Seventeen European surrogate

EGIDIUS BERNS ..............................................................................449

1 Introduction .............................................................................449

2 Analogy ...................................................................................451

3 Gap and churchyard .................................................................454

4 Surrogate .................................................................................457

References .....................................................................................461

Chapter Eighteen He who tears down also builds up. Literature and

the debate on norms and values

JAAP GOEDEGEBUURE ....................................................................463

Chapter Nineteen Europe: A question onto itself?

DONALD LOOSE .............................................................................469

1 2 3 4

Introduction .............................................................................469

The European soul is a question onto itself .............................470

Answers that are no answers ...................................................473

One final question ...................................................................475

About the authors ..........................................................................477

Author index ..................................................................................481

Subject index .................................................................................487

FIGURES AND TABLES Figures 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 4.1 4.2 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 6.1a 6.1b 6.2 6.3 8.1 8.2

Scree plot - eigenvalues ..............................................................27

Countries’ mean scores on the two dimensions .......................... 32

Countries’ scores on ‘autonomy - socio-liberalism’ dimensions 34

Countries’ scores on ‘normative - religious’ dimension ............. 35

Net generational differences in values (1999/2000) ................... 38

Cohort differences on ‘autonomy - socio-liberalism’ ................. 41

Cohort differences on ‘normative-religious’ values ................... 42

Countries'scores on ‘autonomy - socio-liberalism’ in 1981,

1990, 1999 .................................................................................. 43

Countries’ scores on ‘autonomy - socio-liberalism’ in 1990

and 1999 ..................................................................................... 44

Countries’ mean scores on ‘normative - religious’ in 1981,

1990 and 1999 ............................................................................ 45

Countries’ mean scores on ‘normative - religious’ in 1990

and 1999 ..................................................................................... 46

Structure of the chapter ............................................................100

Structure of the empirical part of the chapter ...........................112

Map of European regions .........................................................126

Trust scores at NUTS1 level in Europe ....................................128

An overview of the different measures of social capital ...........130

Regional scores on P-Groups in Europe ...................................131

Regional scores on O-Groups in Europe ..................................132

Regional scores on A-Groups in Europe ..................................132

Regional growth 1950-1998 .....................................................135

Coefficient and bands of Trust based on recursive OLS ..........140

Coefficient and bands of A-groups based on recursive OLS ....140

Semi-reduced form of the model ..............................................159

Semi-reduced form of the model ..............................................160

Graphical representation of the theoretical model ....................167

Graphical representation of the empirical model ......................172

Civil society patterns ................................................................222

Relationships between leisure participation (horizontal axis)

and the alleged benefits of participation ...................................225

XIV

9.1 9.2

FIGURES AND TABLES

Schematic depiction of the underlying logic ............................257 Relation between Putnam Groups and Lijphart’s ExecutivesParties dimension ......................................................................265 9.3 Relation of Generalized Trust with Putnam Groups and Lijphart’s Executives-Parties dimension ...................................267 9.4 Relation of Qualified Tolerance with Putnam Groups and Lijphart’s Executives-Parties dimension ...................................268 9.5 Relation of Political Commitment with Putnam Groups and Lijphart’s Executives-Parties dimension ...................................270 9.6 Relation of Aversion of Right and Left Wing Political Extremism with Lijphart’s Federal-Unitary dimension ............270 9.7 Relation of Belonging first of all to my own region and Proud to be a citizen of one’s own state with Lijphart’s Federal-Unitary dimension .......................................................271 10.1 Hypothetical influences of welfare state type, c.q. welfare state effort, and aggregate social capital on the aggregate level of civic morality and informal social solidarity ...............281 10.2 Hypothetical influences of contextual and individual characteristics on people’s informal civic morality and social solidarity ...................................................................................282 10.3 Country means on concern with the unabled, the unemployed, and immigrants ..............................................286 10.4 Means per type on concern with the unabled, the unemployed, and immigrants ..............................................288 10.5 Variances in informal solidarity within and between welfare state types .................................................................................289 10.6 Country means on civic morality ..............................................291 10.7 Means on civic morality per welfare type ................................292 10.8 Country means on social capital ...............................................293 10.9 Mean level of social capital per welfare state type ...................294 10.10 Country means on social capital and expenditure on social protection as % of GDP ............................................................295 11.1 Traditional gender attitudes and gender differences .................323

11.2 Women' s work and sex-role attitudes .......................................324

11.3 Church attendance and sex-role attitudes .................................325

12.1 Average levels of public approval of divorce, across countries 354 12.2 Linear relationship between public approval of divorce and the divorce rate .........................................................................356 12.3 Non-linear relationship between public approval of divorce and the divorce rate ..................................................................356 12.4 Linear relationship between public approval of divorce and church attendance ..............................................................357

FIGURES AND TABLES

XV

12.5 Linear relationship between public approval of divorce and

the percentage Catholics ...........................................................357

12.6 Linear relationship between public approval of divorce and

the percentage Orthodox persons .............................................358

12.7 Public approval of divorce, across welfare regimes .................358

13.1 Percentages of Catholics, Protestants, Orthodox, and

unchurched in European countries ...........................................381

13.2 Religious and moral pluralism in Europe (standard deviations) .385

13.3 Degree of religious pluralism in four generations ....................391

13.4 Degree of pluralism with regard to personal interest morality

in four generations ....................................................................391

13.5 Degree of pluralism with regard to self-determination

morality in four generations .....................................................392

13.6 Degree of pluralism with regard to civic permissiveness in

four generations ........................................................................393

13.7 Shifts in religious pluralism ......................................................394

13.8 Shifts in pluralism with regard to personal interest morality ....395

13.9 Shifts in degree of pluralism with regard to civic

permissiveness ..........................................................................395

16.1 Capabilities and the ‘inclusive economy’ .................................440

XVI

FIGURES AND TABLES

Tables 2.1 Results of factor analyses on three levels ...................................28

2.2 Correlations between the two basic values dimensions and

some country characteristics.........................................................36

2.3 Principal components reduced set - 1999 rotated component

matrix ..........................................................................................39

4.1 Probability of being self-employed ............................................106

4.2 Component matrix ......................................................................109

4.3a Descriptive statistics...................................................................110

4.3b Correlation table .........................................................................110

4.4 Regression results. Entrepreneurial attitude and regional

economic performance, 1950-1998 ............................................111

5.1 Data for European regions..........................................................127

5.2 Descriptive statistics...................................................................133

5.3a Correlation table of social capital variables................................133

5.3b Correlation table of standard economic variables ......................133

5.4 Regression results: Trust, group memberships and regional

economic performance, 1950-1998 ..........................................137

6.1 Comparative statics, model without investment .......................160

6.2 Comparative statics, model with human capital .......................163

6.3 Descriptive statistics .................................................................171

6.4 IV-regression ............................................................................175

7.1 Random intercept regression of public support for market

justice: Eastern and Western European countries

in 1999/2000 ..............................................................................202

7.2 Random intercept regression of public support for

government intervention: Eastern and Western European

countries in 1999/2000 .............................................................206

7.3 Public support for state intervention and competition ..............210

8.1 Summary multiple regressions analyses per domain ................230

8.2 Final multiple regression analyses ............................................232

8.3 Backgrounds of leisure participation and the benefits of

leisure participation in 20 countries: Odds ratios .....................236

8.4 Proportions members/volunteers for arts/culture activities

organizations by generations in 7 countries 1981-1999 ...........240

8.5 Social trust and political involvement by culture participation

and generations in 7 countries 1981 and 1999: Odds ratios .....241

A8.1 Indicators (%) in 20 countries ..................................................250

9.1 Scores on the Putnam-group variables and on Lijphart’s

Executives-Parties and Federal-Unitary dimension for

eighteen European countries .....................................................262

FIGURES AND TABLES

XVII

9.2 Correlations between Scores on the Putnam-group variable and Lijphart’s Executives-Parties and Federal-Unitary

dimensions for eighteen European countries ............................264

10.1 Correlations (Pearson) between welfare state effort and

informal solidarity with the unabled, unemployed, and

immigrants ................................................................................290

10.2 (Partial) correlations (Pearson) of social capital with

informal solidarity and civic morality (aggregate level) ..........296

10.3 Regression of individual and context level characteristics on

informal solidarity and civic morality ......................................298

11.1 Frequency distribution of six items on sex roles in 1999 .........318

11.2 Means of independent variables ...............................................320

11.3 Deviations from means on traditional sex-role attitudes in

30 countries ..............................................................................322

11.4 Multi-level regression analyses of women’s traditional

sex-role attitudes in 30 countries ..............................................326

11.5 Multi-level regression analyses of men’s traditional sex-role

attitudes in 30 countries ............................................................327

11.6 Contextual effects on women’s traditional sex-role attitudes

in 30 countries ..........................................................................330

11.7 Contextual effects on men’s traditional sex-role attitudes

in 30 countries ..........................................................................331

11.8 Regression residuals of prediction of women’s sex-role

attitudes by country ..................................................................333

12.1 Descriptive statistics of variables .............................................352

12.2 Aggregate-level regression of public consent to divorce on

contextual characteristics ..........................................................355

12.3 Parameters from random intercept models for public consent

to divorce ..................................................................................362

13.1 Results regression analysis religious pluralism ........................386

13.2 Results regression analysis pluralism with regard to personal

interest morality ........................................................................387

13.3 Results regression analysis pluralism with regard to

self-determination morality ......................................................388

13.4 Results regression analysis pluralism with regard to civic

permissiveness ..........................................................................389

This page intentionally left blank

DEDICATION

To Jan Kerkhofs and Ruud de Moor†, the founding fathers of the European Values Study

This page intentionally left blank

SERIES EDITORS’ PREFACE This book is the sixth volume in a series on European values published by Brill Academic Publishers. The main purpose of this series is to pre­ sent and distribute the main findings from the European Values Study (EVS), a large-scale cross-national and longitudinal research project on fundamental values in Western societies. The project was initiated by Jan Kerkhofs of the Catholic University of Leuven (Belgium) and Ruud de Moor of Tilburg University (the Netherlands). At the end of the 1970s, they established the European Value Systems Study Group (EVSSG), a small group of mainly social and political scientists, aiming at empirically investigating the main fundamental value patterns of the Europeans. The group succeeded in conducting a large-scale survey in countries of the European Community (EC) plus Spain, in 1981. The study aroused interest in many other European as well as non-European countries, where colleagues and research institutions joined the project and applied the original EVS questionnaire. In this way, comparable sur­ veys became available also outside Western Europe. Apart from a large series of books on the findings for individual countries, several crossnational comparative studies were published by, e.g., Jean Stoetzel (1983), Les Valeurs du Temps Présent: Une Enquête Européenne, Paris: Presses Universitaires de France; Stephen Harding, David Phillips with Michael Fogarty (1986), Contrasting Values in Western Europe: Unity, Diversity and Change, London: MacMillan; and Loek Halman, Felix Heunks, Ruud de Moor and Harry Zanders (1987), Traditie, Secularisa­ tie en Individualisering, Tilburg: Tilburg University Press. In order to investigate changes in values, a replication study was ne­ cessary. A second wave of surveys was fielded in 1990 again in all EC countries (minus Greece), as well as in Scandinavia, some Eastern Euro­ pean countries and several countries outside Europe. Ronald Inglehart from the University of Michigan, Ann Arbor, has been very active and successful in getting the survey fielded in the early and mid nineties in countries inside and outside Europe other than those participating in the European Values Study (EVS). This worldwide project is called the World Values Surveys (WVS). For results and more information on the EVS 1990 study, we can re­ fer to many publications, for instance, David Barker, Loek Halman and Astrid Vloet (1992), The European Values Study 1981-1990 Summary Report, London: Gordon Cook Foundation; Sheena Ashford and Noel

XXII

SERIES EDITORS’ PREFACE

Timms (1992), What Europe thinks: A Study of Western European Va­ lues, Aldershot: Dartmouth; and Loek Halman and Astrid Vloet (1994), Measuring and Comparing Values in 16 Countries of the Western World, Tilburg: WORC. To monitor further changes in values and explore the dynamics in at­ titudes, beliefs and values in Europe, a third wave of surveys has been conducted in almost all European countries in 1999/2000. A source book on this 1999/2000 EVS study has been produced and published (Loek Halman (2001), The European Values Study: A Third Wave, Tilburg: EVS, WORC, Tilburg University). The international data set is deposi­ ted at Zentralarchiv für Empirische Sozialforschung an der Universität zu Köln (ZA) in Cologne and can be obtained from there (www.gesis.or/za). For more information on the European Values Study project and recent developments, we can refer to the EVS website (www.europeanvalues.nl) and for information on the World Values Sur­ veys to their website (www.worldvaluessurveys.org). This book is, as mentioned before, the sixth volume in this series. The first volume, published by Tilburg University Press, was The Individua­ lizing Society; Value Change in Europe and North America (1993; sec­ ond revised edition 1994) edited by Peter Ester, Loek Halman and Ruud de Moor. The second book Values in Western Societies (1995) was edi­ ted by Ruud de Moor, and the third book Political Value Change in Western Democracies (1996) was edited by Loek Halman and Neil Nevitte. The fourth volume, From Cold War to Cold Peace (1997), was written by Peter Ester, Loek Halman and Vladimir Rukavishnikov. It contained a comparison of Russian and European values. This book has also been published in Russian (in 1998). In 1999, a fifth volume was re­ leased on Religion in Secularizing Society, edited by Loek Halman and Ole Riis. A second printing of this book appeared in 2003 in the re­ established series on European Values Study now at Brill Academic Publishers. The contributions in the sixth volume on The Cultural Diversity of European Unity examine Europe’s values from an economic, political, social, and religious-moral point of view. The focus is not only on the current value patterns but also on the dynamics of value changes over time using the data from the three time points and all contributions deal with the relationships between values and societal structural characteris­ tics. The authors are all employed at Tilburg University and represent the diversity of scientific disciplines currently present at Tilburg Univer­ sity. The seventh volume in this series is in preparation and focuses on current value patterns in Europe from a variety of scientific viewpoints,

SERIES EDITORS’ PREFACE

XXIII

interests and analytic strategies: European Values at the Turn of the Mil­ lennium.

Wil Arts & Loek Halman

This page intentionally left blank

ACKNOWLEDGEMENTS Tilburg University is proud of its leading part in the European Values Study. When the data from the third wave (1999/2000) of this crossnational values survey came at the disposal of the national EVS teams, the Tilburg group decided to set into motion an ambitious project. Rep­ resentatives of all faculties of Tilburg University, a university special­ ized in the social and behavioral sciences in the broad sense, have been exploring in the last two years the major societal domains of values, be­ liefs and attitudes from the perspective of European unity and diversity. This bulky volume is the result of this project. It is published on the oc­ casion of the 75th anniversary of the founding of Tilburg University. This is the place to thank all contributors to this volume for providing us with highly enlightening analyses, intriguing explanations and chal­ lenging reflections about value differences and value changes in con­ temporary Europe. We are very much indebted to Wim van de Donk and Ton van Schaik for their valuable assistance in preparing this project. Without the help of many it could not have succeeded. We thank first of all the national teams and their sponsors for collecting the EVS data. We thank the data management and archiving team at Tilburg University for cleaning, merging and disseminating the unique and highly informative cross-national and longitudinal data set that we analyzed and reflected on in this book. We are grateful to the Board of the EVS foundation and the EVS steering committee for their perseverance and for their con­ tinual organizational support. There are, of course, many others to whom the editors are indebted for the production of this volume. We gratefully acknowledge financial support from the Board of Tilburg University, the Work and Organization Research Center of Tilburg University’s Faculty of Social and Behavioral Sciences and its Department of Sociology. Without Kees Boos and his indispensable expertise the layout process of this book would not have proceeded as well and as smoothly as has been the case. Tilburg May, 2003

Wil Arts, Jacques Hagenaars & Loek Halman

EVS-6-serie.qxd

14/07/2003

19:45

Page 1

EUROPEAN VALUES STUDIES ISSN 1568-5926

1. P. Ester, L. Halman & R. de Moor (eds.), The Individualizing Society. Value

Change in Europe and North America. 1994. ISBN 90 361 9993 X 2. R. de Moor (ed.), Values inWestern Societies. 1995. ISBN 90 361 9636 1 3. L. Halman & N. Nevitte (eds.), Political Value Change in Western Democracies.

1996. ISBN 90 361 9717 1 4. P. Ester, L. Halman & V. Rukavishnikov (eds.), From Cold War to Cold

Peace? A Comparative Empirical Study of Russian and Western Political

Cultures. 1997. ISBN 90 361 9737 6 5. L. Halman & O. Riis (eds.), Religion in Secularizing Society. The Europeans’

Religion at the End of the 20th Century. 1999; 2002. ISBN 90 361 9740 6 (1999), 90 041 2622 8 (2002)

6. W. Arts, J. Hagenaars, J. & L. Halman (eds.), The Cultural Diversity of European Unity. Findings, Explanations and Reflections from the European Values Study. 2003. ISBN 90 04 12299 0

CHAPTER ONE

THE CULTURAL DIVERSITY OF EUROPEAN UNITY: AN INTRODUCTION WIL ARTS, JACQUES HAGENAARS & LOEK HALMAN 1 Introduction E Pluribus Unum—Out of Many, One—is the well-known motto of the United States of America expressing a sentiment that many wish to be true for the United (or Federated) States of Europe. European politicians have often referred enviously to how the founding fathers created unity out of diversity by writing the American constitution and they have con­ tinually raised the question of how, e pluribus, Europeans can become unum. Today Americans not only rally around the flag and sing the na­ tional anthem like so many nationalities do, but also have a very earnest common allegiance to the high law of the constitution. Valéry Giscard d’Estaing, the president of the European Union’s constitutional conven­ tion in Brussels, has often compared the work of the Brussels delegates to the work of the American founding fathers. Perhaps Brussels 2003 will one-day rank alongside Philadelphia 1787. Mr. Giscard d’Estaing has shown a marvelous ability to discern consensus where others see only discord. When he and his 12-strong presidium issued the first 16 ar­ ticles of a draft constitution in February 2003, they were deluged with over 1,000 proposed amendments from all sides. A fair number of dele­ gates are demanding, for example, an explicit mention of religious va­ lues and even of God’s name in the constitution. Others, however, want to emphasize the separation of church and state. These amendments re­ veal the great dissension and the many difficulties that must be over­ come before the dream of Europe as a real constitutional unity can come true (Charlemagne, 2003). Why is it so difficult to accomplish this feat? One of the problems is that national identities are so much stronger in Europe today than they were in America in the 18th century. A constitution of the European Un­ ion could declare by law the British, the Dutch, the French, the Germans, the Italians, the Poles et cetera to be citizens of a United Europe and simply cast the Norwegians, the Russians and the Swiss out altogether.

2

CHAPTER ONE

They would, however, still be the British, the Dutch, the French, the German, the Italian, the Norwegian, the Polish, the Russian and the Swiss men and women they are. The reason why this is the case is that their national identities are an important part of the way they define themselves and are seen by others. Europeans perceive themselves as different from other Europeans not only in nationality, but also in many other important respects. Scores of languages are spoken by them and in many accents. They have a superfi­ cial diversity of appearance, and also deeper diversities of aspiration and habit and belief. They have many different religions, and some of them are atheists or agnostics. In a way, it is only the common interwoven his­ tories that make Europeans one. In 1933, the year that the national-socialists came to power in Ger­ many, the great Dutch historian Johan Huizinga went in search for the future of the European spirit. He was convinced that he could find this spirit not only in contemporary Europe, but also in the common inter­ woven histories mentioned above. He argued: ‘(...) au fond les forces qui travaillent pour ou contre cette unité de travail de l’esprit de l’Europe ont été toujours les mêmes. Nous les trouvons dans les âges passés comme nous les trouvons maintenant.’ Later he came to realize that invoking an imaginary Europe of the past where Christian harmony—and therefore every other kind of harmony—reigned was not the right way to proceed. By idealizing the past our own time can pale into insignificance. Mour­ ning over a cultural and political decline could easily overshadow a view of the future. He recognized this danger at the nadir of European history in the dark years of the Second World War. In a moving essay Defiled World; A Reflection on the Chances of Recovery of our Civilization, his last will and testament, he had to acknowledge that history did not, or would not, give him the consolation he sought in it (Hanssen, 1996). Looking back to the interwoven histories that make Europe a unity he came to the following conclusion: ‘As such, it was certainly not harmful, on the contrary a blessing for Europe that it contained so many entities and so many differences in civilization. Even if each of these peoples had strong national feelings and felt itself independent in its own sphere, this did not need to exclude completely a peaceful cohabitation of them all. (...) It is the uniformity that kills and the diversity that gives birth. To all appearances, a religious or other idealistic unison would not be neces­ sary to warrant or at least guard the harmony of a continent. A small amount of human reason and of common good will would be enough to

THE CULTURAL DIVERSITY OF EUROPEAN UNITY

3

bring about such a situation. (…) Alas, the collectivity that connects the peoples did not have both these essential properties at its disposal (…)’1. In Huizinga’s view, good will and reason had been eroded by the rise of a wrong kind of nationalism that isolates and separates people’s own identity and culture from that of others. Along with their many interes­ ting and productive differences, Europeans need to share some common background for reason and good will to be able to operate and to achieve mutual understanding, fruitful cooperation and harmonious living toge­ ther. In spite of the disenchantment engulfing him he did not allow him­ self, however, to be overwhelmed by somber moods. He was enough of an ethicist and believer to argue that the only thing left for Europeans to do is to hope and trust. 2 The European Values Study as a research program Huizinga himself continued to trust in the wise rule of God, even if in the final analysis this God for him was unknowable (Hanssen, 1996). Fear of religious and moral decay, but also hope and trust, very much in Huizinga’s line of reasoning, occupied the minds of a small group of so­ cial researchers headed by Ruud de Moor from Tilburg University and Jan Kerkhofs from Catholic University of Leuven at the end of the 1970s. They raised the question of whether Europe was at that moment still the cultural unity that it once became under the influence of Chris­ tianity. Is it true that traditional norms and values have been shaken fun­ damentally in modern times and that Christendom is no longer the cor­ nerstone of European societies and its dominant frame of reference? If this is the case, so they asked, have long-term social processes such as modernization, secularization and individualization then led to the re­ placement of religion by another value system that can keep Europe to­ gether. Or has the result of these secular processes sooner been a frag­ mentation of value patterns? To answer these questions, they decided to ______________ 1

Huizinga wrote his short paper ‘Discours sur l’avenir de l’esprit Européen’ under the auspices of the League of Nations. The essay Geschonden Wereld (Defiled World) he wrote in 1943 in voluntary internal exile in a little village De Steeg after being fired as a professor at Leyden University by the German occupation forces and after a period as a hostage in the St. Michielsgestel prisoner camp. He died before the end of the war on the first of February 1945 and did not see its publication after the liberation. Both pieces can be found in volume 7 of the complete works of Huizinga that were published from 1948­ 1953 in 9 volumes by Tjeenk Willink in Haarlem. The French citation can be found on p. 262, the English ones (originals in Dutch) can be found on pp. 584-585, authors’ transla­ tion.

4

CHAPTER ONE

conduct a survey on basic human values in all the European Community member states and started what came to be known as the European Va­ lues Study (EVS). The importance of their work and the urgency and relevance of their basic questions is proven by the fact that it did arouse such a broad interest and attracted so much extra intellectual and organi­ zational talent that within two decades the project developed into a glo­ bal project. The surveys have been repeated several times and nowadays EVS is widely accepted as one of the most important and unique data sources for the comparison of values and beliefs of people throughout the world. Because of this, it seemed more than fit to dedicate this book to the two founding fathers of European and worldwide value surveys. In the nineteen seventies, comparative research on values was rare and mainly concerned with beliefs and attitudes towards very specific is­ sues. The EVS researchers wanted to empirically study a wide variety of social and moral values, regarding a broad variety of life domains, such as work and leisure time, society and politics, religion and morality, marriage and family life. They were interested in the linkages between the value orientations identified in these different life spheres and wanted to detect the basic value patterns underlying the specific value orientations. Finally, they wanted to study the interrelationships between people’s basic value orientations and European social institutions (Har­ ding et al., 1986). Following the ideas of the founding fathers of socio­ logy, Weber and Durkheim, they believed that values are prime guide­ lines in people’s lives. Where European integration was seen and led al­ most exclusively from the political and economic points of view, they wanted to study the cultural and then especially the value dimension of European integration, being convinced that contrasting values held in the various European countries could hamper the unification of Western Europe. Values play a key role in integration processes. Economic inte­ gration leads to high volumes of transactions between nations and en­ courage greater similarities in main values. But ‘(...) similarities in main values, in turn, are conducive to greater trust between people (...). Higher levels of trust encourage greater cooperation and economic integration. And economic integration, Deutsch concludes, is conducive to greater political integration’ (Nevitte & Inglehart, 1995: 108–109; see also Deutsch, 1957). The EVS group managed to conduct a major cross-national survey in 1981 in all countries of the European Community, excluding Greece, but including Spain (at that time not yet a member state) using a standardi­ zed common questionnaire. In the subsequent years, researchers from all over the world became interested in the project and they have applied the same questionnaire in their own countries. By 1984 the survey had been

THE CULTURAL DIVERSITY OF EUROPEAN UNITY

5

conducted in more than 25 countries worldwide. As a result a unique da­ ta set became available on mass believes, attitudes and values making it possible for the first time to examine in a systematic and comparable manner the value profiles of the Western European countries (along with a few other nations). Scores of articles and books have been written u­ sing this data set, many of which concentrate on the question of how si­ milar or different European countries are and why these similarities and differences are as they are. In order to be able to investigate the dynamics of value change in Eu­ rope, a repeat survey was conducted in 1990. This second wave of the European Values Study was carried out in almost all European countries, including now also a number of Central and Eastern European countries, and, within the framework of the World Values Surveys, data in a large number of countries outside Europe were also collected in the early nine­ ties. Analyses of this international comparative and longitudinal dataset again demonstrated important cross-national similarities but also signifi­ cant cross-national variations, not only at both points in time but also in trajectories of value change. In order to further and better investigate the dynamics of value change and to empirically address the competing ideas of convergence and divergence of values in Europe, a third wave of surveys was fielded in 1999/2000. Apart from Albania, Cyprus, Moldova, Norway, Switzer­ land, and some former Yugoslavian countries and the Caucasian repub­ lics, all European countries are included in this third wave. For more in­ formation we refer to Halman (2001) and the EVS website mentioned at the end of this chapter. The latest wave of surveys has also been fielded in a large number of countries outside Europe within the framework of the World Values Surveys. Apart from some studies WVS conducted during the 1995-1997 period outside and partly inside Europe, this group managed to field surveys in 2000-2001 in representative countries of the major cultural zones of the world, including the Islamic world. More in­ formation on the World Values Surveys and the countries included can be obtained from the WVS website also mentioned at the end of this chapter (see also Inglehart, 2003). We will not try to summarize the very large amount of literature that EVS produced, but leave it at a general remark. Although the immediate reason for starting EVS was moral and political concern, the project was designed as a problem and theory oriented research program. In a later stage, however, the project became increasingly data driven. The reason for this change was twofold. First, the EVS data source proved to be a gold mine for a host of social researchers. Because of the wide range of issues addressed in the questionnaire, researchers from diverse back­

6

CHAPTER ONE

grounds and different scientific disciplines had a considerable chance to find data to their liking. This led to a proliferation of research questions addressed in data analysis. Second, the dominant theoretical perspective of EVS has for a long time been modernization theory (Arts & Halman, 2002). Although this theory has provided some stimulating insights into the similarities and differences observed among the Western European countries, it was not capable to explain all or even a great part of the variation among countries. The differences and similarities in basic val­ ues appeared to be not only a product of secular modernization trends, but also to a large extent the products of each country’s unique trajectory of social development, its historical heritage, political ideology and cul­ tural experiences and traditions (Bailey, 1992). It proved not only to be difficult to find general satisfying explanations for all specific crosssectional, but also for the longitudinal differences and similarities (De Moor, 1994). Ronald Inglehart (1997) has suggested that several amendments are necessary if value researchers want to continue working with moderniza­ tion theory. The first amendment is that humankind has entered a new historical stage, that of postmodernity or post-industrialism. This new stage is not only accompanied by new technological developments (in­ formation and communication technology) and economic changes (glo­ balization of markets, flexibility of work), but it also brings new values, particularly postmaterialistic rather than materialistic ones. The second one is that we must not focus all our attention on long-term develop­ ments, but that we also need to take into consideration short-term changes, such as the different phases of the business cycle, and short term events, such as wars and revolutions. A third amendment is that we must incorporate in modernization theory the theoretical notion of path dependency. This notion refers to the idea that although the direction of change may be common in various societies, each society develops ac­ cording to its own speed and in distinctive way reflecting a society’s social-cultural experiences and historical heritage. ‘Economic development tends to push societies in a common direction, but rather than conver­ ging, they seem to move on parallel trajectories shaped by their cultural heritage’ (Inglehart & Baker, 2000: 49). Some value researchers have heeded Inglehart’s call and have started working with a more complex version of modernization theory, whereas others have embraced other general theories, such as utility theory, or preferred to apply middle range theories to their research questions. However, they all agree that people’s attitudes and values do not occur in a vacuum. They are embedded in the broader social and historical context. These histories and contexts define the constraints and deter­

THE CULTURAL DIVERSITY OF EUROPEAN UNITY

7

mine opportunities affecting actors in their choices, priorities, and value preferences (Hechter, Opp & Wippler, 1990). People’s views, their acti­ vities, ideas, attitudes, and values are related to the institutional, eco­ nomic, and cultural structure of the society in which they live. However, the social, economic, cultural and institutional context is not only an in­ dependent variable explaining values, attitudes and priorities, it is also a dependent variable in the sense that it is shaped by dominant values, atti­ tudes, and priorities in a society. In the course of history, human actors have created and changed institutions within the societal constraints and in accordance with prevailing values that existed at the time. Once these institutions had come into being, they obtained an ‘objective’ momen­ tum of their own and influenced the actors concerned, and these pro­ cesses go on continuously. The diversity in institutional patterns can be traced to specific conditions at the time of their emergence. Accepting the idea of path dependency and the idea of nation specific values and institutions that are rooted in different historical settings and socio-economic circumstances, it seems warranted to assume that con­ temporary and future Europe will still display a variety of value orienta­ tions. On the other hand, however, within Europe the accelerated process of political and economic unification of the European Union, the rapid developments in Central and Eastern Europe since the Velvet revolution, and the imminent entry of a number of former Communist countries into the European Union, raise the question as to whether national cultures in Europe will gradually disappear and promote a kind of cultural unifica­ tion in terms of world views, values, and lifestyles. The pivotal question is therefore whether cultural unity or diversity will prevail in Europe in the near future. 3 The plan of this book This latter question underlies this book. The focus is on both the diffe­ rent and the common value patterns that Europeans hold and their devel­ opments in many of the most important life domains. Using the EVS data set, in the core analysis part of this book, Europe’s values will be examined from an economic, political, social, and religious-moral point of view. At several places ‘Europe’ will be treated not only as a collec­ tion of nations to be compared, but also as a collection of super and sub national regions in order to overcome the problems caused by the rigo­ rous and partly arbitrarily drawn national borders. The dynamic, deve­ lopmental perspective is obtained not only by comparing the data from three points in time (1981, 1990, 1999), but also where possible and relevant by comparing successive cohorts or generations. Finally, the

8

CHAPTER ONE

contributions in this book all focus one way or the other on the intrigu­ ing relationship between values on the one hand and a society’s struc­ tural characteristics on the other. Part 1 places Europe’s values in a broad perspective, from three diffe­ rent angles. From an empirical-analytical point of view, Jacques Hage­ naars, Loek Halman, and Guy Moors explore in Chapter 2 the basic di­ mensions underlying Europe’s attitudes and values in the main life do­ mains, as they follow from empirical analyses of the EVS data set. Two basic dimensions were identified, labeled the Autonomous/socio-liberal dimension and the Normative/religious one. The positions of the coun­ tries on these dimensions are indicated as well as changes in these posi­ tions over time. As argued above, value patterns in Europe cannot be understood unless one takes history into account. Wil Arts, Arnoud-Jan Bijsterveld, and Karel Veraghtert sketch some of the more recent developments in European history in Chapter 3 against the background of a more distant past, focusing on those trends that have had a lasting impact on Europe’s common value patterns and on historical-cultural fault lines within Europe. A perhaps unexpected, but very interesting point of view is offered by three interviews conducted by Marga van Zundert with three eminent Europeans: Cardinal Godfried Danneels, Arch-Bishop of MechelenBrussels, Ruud Lubbers, UN High Commissioner for Refugees and for­ mer Prime Minister of the Netherlands, and Herman Wijffels, chair of the Dutch Social and Economic Council and former CEO of the RABO Bank. Their reflections on Europe’s changing values and on their role in the world of tomorrow give the forthcoming analyses a sense of urgency doing away with the lack of commitment from which many academic discussions suffer. The in-depths analyses of European values, based on the EVS data set are presented in Part 2 and cover most of the life domains investigated in EVS. The economic point of view is the focus of Chapters four through seven. First, in Chapter 4, Sjoerd Beugelsdijk and Niels Noorderhaven con­ struct an index to measure the entrepreneurial attitude and aggregate it to characterize 54 European sub national regions. They then investigate how these regional scores are related to regional economic performance, thus linking culture and economy at the regional level. A similar (re­ gional) link is investigated in Chapter 5 by Sjoerd Beugelsdijk and Ton van Schaik, but now between economic growth and social capital in the form of generalized trust and associational activity.

THE CULTURAL DIVERSITY OF EUROPEAN UNITY

9

While the empirical correlations between cultural and economic indi­ cators form the kernel of Chapters 4 and 5, in Chapter 6, Sjoerd Beugels­ dijk and Sjaak Smulders develop an explicit model indicating the mecha­ nisms connecting social capital and economic growth. This model is tes­ ted (and modified) using the EVS data. Wil Arts, John Gelissen and Ruud Luijkx discuss in Chapter 7 the transition in Eastern Europe from a planned to a market economy. They argue that such a transition not only involves developing the proper market and legal institutions, but that it is necessary for a popular legiti­ mization of the transition to arise that people believe that this market economic system is just. In other words a sense of market justice has to be developed and the belief in socialist justice abandoned. They investi­ gate whether there are any differences in this respect between Western and Eastern Europe and if so, how these differences have developed over time. After the economic chapters, the political domain comes to the fore in the next three chapters. In Chapter 8, Paul Dekker, Peter Ester, and Henk Vinken try to answer the question: ‘More civil society, better citizens?’ They investigate the relationships between civil society, social trust and democratic involvement and how these are based on a cultural founda­ tion. Extra attention is paid to generational differences and developments to see whether indeed the young take the lead in ‘killing civic engage­ ment’ as many believe. The link between Putnam’s social capital theory and ‘good govern­ ance’, and between Lijphart’s consensus democracy and political per­ formance has been extensively studied. Wim van de Donk, Frank Hendriks, and Ruud Luijkx take in Chapter 9 an unexpected look at these relationships by studying two ‘missing links’: the connections be­ tween Putnam’s and Lijphart’s central concepts and the relationships consensus democracy and social capital have with basic value orienta­ tions, thus shedding further light on the workings of our democratic in­ stitutions. The consequences that the welfare state may have for social capital, informal solidarity, and civic morality are the topic of Chapter 10, writ­ ten by Wil Arts, Loek Halman, and Wim van Oorschot. They want to know whether, as some believe, the welfare state erodes social capital, informal solidarity and civic morality or whether social capital, informal solidarity and civic morality are still in good shape thanks to the welfare state, as others believe. They also inquire whether welfare regimes come in types and if so, whether Europe can be divided accordingly in super national regions.

10

CHAPTER ONE

Men-Women relationships form another value domain. In Chapter 11, Matthijs Kalmijn focuses on the socially defined appropriate roles of men and women and investigates two explanations of the sex-role revo­ lution that has taken place: a cultural and an economic explanation. More specifically he wants to know how the European societies differ in their attitudes about sex roles and to what extent these differences can be explained by differences in religious and labor force characteristics of individuals and societies. Chapter 12, written by John Gelissen, deals with public consent to divorce. Cross-national differences are described and explanations sought in the institutional characteristics of different welfare regimes, in the cultural characteristics and specific histories of the countries under scrutiny, as well as in individual background charac­ teristics. The final ‘in depth’ Chapter 13 lies at the heart of the original EVS problematic and is about religious and moral values. Veerle Draulans and Loek Halman discuss the decline of traditional, institutional religion and the consequences this may have (or not) for morality in society. One of the assumed consequences is an increase in religious and moral plu­ ralism. They explore the European patterns of religious and moral pluralism, across nations and over time and try to find explanations for the differences and developments found. The last part of the book, Part 3, contains reflections on the findings of EVS, on the value of value survey research, on the hidden assump­ tions behind it, on the deeper layers of Europe’s position and European developments. These essays are broad in scope and simultaneously treat their material in depth. The authors were not asked to give specific comments on concrete findings, but to reflect on one or more of the main themes in this book. The results are intriguing and gratifyingly excellent (although the authors were of course selected for their excellence and their expert knowledge in particular domains). The essays invite the reader to re-read or rather re-think the analyses in this book from a new and more profound perspective. Lans Bovenberg, in the first essay (Chapter 14) re-thinks the question of whether and if so, how cultural homogeneity is a necessary condition for economic integration. Ernst Hirsch Ballin reflects on European tolerance of people with a different belief. His essay on interreligious dialogue (Chapter 15) deals especially with Europe’s dialogue with the Muslim world. Ruud Muffels tries to develop a framework in which values and institutional differences in the socio-economic domain can be examined in an integrated fashion (Chap­ ter 16). To attain this goal he uses explanations and findings from both cross-national value research and comparative welfare regime studies. Gido Berns observes (Chapter 17) that the European countries are in

THE CULTURAL DIVERSITY OF EUROPEAN UNITY

11

many important respects destined to seek cooperation, but that the United States of Europe are still far away. He tries to find a philosophi­ cal foundation for this awkward situation, starting from Kant’s Zum ewigen Frieden. Jaap Goedegebuure (Chapter 18) reflects on the rela­ tionship between literature and moral values: only those authors who have deep knowledge of Evil may become great writers. Perhaps, and that is an implicit lesson to be learned from his essay, value studies must much more inquire into vices, than into virtues. In the final essay (Chap­ ter 19), Donald Loose launches the hypothesis that in the end EVS tries to answer the question: ‘What is a European soul?’ He paraphrases Saint Augustine’s ‘I have become a question onto myself’ into: ‘Europe: a question onto itself?’ This is a question that can only be answered by raising other questions. And what about ‘answers’ from objective, hard, solid ‘facts’ such as those provided by EVS? To quote him: ‘In that sense the European Values Study survey functions as a consolation and a reassurance: a reflecting mirror of the general answers which are still there, should we raise a question. A perfect objective Platonism for the European people in general.’ Is that what it is all about? More detailed information on the European Values Study and the World Values Survey can be found on the websites www.europeanvalues.nl, and www.worldvaluessurvey.org. References Arts W. & L. Halman 2002. Risk and Trust. Value Change in the Second Age of Modernity. Pp. 321–348 in P. Chiemelski, T. Krauze & W. Wesolowski (eds.), Kultura, Osobowosc, Polityka, Culture, Personality, Politics. Essays in Honour of Aleksandra Jasinska-Kania. Warzawa: Wydawnictwo Naukowe ‘Scholar’. Bailey, J. 1992. Social Europe: Unity and Diversity – an Introduction. Pp. 1–16 in J. Bailey (ed.), Social Europe. London & New York: Longman. Charlemagne 2003. Philadelphia or Frankfurt? The Economist, March 8th: 37. De Moor, R. 1994. Epilogue. Pp. 229–232 in P. Ester, L. Halman & R. de Moor (eds.), The In­ dividualizing Society. Tilburg: Tilburg University Press. Deutsch, K.W. 1957. Political Community and the North Atlantic Area. Princeton: Princeton University Press. Halman, L. 2001. The European Values Study: A Third Wave. Sourcebook of the 1999/2000 European Values Study surveys. Tilburg: EVS, WORC, Tilburg University. Hanssen, L. 1996. Huizinga en de troost van de geschiedenis.Tilburg University: Doctoral Dis­ sertation. Harding, S., D. Phillips & M. Fogarty 1986. Contrasting Values in Western Europe. London: MacMillan. Hechter, M., K.D. Opp & R. Wippler 1990. Introduction. Pp. 1–9 in M. Hechter, K.D. Opp & R. Wippler (eds.), Social Institutions. Their Emergence, Maintenance and Effects. Berlin: Walter de Gruyter. Huizinga. J. 1950. Verzameld Werk, Vol. 7. Haarlem: Tjeenk Willink. Inglehart, R. 1977. The Silent Revolution. Princeton: Princeton University Press. —— 1990. Culture Shift in Advanced Industrial Society. Princeton: Princeton University Press.

12

CHAPTER ONE

—— 1997. Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies. Princeton: Princeton University Press. —— (ed.) 2003. Human Values and Social Change. Findings from the Values Surveys. Leiden & Boston: Brill. —— & W.E. Baker 2000. Modernization, Cultural Change, and the Persistence of Traditional Values. American Sociological Review 65: 19–51. Nevitte, N. & R. Inglehart 1995. North American Value Change and Integration: Lessons from Western Europe? Pp. 107–135 in R. de Moor (ed.), Values in Western Societies. Tilburg: Tilburg University Press.

PART ONE

EUROPEAN VALUES IN A BROAD PERSPECTIVE

This page intentionally left blank

INTERMEZZO

GODFRIED DANNEELS ON INDIVIDUAL AUTONOMY: ‘THIS EXTREME INDIVIDUALISM AND HEDONISM HAS NO FUTURE’ INTERVIEW BY MARGA VAN ZUNDERT

Our European values can all be traced back to the age-old, JudeoChristian principle of human dignity, says Cardinal Godfried Danneels. Still, today’s European culture is following a downward curve. Not be­ cause of secularization—that is essentially a good thing—but we live too fast. ‘Nothing sinks in; nothing internalizes. There is no time to antici­ pate, no time for highlights and no time to enjoy memories.’ Godfried Danneels is a man of the church: he is Archbishop of Mechelen-Brussels and chairman of the Belgium Bishops Conference. He is also a man of science: he was professor of theology for almost twenty years. Danneels believes that the confrontation between reason and be­ lief is inevitable in our modern society, and that it is a good thing since a free, conscious choice to believe is to be preferred. Still, the archbishop is not optimistic about Europe’s culture, which is demonstrating a downward trend. Each individual is living in his own empire. But hu­ manity will regenerate: ‘One always sees a countermovement when mankind approaches the borders of self-destruction.’

Cardinal Danneels, do you consider yourself a European? I am a European, by birth, by education and by total ambiance; I’m to­ tally embedded in Europe. And I also feel European; of course, I’m Fle­ mish, but I never have the urge to underline that fact, I speak as much French as Dutch. To me, participating in Europe’s multiple cultures is something that goes without saying and is a great enrichment in my life. I regret knowing so little about the European cultures on the other side of the Danube. Unfortunately, I also lack the key to these cultures, as I don’t speak any Slavonic languages. If someone had asked me thirty years ago: ‘What is Europe?’ I would probably have answered: ‘All the countries between the Atlantic coast and Vienna.’ I wouldn’t dare say that today; in that sense I have evolved over the past thirty years. What is Europe anyway? Isn’t it just a peninsula, an appendix, of Russia? Take

16

INTERMEZZO

Dostoevski away from the European culture: it would be amputated, hobbling on one leg. If only I were a bit younger, I would try to learn Russian. As far as that is concerned, I envy the Pope, who is familiar with both Western and Eastern Europe. How would you describe our European values? European values have their origin in the Judeo-Christian tradition. There may have been influences from outside, from the Arabs, from Slavic people, from Greek philosophers and the Romans, but basically Euro­ pean values have a Biblical foundation. They are the result of a historic miracle, the revelation, and as such they are an inexplicable, historical phenomenon. Asking whether we would have had the same set of values without Judaism or Christianity is asking Why is the grass green? or Why is the sea salty? We cannot say more other than that the values are here. During the French Revolution the Judeo-Christian values were secu­ larized; the umbilical cord that connected the value set with a transcen­ dent God was cut through, resulting in moral values based on the human being itself, which was popularized in the expression: freedom, equality and brotherhood. I would summarize European values in one principle: the principle of human dignity. Europeans put the human being first; a man cannot be subordinated to any power, not to the military and not to economic or political power. The human being is no function of anything else. On the contrary, everything stands in function of the human being. ‘Man is the master of the universe’ and ‘lord of creation’ are typical Biblical expres­ sions. From that fundamental religious principle of human dignity all other values can be deduced: respect for life, brotherhood, et cetera. Are today’s European values at odds with the Catholic morality? There is no continuous duel going on, but there are certainly friction points. These all center around a difference in opinion on the autonomy of the human being. The church sees the human being as a creature with relative autonomy. The human being is free and independent, firmly standing on his own two feet. However, he is also dependent on his source of existence, a transcendent God. This is a complex paradox of heteronymous autonomy. Most secularized value systems reason that a human being has total autonomy; an individual doesn’t have to give account to anything or anyone, except to himself and—to a certain extent—to the society he lives in. This difference between relative and total autonomy is the main

GODFRIED DANNEELS ON INDIVIDUAL AUTONOMY

17

friction point between the Catholic Church and today’s society, all con­ troversies can be traced back to this disagreement. Take, for example, the question: Can a human being decide about his own life? Or in more concrete terms: Can an individual carry out euthanasia or abortion when there are serious reasons to do so? The autonomous morality would say: yes; the church says: no. Apart from a responsibility towards myself and society, I feel a re­ sponsibility towards God. Often people reply that in my view men are not autonomous, that they are deprived of true freedom; deep down they are shackled, slaves of God. I always answer this criticism with the fol­ lowing comparison: When you lay a diamond in the sun, it doesn’t say: ‘Please, sun don’t shine on me because you will crush me.’ It will say: ‘Please, sun shine on me, because the more you shine, the more I twin­ kle, and the more a diamond I am.’ But people have difficulties belie­ ving that the greater God is, the greater they are. They presume a kind of competition: the bigger God is, the smaller they must be. Are European values changing? I don’t think so, the large changes took place in the seventeenth century when the concept of total autonomy of the human being was formulated. Today people merely draw out the further consequences of this opinion. The choice of with or without God has more and more impact today be­ cause of the continuous developments in science and medicine. Take for instance that Italian who claims to be cloning humans. To me he is just a wizard’s pupil; no serious scientist would act like he does considering the current knowledge. I notice some opposition to the principal of total autonomy, predomi­ nantly from religious people, but on the other hand from a number of non-believers as well. They also find these manipulations a bridge too far. And also the recent longing for homeopathy, the back-to-nature phi­ losophy, are signs of protests against total autonomy. However, the op­ position is small; the protests are only murmuring waves on the surface, they are certainly not the tidal wave that can turn the stream. Still, one always sees a countermovement when mankind approaches the borders of self-destruction. Consider, for example, the nuclear arms race. Twenty-five years ago nobody seriously considered diminishing the ar­ maments, whereas both sides are reducing their stockpiles without much fuss now.

18

INTERMEZZO

You once compared today’s Europe with the Roman Empire prior to its fall. Do you still believe our civilization will end soon? I haven’t said that the European civilization will end; one would have to be a prophet to say that. But there are a striking number of similarities: the bread-and-circus mentality, an overwhelming fatigue, a lack of ideal­ ism and public spirit, and an excess of debates in parliament. The senate in Rome was debating day and night, meanwhile the Barbarians where conquering one fortress after the other on the borders. Another similarity is the overheated emphasis on eroticism and sexuality. People behave as if they were one-day flies that buzz feverishly and frantically in their last hours. We live too fast, there is no time to anticipate, no time for high­ lights and no time to enjoy memories. Nothing sinks in, nothing internal­ izes. I think that the Western European culture has the characteristics of a downward curve; this extreme individualism and hedonism has no fu­ ture. But nobody can predict how far and how fast it will collapse. And the Roman Empire fell, but not humanity. Humanity regenerates at a cer­ tain point, it always does. There will be another time, a time in which we are masters of the communication and mobility technology that control our lives today. We may have achieved some things in economic terms and gained personal freedom, but what have we accomplished as far as giving meaning to our lives? Belgium has the highest suicide rates in Europe. Young people leave little, heart breaking notes behind: ‘Mum and dad, you have given me everything, but I could never comprehend what my life was about.’ There are large differences in secularization within Europe. In Poland eighty percent of the citizens visit church regularly, in the Netherlands only four percent. Do you have an explanation for these figures? In Poland, all Catholics attend Sunday’s service; believing and praxis are closely tied. Almost all Flemish people have been brought up in that same kind of catholic tradition, yet only about fourteen percent attend the services weekly. In Western Europe many Christians have developed a kind of warehouse mentality. Like a bee that picks and chooses the most brightly flourishing flowers, they only select the religious traditions that are to their liking. I don’t have a good explanation for the differences, but I know that the test of secularization will happen everywhere, also in Poland, and also in Africa. The confrontation between reason and belief is inevitable, because if God’s existence were obvious, there would be no need for faith.

GODFRIED DANNEELS ON INDIVIDUAL AUTONOMY

19

Secularization cannot be stopped, and must not be stopped. I’m also secularized. When I’m ill, I may say a rosary, but I certainly will go to a doctor. In essence secularization is a legitimate event, because individual autonomy is progress. However, I don’t agree with the often-drawn con­ clusion that individual autonomy leaves no room for God. Secularization will happen everywhere, you say. What about in Islam? In Islam there is monolithicism: language, culture, religion, military po­ wer, politics and law – all are one and the same. I have grave doubts that this monolithicism can last in a civilized society. It also existed in Chris­ tianity in the Middle Ages: cuius regio, illius et religio – Claudius bap­ tized, all citizens of Gallia baptized. The French Revolution taught us that church and state can be separate. I can speak Arabic, be a supporter of Marx’s economic theories, participate in the Catholic Church and en­ joy Dostoevski. You say that secularization is in essence a good thing, but it reduces the number of Catholics dramatically. Secularization forces one to make a very conscious choice about one’s beliefs. One is no longer a Catholic by birth. That is good. I think that a moral choice is more perfect when it is made by a free man. And I pre­ pare myself for a smaller church because I know very well that a deci­ sion made freely is always more demanding than being one of the herd. But don’t be mistaken; being an atheist is as difficult as being a Christian. The atheist also has to make a conscious decision: the decision that God doesn’t exists. He stands on no firmer grounds than I; there are no compelling arguments for his choice, nor for mine. Atheists always are busy disproving my arguments, but they don’t come forward with their own arguments, because they simply haven’t got any. In practice, most people don’t know whether God exists or not. They don’t make a decision and live as agnostics, mainly because they have other things on their minds. However, being an agnostic is a makeshift solution; at some time there will come the moment to decide. Women are more religious than men in all European countries. Do you have an explanation for this fact? That is a highly intriguing matter. But I only have hypotheses about it, no explanation. I think that women are more sensitive, more perceptive to non-material matters such as religion, art or love. In that sense, they have a thinner skin. They are more connected with or embedded in the cosmos, in nature and life, perhaps also because the woman carries the

20

INTERMEZZO

child. It has always captivated me that most revelations in the Bible are made to women, and that the Pythia of Delphi were also women. But who am I to talk about this subject? I’m not an expert in the psychology of women. Although the women are more religious, the church is a man’s world. The church may be a man in composition, but in itself she is a woman. The church is never called our father, always our mother. Today the ac­ tual power structure in the church is male, but it shouldn’t have to be that way. It is just that government in the church has long been closely linked with the priesthood. But I think that priest structure and power structure in principle don’t need to be one and the same. Both my vicars are women, and I see no reason why a woman should not lead a Roman congregation. I don’t see that happening, a woman in charge in Rome? Maybe not in Rome, but Rome is not the whole church. We, the peri­ phery, are also part of the Catholic Church. Rome has always been con­ servative, but so is Brussels or The Hague. The EVS studies show a very poor trust in authorities. How do you feel about that? People don’t trust anybody except themselves anymore. That is the re­ sult of the individualism of the past thirty, forty years. Every human has become an empire in himself. But the lack of trust is also a counter reac­ tion to the fact that the government and other authorities decide upon more and more issues of daily life. Every day a new regulation is an­ nounced in Flanders, and in such an authoritative way that it inspires no confidence: you cannot sell chocolate cigarettes because that would promote smoking; your car’s front number plate has to be reflective, while the police always flash from behind; you may not advertise to­ bacco except, of course, at the racing circuit of Francorchamps since the financial consequences of not advertising there would be too great; you may not use your mobile phone in your car, but you may type in a num­ ber to Singapore on your hands-free telephone (…). I can understand why people want to go live on a Caribbean island. People want to have something that is theirs to decide. That also explains the fierce opposi­ tion against the sexual morality of the Catholic Church. The bed is al­ most the only place where one can still hold absolute sway.

GODFRIED DANNEELS ON INDIVIDUAL AUTONOMY

21

Mgr. dr. Godfried Danneels (1933) is Cardinal-Archbishop of Mechelen-Brussels and chairman of the Belgium Bishops Conference. Godfried Danneels was born in the West-Flanders countryside, the eldest of six children. He studied at the seminary in Bruges, received a degree in philosophy from the Catholic University Leuven and a doctoral degree in theology from the Gregoriana in Rome. On 17 August 1957 he was ordained as a priest. From 1959 to 1977 he was professor of liturgy at the seminary in Bruges, and professor of theology at Leuven University. In 1977, Danneels was ordained as Bishop of Antwerp; three years later he became Archbishop of Mechelen-Brussels. Danneels is a member of several Roman congregations (e.g., education and eastern churches) and for almost ten years Danneels was chairman of Pax Christi. His open at­ titude towards the media has made him a well-known person in Belgium and the Netherlands. In 2002 he received an honorary doctorate from Tilburg University.

This page intentionally left blank

CHAPTER TWO

EXPLORING EUROPE’S BASIC VALUES MAP JACQUES HAGENAARS, LOEK HALMAN & GUY MOORS 1 Introduction The European Values Study (EVS) is the most comprehensive data source on basic values held by Europeans. It is a unique data source, not only because a great number of European countries were included in the 1999/2000 survey (33 in total), but also because many of those countries were also represented in two previous waves of investigation in 1990 and 1981. The comparability between countries seems assured because the questionnaire used in the countries is the same to a very large extent. Furthermore, there is a fair amount of overlap between the question­ naires used in the three waves of surveys1. A unique feature of EVS is also the wide range of orientations, preferences, and attitudes in the most important life domains: religion, morality, family and marriage, politics, economics, work, and leisure time. The data in each of these separate life domains has been investigated extensively to identify the relevant under­ lying value dimensions. Individual differences and similarities on these dimensions have been further explored by relating them to (background) characteristics of individuals as well as to the scores on other value di­ mensions in the other life spheres. At the aggregate level, countries or regions variations have been examined, using aggregate (average) scores on the relevant indices and variables. (See Halman et al., 1987; Halman, 1991; Ester, Halman & De Moor, 1993, 1994; Halman & Vloet, 1994; the chapters in this book; the many publications mentioned on the EVS website www.europeanvalues.nl.) The main focus of this chapter is somewhat different. We are not so much concerned with the distinctive orientations in the various life ______________ 1

This is not the place to discuss issues of comparability in cross-national or crosstemporal research, but the authors share the opinion that identity of question wording does not guarantee comparability of outcomes: the same item may have very different mean­ ings in different cultures or periods.

24

CHAPTER TWO

spheres, but sooner aim at uncovering the pattern beyond these orienta­ tions. In other words, we want to answer the question whether and how the dimensions, indices, and basic values ascertained so far in each of the distinctive life domains (plus a few indices constructed for the pur­ poses of this chapter) might be reduced to a limited number of even more fundamental, underlying value dimensions. It implies that more and more we will abstract from the concrete contents of the separate questions and items. In the end we hope to achieve a kind of cultural map of Europe that captures the diversities and similarities in value pro­ files of European publics. In cross-national or cross-cultural values research there have been several important and highly influential attempts to identify the most ba­ sic value dimensions, e.g., Hofstede (1980, 2001), Schwartz (1994), Tri­ andis (2000, 2001), Inglehart (1990, 1997), Inglehart & Baker (2000). We will not try to compare the results of our exploratory analyses syste­ matically with their (differing) findings. A more refined elaborate evaluation of the (dis)similarities is outside the scope of this chapter, be­ cause our present and their previous results depend on the particular countries chosen, the specific items, variables, and domains included in the analysis, the sample actually investigated, and on many other (statis­ tical) issues, all of which are rather different in the various studies. The exploratory analyses in this chapter have three important fea­ tures. First, we try to identify a meaningful pattern in the indices distin­ guished in the distinctive life spheres included in EVS and thus cover a wider range of orientations than ever before2. Secondly, the basic value dimensions are investigated both at the aggregate and the individual level. It is a regular practice in analyses of these kinds to compare country-level scores by employing some average of the individual scores and then analyze the correlations between these average scores. Of course, there is nothing wrong with such an approach, as long as one realizes that not the individual but the country is the research unit, and that con­ clusions based on the correlations at the aggregate, country level do not automatically apply to the individual level scores (and vice versa). For ______________ 2

When investigating the contents from these life domains, we do not use the original separate questions and variables, but indices, combining several aspects of a particular domain. If one tries to find patterns in a large set of separate items, many methodological artifacts may strongly influence or bias the results, such as particular items clustering to­ gether not because they refer to the same content, but simply because they are located close together in the questionnaire, or because they have the same question format, or be­ cause the same ‘accidental’ words occur in their formulation, etc. Working with indices at least partially overcomes these problems.

EXPLORING EUROPE’S BASIC VALUES MAP

25

example, it may well be that there is no or a very low positive correla­ tion at the individual level within countries between one’s income and one’s overall satisfaction with life, but at the same time a high positive correlation between the gross national product per capita (or mean in­ come) of a country and its average level of life satisfaction. Or to men­ tion another famous (or notorious) example: large aggregate differences between religious and nonreligious countries (using the percentage of regular church attendees in the country as an indicator) with regard to the average attitude towards abortion does not exclude the possibility that there are no individual within country differences with regard to abortion between church attendees and non-attendees. Although often implicitly, too easily have conclusions drawn from findings at the aggre­ gate level been applied to the individual level. In order to prevent such unwarranted conclusions, the relationships among the indices obtained from our aggregate analyses will be compared explicitly with the rela­ tionships obtained from analyses at the individual level. Finally, we in­ vestigate whether the positions of the countries on the fundamental value dimensions have changed over time. To this purpose, the data sets of the three survey waves, covering a range of 20 years, will be used. In these analyses we also use a generational perspective, as advocated in this re­ search area by, among others, Inglehart (1990, 1997). 2 Exploring value dimensions in Europe: The 1999/2000 EVS Starting point of our analyses is the 1999/2000 data as this data contains the largest number of participating countries yielding the most encom­ passing and complete picture of Europe. Forty attitudinal scales have been constructed in the various life domains; they are presented in Ap­ pendix A. In principle, each individual has a score on all indices, and mean scores per index per country have been computed. Looking for the basic dimensions that underlie this set of 40 indices means looking for clusters of variables that are (strongly) related to one another. It is as­ sumed that high correlations between indices reflect a common value dimension. There are several statistical tools for identifying such ‘com­ munalities between indices’. Here, we will use the one most commonly employed in this area of research, viz. principal components analysis, a variant of (orthogonal) factor analysis. This technique linearly trans­

26

CHAPTER TWO

forms a set of variables into a smaller set of variables (principal compo­ nents) that represents most of the information of the original set3. The principal components analysis will be applied in the first instance to all 40 indices, and that at three levels of analysis. As said above, at the individual level, the correlations among the individual scores are ana­ lyzed. Further, in the aggregate country-level analysis, the correlations among the country mean scores are used. And finally, because of our in­ terest in generational analysis, the third level concerns an age group by country analysis. Six identical, (roughly) 10-years age groups have been defined for each country, resulting in (6 x 33 =) 198 age-country groups. For each of these 198 groups mean scores on all indices have been com­ puted and correlated. 2.1 The number of dimensions The very first question that has to be answered when one is interested in the comparability of the results of the analyses from the three different levels is whether or not the same number of fundamental dimensions, i.e., principal components, underlies the scores on the 40 indices. To de­ termine the number of dimensions implied by the data, i.e., how many principal components should be extracted, several criteria are available. In this context, the most useful criterion is based on Cattell’s scree plot of the eigenvalues: one looks for the point on the (monotonically declin­ ing) line connecting the plotted eigenvalues where the slope of the line changes from large to small4. The part of the line connecting the com­ ______________ 3

Applying different statistical techniques to the data results in somewhat different op­ erational definitions of the concept ‘underlying dimensions’ or ‘fundamental variables’. In principle, besides principal components analysis, proper factor analyses can be carried out or other latent structure models like (ordered) latent class analysis or latent trait models. Very often the results of all these techniques yield very similar results, but sometimes the results will be different. It might be interesting to further investigate whether this is the case here (or in similar investigations where almost exclusively principal components analysis has been used). 4 The criteria for determining the appropriate number of principal components (or fac­ tors in factor analysis) are based on the eigenvalues of the components where an eigen­ value reflects the amount of variation in all the original variables (indices) that is ex­ plained by a particular principal component. An often used criterion (e.g., the default cri­ s classical criterion to retain principal components with an ei­ terion in SPSS) is Kaiser' genvalue equal or greater than 1. However, this criterion often leads to overfitting and to choosing too many principal components; moreover, in the context of this study it will almost necessarily lead to the conclusion that aggregate data need fewer principal compo­ nents than individual data. Aggregate data generally produce stronger correlations than individual level data and the larger the intercorrelations among the original variables, the larger the eigenvalues of the first principal components. Inspection of the scree plot usu­ ally leads to choosing less principal components (factors) and to components that appear

EXPLORING EUROPE’S BASIC VALUES MAP

27

paratively large components representing many indices will be steep and the part connecting singular components that represent single unique parts of the indices will gradually trail off. The three scree plots (one for each level) are presented in Figure 2.1. 11 10 9 8 7 6 5 4 3 2 1 0 1

2

3

4

in d iv id u a l

5

6

c o u n try

7

8

9

10

c o u n try -a g e

Figure 2.1 Scree plot - eigenvalues

It can be concluded from Figure 2.1 that for all three levels of analysis a two components solution has to be chosen. There is a clear drop in the sizes of the eigenvalues from the first to the second and from the second ______________

to be robust when cross-validating the outcomes (much more so than when using Kaiser’s criterion).

28

CHAPTER TWO

to the third component and afterwards they clearly trail off. Hence, it was decided to apply (varimax) rotation to the two-component solution in order to obtain a simple structure solution, i.e., a solution in which the variables (indices) are assigned as ‘belonging’ as much as possible to just one of the two components. Then, the important question arises whether the two rotated components have the same meanings for all three levels. Because the meaning of a component is defined by means of the relationships between the component and the variables (indices), the component (factor) loadings representing the correlation between a component and a variable must be inspected. Table 2.1

Results of factor analyses on three levels Unit of analysis Individual Country Countryage group Component Component Component

Authority above autonomy Intolerance towards disturbing groups NOT protest prone Intolerance towards ethnic groups Work ethos Material conditions important for marriage Working woman is bad for child Less emphasis on material aspects in life is good Readiness to sacrifice for the environment Working women is ok Postmaterialism Children and marriage NOT necessary In favor of democracy Personal-sexual permissiveness

1 2 –.61 –.56 –.48 –.44 –.43 .30 –.38 –.31 .30 .36 .26 .41 –.20 .49 .50 .56 .23 .59 –.43

1 –.78 –.87 –.74 –.65 –.72 –.72 –.44 .67 .39 .67 .79 .83 .82 .69

2 1 2 .37 –.84 –.85 –.78 –.66 –.77 –.65 .42 –.51 .33 .38 .65 .38 .45 .68 .80 .82 .75 .39 –.59 .79 –.48

Self-interest permissiveness Legal-illegal permissiveness Abortion approved Expressive work qualities Confidence in international organizations Solidarity Cultural homogeneity important for marriage Emphasis on authorities is good

–.53 –.53 –.59 –.42 –.34 –.41 .25 –.32 –.76 –.72 .34 .52 .44 .38 .35 .37 .40 .41 .43 –.32 .40 –.30 .80 –.46 .71 .42 –.31 .69 –.43 .59

EXPLORING EUROPE’S BASIC VALUES MAP

29

Table 2.1 (continued) Emotional qualities important for marriage Mutual bond important for marriage Religiosity Confidence in legitimacy institutions Confidence in authority institutions State responsibility above individual responsibility Europe (feeling of belongingness to Europe) Women should NOT have job and NOT contribute to income Unlucky is reason for poverty Acceptance of injustice in society Progress is reason for poverty World (feeling of belongingness to world) Area (feeling of belongingness to region) Laziness is reason for poverty Town (feeling of belongingness to town) Intolerance towards extremists Injustice is reason for poverty Nation (feeling of belongingness to nation) Explained variance (after varimax rotation) =

–.37

.44 .45 .50 .52 –.33

.78 .51 .38 .75 –.49

–.32 –.23

.48 .71 .51 –.43 –.69

.35 .53 .46 –.59 .43 –.56 –.38

.26

.35 .48

.48 .36

.52 .58

.32 .58 –.49 –.34 .40 .58

–.28

9.9 7.8 25.3 16.1

.66 .57 .65

.42 .32 –.51 –.45 .40 .49

–.41

26

16

The rotated component matrix of the three analyses is presented in Table 2.1. The rows of this matrix are arranged in such a way that the first group of variables primarily relates to the first principal component for all three levels of analysis, the second set identifies the second compo­ nent and the third set of issues is not as consistently as the other two groups related to the two dimensions for all three levels for various rea­ sons. The last three items of the third group do not belong to any of the two dimensions and could be dropped for that reason. Then, for many items of the third group, the differences are between the two aggregate solutions and the individual one, where the latter solution has much lower loadings than the former. Given the common fact that individual level analyses yield lower correlations (and consequently component loadings) these outcomes are not really surprising. Although one should be cautious to interpret low loadings, the signs of the individual (not re­ ported here) and aggregate analyses are in agreement with each other, revealing at least a partial consistency. For some of the items of the third

30

CHAPTER TWO

group, however, the main difference is that a particular item is more strongly related to the first (second) than the second (first) component at one level of analysis while at another level the results are just the oppo­ site. Nevertheless, the signs of the loadings are fairly consistent. Irres­ pective of this, we decided to work only with the first two groups of items to avoid as much as possible ecological fallacies (where an addi­ tional considerations was that the substantive interpretations of the two dimensions are not different with or without the third set of items). The analysis was rerun based on only the first 25 items in Table 2.1. The out­ comes were almost identical to the ones reported in Table 2.1 and will not be reported separately. The patterns of loadings of the final analysis are highly similar for the individual and macro level analysis. For example, the loadings of the items on the first factor at the individual level correlate strongly with the loadings at the aggregate (country-age group) level: product-moment correlation r = .988, Spearman’s rank correlation rs = .943 and similar re­ sults are obtained for the second factor: r = .942, rs = .729. From all this, it follows that country comparisons on individual-level estimates of the principal components will be similar to the one’s based on aggregate data. 2.2 Interpretation of the basic value dimensions The first dimension in Table 2.1 appears to reflect personal autonomy. It echoes the attitudes of the USA ‘liberals’ or the European continental ‘social liberals’. Individual freedom and personal autonomy are key cha­ racteristics of this dimension, but they are linked with non- or postmate­ rialistic and pro-democratic views. Countries or individuals that rank high on this dimension favor personal autonomy more than authority. They are also more permissive regarding personal sexual matters. Fur­ ther, the ‘high-scorers’ do not regard children and marriage an absolute necessity, which seems logical as it is up to the individual to auto­ nomously develop her or himself and to decide about children and mar­ riage. The importance of individual development also appears from the high acceptance of working women, which is considered to be good for the women themselves and not bad for the development of the child. These high scoring countries and individuals are also tolerant with re­ gard to people of different ethnic background as well as regarding people with deviant behavior. As Turner pointed out, individual freedom and personal autonomy require the acceptance of the right to act and behave in the way the individual wants, and thus allow for considerable diver­ sity (Turner, 1988). The high-scorers also have a high level of protest proneness and are more ready to make personal sacrifices to preserve the

EXPLORING EUROPE’S BASIC VALUES MAP

31

environment. The kind of individualism that characterizes this dimen­ sion should not be interpreted in terms of egoism, narcissism, hedonism or ethical relativism; it does not lack a spirit of community (Etzioni, 1993), but it is socially committed. The dimension can be labeled in terms of autonomy, non-materialism, and social engagement. It repre­ sents the ‘social-liberal attitude’. The second dimension, which is (statistically) independent from the first one due to the orthogonal varimax rotation, combines a number of ‘normative’ issues. Individuals and countries that score high on this di­ mension maintain strict moral standards, they highly value societal norms and institutions and stress solidarity. Civil morality is high since self-interest and illegal behavior is rejected. Abortion is disapproved of and marriage and the intrinsic meaning of work are important. Not sur­ prisingly, religiosity scores high on this ‘normative’ dimension. Indi­ viduals and countries that classify high on this dimension can be labeled as ‘normative’ in the sense that religion and societal norms and institu­ tions are highly valued. 3 Comparing countries and generations in the 1999/2000 EVS The main conclusion from the analyses so far is that Europe’s values map can be drawn by means of two independent, meaningful and well interpretable dimensions, labeled ‘socio-liberal’ and ‘normative’ for all three levels of analysis. From now on, the aggregate country by genera­ tions (age groups) analysis outcomes will be the focus of the discussion. For each of the 198 (= 33 x 6) country by age groups, (standardized component) scores on the two dimensions has been computed. It ap­ peared that the variation in the scores for a particular dimension can be (completely) attributed to the main effects of country (the set of dummy variables representing the countries), the main effects of generation (the set of dummy variables representing the generations) and their interac­ tion effects (the product terms of the two sets of dummies). The two main effects are by far the most important and account for more than 92% of the variance of the scores of both dimension 1 and dimension 2. This implies that the country differences are remarkably similar for all generations or, to say the same thing differently, the generational differ­ ences are surprisingly similar in all countries. Despite the large socio­ economic and political differences between the countries in Europe, similar age groups seem to react in a very similar manner to events and trends in all countries. Based on these component scores several comparisons were made. First, we present a two-dimensional values map of Europe, locating the

32

CHAPTER TWO

countries on the map (Figure 2.2). Next, the ranking of the European countries on each dimension will be explored in more details (Figures 2.3 and 2.4) and the scores on the dimensions will be correlated with a number of important country characteristics, in order to find a first inter­ pretation of the countries’ positions. Then, the overall generational dif­ ferences are presented, along with a few remarks on the (small) generation-country interaction effects. 3.1 Comparing countries The horizontal axis in Figure 2.2 represents the ‘social-autonomy’, the socio-liberal dimension from non-liberal (–2.0) to liberal (2.0). The ver­ tical axis is the normative dimension and runs from non-normative (–2.0) to normative (2.0). 2.0

1.5

IR

1.0

PT

GR

IT

NI

RO ‘Normative - religious’

0.5 IS

HR HU

AT ES BE UK SL DE LX

PL

0 BU UA

FR

Sl

NL

–0.5 CZ –1.0 RU

SE

FI EG

DK

LV LT Bl EE

–1.5

–2.0 –2.0

–1.5

–1.0

–0.5

0

0.5

1.0

‘Autonomy - socio-liberalism’

Figure 2.2 Countries’ mean scores on the two dimensions

1.5

2.0

EXPLORING EUROPE’S BASIC VALUES MAP

33

The component scores of the countries are from the country by age groups analysis, and as such control for the (compositional) age group differences between the countries. The countries in the upper right quadrant formed by the perpendicular vertical and horizontal lines through the zero (mid) points are the socio­ liberal, and religious-normative countries. The most ‘typical’ examples in this quadrant lying on or above the .5 line on both dimensions are Ire­ land, Northern-Ireland, and Italy. Poland and Romania form the typical less liberal and the more religious, normative countries in the upper left quadrant. Sweden and Denmark are the most outspoken instances of the liberal, nonreligious (lower right) quadrant while Russia, Lithuania, Bel­ arus, Estonia and Latvia form a typical cluster in the nonliberal, nonreli­ gious (lower left) quadrant. The closer countries are on the map, the closer their scores on the two dimensions and thus the more similar they are. A complete interpretation of the locations of countries on the map must involve not only their present situation, but also their particular, idiosyncratic histories. We will not make an attempt to accomplish this here, but rather look at the order of the countries on each separate di­ mension and investigate how this correlates with some overall ‘adminis­ trative’ measures. To get some order in the pictures in Figures 2.3 and 2.4, the Eastern/Central European countries are grouped together as are the Western European countries. Within these clusters the countries are ordered on the basis of their score on the dimensions. On the first socio-liberal di­ mension, a clear demarcation appears between West and Eastern Euro­ pean countries where the Western European countries are on the liberal, autonomous, and the Central and Eastern European countries on the less socio liberal, autonomous side of the continuum. In particular Russia and its former states Lithuania and Belarus, and Hungary appear far less autonomous than countries like Sweden, the Netherlands, and Denmark. There are exceptions to this East-West cleavage: Portugal, Greece, and Malta are less autonomous or socio-liberal than Croatia, Slovenia and former East Germany. Note also that there is still a big difference be­ tween former West and East Germany, i.e., West Germany is much more on the socio-liberal side than East Germany. Country differences on this socio-liberal dimension may be under­ stood in terms of economic welfare. As Inglehart and others have ar­ gued, the level of economic development and the corresponding level of existential security push values in the direction of personal autonomy and postmaterialism (Inglehart, 1997). In agreement with this, the more wealthy (Western) countries display higher levels of personal autonomy than the Central and Eastern countries and within Western Europe, the

34

CHAPTER TWO

economically less favored societies of Portugal and Greece are less socio-liberal than the more economically advanced Nordic countries and the Netherlands. Of course, not all differences between the countries can be completely explained in terms of economic welfare (see the relative high position of Croatia) and, as argued before, other (historical) circum­ stances must be taken into account. But overall the picture is fairly clear. 2.0

1.5

1.0

0.5

0

–0.5

–1.0

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

–2.0

Russia Lithuania Hungary Belarus Estonia Romania Ukraine Latvia Poland Slovakia Bulgaria Czech Republic Croatia Slovenia East Germany

–1.5

Figure 2.3 Countries’ scores on ‘autonomy - socio-liberalism’ dimensions

The position of the countries on the second dimension cannot be ‘sim­ ply’ understood from their level of economic development and existen­ tial security. The economically more advanced countries are now in the middle, in-between religious, normative oriented societies on the right side of the dimension and more secular, less normative oriented societies on the left side. In fact, the economically most advanced countries re­ semble some of the least wealthy and least prosperous countries in

35

EXPLORING EUROPE’S BASIC VALUES MAP

Europe. Perhaps, the order of the countries reflects the cultural heritages of Catholics, Protestants, and Orthodox traditions and experiences5. MT = 3.18 2.0

1.5

1.0

0.5

0

–0.5

–1.0

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

–2.0

Estonia Belarus Lithuania Russia Latvia East Germany Czech Republic Slovakia Ukraine Bulgaria Slovenia Hungary Croatia Romania Poland

–1.5

Figure 2.4 Countries’ scores on ‘normative - religious’ dimension

Generally speaking, Catholic societies appear more religious and moral­ ly stricter than Protestant and Orthodox countries. Iceland appears an exceptional case in Western Europe. The high level of religiosity seems to reflect ‘a curious mixture of traditional values individualism’ (Péturs­ son & Jonsson, 1994: 164), but it is hard to explain this curious mix. Ro­ mania is a conspicuous exception being more religious/normative than expected given the more secular image of Orthodox Europe. We cannot ______________ 5

Note that religious denomination is not included in the religiosity scale.

36

CHAPTER TWO

offer a simple and conclusive explanation for this finding but one specu­ lation may be that it is due to the fact that Communism did not destroy the traditional economy and social system. The same happened in Po­ land. And where the traditional economy and social system persisted, culture remained traditional and religious (Tomka, 1998: 242). 3.2 Value profiles and country characteristics In order to obtain further insights into the nature of the country differ­ ences and check and elaborate the interpretations of our two fundamental values dimensions, the component scores have been correlated with a few available statistics. Two countries, namely, East Germany and Northern Ireland, were excluded from the analysis because not all statis­ tics were available. Table 2.2

Correlations between the two basic values dimensions and some country characteristics

Dummy: Eastern (= 0) versus Western (= 1) Europe Seats in parliament held by women as % of total HDI human development index Employment rate women as % of male rate 1999 Proportion of (self reported) Catholics

PC1 .80 ** .70 **

PC2 .36 ** –.25

.89 ** –.24

.22 –.75 **

.15

.54 **

** p < .05 (two-tailed).

Several significant correlations are found. The most interesting ones from our present point of view are presented in Table 2.2. The first di­ mension correlates highly with the dummy variable East versus West, reflecting the large differences between Eastern and Western European countries with regard to autonomy and socio-liberalism. Also, the much lower correlation of this dummy variable with the second, normative dimension follows from the discussion above. We also argued that the first dimension appears as a ‘welfare’ dimension. The rather strong cor­ relation with HDI6 supports this interpretation. It is further interesting to see that the ‘autonomy - socio-liberalism’ dimension which emphasizes the right for women to develop themselves socially, nicely correlates ______________ 6

HDI is the Human Development Index and incorporates capabilities such as ‘living a long and healthy life, being knowledgeable and enjoying a decent standard of living’ (UNDP, 2000: 20).

EXPLORING EUROPE’S BASIC VALUES MAP

37

with the percentage of women members in parliament. The second ‘nor­ mative’ dimension emphasizing the importance of norms and the insti­ tutes such as work and family in their traditional form, correlates as ex­ pected with the percentage of women working, i.e., the more ‘normative’ the less women work. Finally, the impression that the ranking of the countries on this second dimension had to do with the diverse cultural background of Catholic versus Protestant and Orthodox countries is cor­ roborated by means of the correlation with the percentage of Catholics in the country. Thus, while the first dimension of ‘autonomy’ and ‘socio-liberalism’ can be linked with an economic dimension—or at least strongly relates with economic development—the second dimension is more to be seen in terms of cultural heritage, especially the religious tradition. Although one cannot conclude to a simple one-directional causality from these correlations, it seems justified to conclude that culture matters. 3.3 Comparing generations In order to explore (net) generational effects, we have distinguished six ten-year age groups (Figure 2.5). The first age group consists of those respondents that were born before 1930, than those born in the thirties, the forties, the fifties, the sixties, respectively, and finally those born in the seventies and early eighties. Generational differences on the two values dimensions are each other’s mirror image, i.e., ‘socio-liberal’ values continuously increase with birth cohort, while ' normative-religious'values continuously de­ crease. Whether these trends reflect ‘real’ generational differences or life course differences cannot be answered having only data from one parti­ cular point in time. The next section will elaborate on this issue. As mentioned before, cohort-country interactions were very small. As such the major trend in generational differences applies in all countries. How­ ever, separate analyses for different groups of countries (analyses not re­ ported here) revealed a number of peculiarities that are worth mentio­ ning. The overall trend in Figure 2.5 reveals a continuation of genera­ tional differences for cohorts born after the Second World War as far as ‘autonomy and socio-liberal’ values are concerned. In Denmark, Swe­ den, the Netherlands, Belgium, Iceland and—to a lesser degree—France, Luxemburg, and Great Britain, these post-war differences virtually com­ pletely level off. Overall these countries rank high on ‘socio-liberalism’ suggesting that the youngest generations in these countries may have reached the ceiling in ‘autonomy’ and ‘socio-liberal’ orientation. This leveling off of post-war generational differences is only reconfirmed inthe Netherlands and Denmark as far as the ‘normative’ values dimension

38

CHAPTER TWO

is concerned. In the Czech Republic, Latvia and Lithuania, post-war dif­ ferences are also small. The overall finding that the differences in ‘nor­ mative’ values between the two oldest cohorts are (slightly) less pro­ nounced than the differences among other cohorts proved to be partially an artifact from pooling the countries. In Lithuania, Northern Ireland, United Kingdom, Romania, Slovenia, Croatia, Poland, Ukraine and Greece the oldest cohort (born before 1930) is less ‘normative’ than the cohort born in the 30s, which is opposite to the overall trend. With the data at hand it is hard to imagine what causes these particularities. Irre­ spective of that, the deviations from the overall pattern are relatively small as indicated by the small insignificant sizes of the cohort-country interaction effects. 0.7 0.6 0.5

0.4 0.3 0.2

0.1 0 –0.1

–0.2 –0.3 –0.4 –0.5 –0.6 –0.7 –0.8

until 1929

1930-39

1940-49

PC1 ‘socio-liberalism’

1950-59

1960-69

1970­ early 80s

PC2 ‘normative-religious’

Figure 2.5 Net generational differences in values (1999/2000)

EXPLORING EUROPE’S BASIC VALUES MAP

39

4 Trends in the two basic value dimensions: 1981-1990-1999 In the previous section a picture was taken from the generation-country value profiles. As any picture, it reflects the situation at a particular point in time (i.e., 1999), but at the same time it captures part of Europe’s his­ tory. In principle, however, a much better insight into the ways countries and generations change over time can be obtained by making use of all three waves of EVS. We added ‘in principle’ because these analyses have their own limitations. In the first place, not all countries included in the 1999/2000 study participated in all three waves. Actually, only a number of Western European countries was involved three times and many of the Eastern European countries participated only in the 1999 survey. Table 2.3

Principal components reduced set - 1999 rotated component matrix 1

2

Authority above autonomy

–.78

.39

Intolerance towards disturbing groups

–.82

NOT protest prone Intolerance towards ethnic groups

–.83 –.59

.24 .26

Material conditions important for marriage

–.61

.33

Less emphasis on material aspects in life is good

.73

.38

Postmaterialism Children and marriage NOT necessary

.84 .76

.20

Personal-sexual permissiveness

.79

–.47

Self-interest permissiveness

.11

–.51

Abortion approved Expressive work qualities

.22 .29

–.70 .45

Cultural homogeneity important for marriage

–.36

.79

Emphasis on authorities is good

–.38

.64

.19 .48

.59 .63

–.28

.75

Emotional qualities important for marriage Mutual bond important for marriage Religiosity

Explained variance (after varimax rotation) =

34.6

24.1

Secondly, not all indices and items that were used in the analyses thus far to identify the two dimensions have been included in the previous surveys. However, a reduced set of 17 attitudinal scales that was com­

40

CHAPTER TWO

mon in all the three waves reproduced fairly well the aforementioned values dimensions (see Table 2.3). The principal components weights were calculated from the 1999 data, and they were imposed upon the 1981 and 1990 data to enhance the comparability. To obtain an overview of the changes in (Western) Europe between 1981 and 1999/2000, generational (six cohort) differences and net changes over time (three periods), the aggregate of all Western European coun­ tries that were part of the European Values Study in 1981 were studied. Thereafter, the discussion of the country comparisons in the previous section will be extended with findings from 1981 and 1990, as far as the data allow it. 4.1 Cohort-period trends Figures 2.6 and 2.7 compare the cohort value profiles of the pooled dataset for each observation year on the two latent dimensions. Presented are the mean scores of the different cohorts in the successive years of inves­ tigation. When looking at these figures, the results of the youngest co­ hort should be interpreted with care, as the youngest cohort is not com­ pletely observed. E.g., in 1999, the youngest cohort only included people born between 1980 and 1983: the people born after 1983 were too young to be included in the sample. Period effects can be seen from the (vertical) distances between the three lines; cohort differences for a particular year follow from the course of the line connecting the cohort means. For both dimensions, it follows clearly from Figures 2.6 and 2.7 that cohort differences are more important than period effects. In most cases the differences between ob­ servation years are smaller than the differences of two consecutive co­ horts. As far as ‘autonomy and socio-liberal’ values are concerned, a steady (small) increase is observed from 1981 to 1999 for each cohort. The (large) cohort differences show that for all three periods the later, younger cohorts are more socio-liberal and autonomous than the older generations. A conspicuous and interesting result is that the continuous upward trends over the generations towards more autonomy and socio­ liberalism slows down starting from the post-war birth cohort 1950­ 1945. So, a general conclusion can be that there is a steady increase in Europe of socio-liberal values and that this increase is true for all Western-European countries given the fact that cohort-country interaction ef­ fects are virtually non-existing. The mechanisms of change are not so much period effects or due to particular events, but the result of the re­ placement of older, more ‘conservative’ generations by younger, more socio-liberal ones. And if the post war trend continues, the trend stabi­ lizes at a high level of autonomy and socio-liberalism.

EXPLORING EUROPE’S BASIC VALUES MAP

41

1.5

1.0

0.5

0

–0.5

–1.0

–1.5 1900-09 1910-19 1920-29 1930-39 1940-49 1950-59 1960-69 1970-79

1981

1990

1980+

1999

Figure 2.6 Cohort differences on ‘autonomy - socio-liberalism’

Figure 2.7 shows a somewhat different result for the ‘normative-religious’ dimension. The (small) period effects indicate that there has been an increase of ‘normative-religious values’ from 1981 to 1990 but no more or only modest change between 1990 and 1999/2000. Period and cohort effects, however, do not point into the same direction. Normativereligious values decrease with birth cohort, but rise with observation pe­ riod. It is not completely clear how to interpret these results. There is the suggestion, both in the data and in theory, that these co­ hort differences are not only generational differences (the older cohorts are more normative oriented than the younger cohorts) but also or espe­ cially life course differences (the older one is, the more normative ori­ ented).

42

CHAPTER TWO

1.5

1.0

0.5

0

–0.5

–1.0

–1.5

1900-09

1910-19

1920-29

1930-39

1940-49

1950-59

1981

1990

1999

1960-69 1970-79

1980+

Figure 2.7 Cohort differences on ‘normative-religious’ values

The data pattern points at least partially into the latter direction. For in­ stance the cohort born between 1920 and 1929 is in 1990 approximately of similar age as the cohort born between 1910 and 1919 is in 1981 and their value orientation is quite similar, a pattern that is repeated at sev­ eral places in Figure 2.7. However, the growing ‘conservatism’ on this dimension in the eighties is not reconfirmed in the nineties. Hence, no definite claims regarding the relative significance of a life-cycle interpre­ tation versus cohort interpretation of the generational differences in nor­ mative values can be made. The general picture regarding period changes in values that emerged in these analyses is that the eighties were marked by a growing ‘conser­ vatism’ regarding the ‘normative’ value dimension, while at the same time the support for ‘autonomy’ and ‘socio-liberal’ values tended to in­

EXPLORING EUROPE’S BASIC VALUES MAP

43

crease. The latter dimension continues to develop into that direction in the nineties, whereas the ‘normative’ dimension more or less consoli­ dates. An uncertain factor as far as the latter dimension is concerned is the impossibility to separate clearly the effects of age, period, and cohort (generation). Therefore, additional analyses using additional data are re­ quired (Hagenaars, 1990, Chapter 6). 4.2 Countries on the move? Apart from the overall sketch of Western Europe given in the previous section, it is interesting to look at the changes in the separate countries and where possible in the Eastern European countries. For each values dimension we present figures for the countries that have participated in each wave (Figures 2.8 and 2.10) and the countries that joined the EVS project in 1990 (Figures 2.9 and 2.11). We controlled for differences in cohort composition by restricting the comparison to cohorts that were observed in each wave of the EVS, i.e., born between 1920 and 1969. Hence the periods lines only reflect period differences and not trend ef­ fects due to the process of social metabolism.

2.5 2.0 1.5 Sweden Spain

Netherlands

Italy

Ireland

Iceland

Northern Ireland

–1.5

West Germany

–1.0

France

–0.5

Belgium

0

United Kingdom

0.5

Denmark

1.0

Figure 2.8 Countries' scores on ‘autonomy - socio-liberalism’ in 1981, 1990, 1999

As is clear from Figure 2.8 and in agreement with Figures 2.6 and 2.7, the period effects observed for the countries are not very impressive. However, they do differ among the countries. If we look first at the

44

CHAPTER TWO

socio-liberal dimension in Figure 2.8 (Western European countries, 1981-1990-1999), it is clear that in Belgium, Ireland, Italy, Northern Ire­ land, and Sweden the trend toward autonomy and socio-liberal values is continuing from 1981 through 1990 to 1999. In Denmark, France and Iceland the trajectories are v-shaped with the lowest value in 1990. In Great Britain, West Germany, and Spain the opposite, inverted v-shape trend is seen. In the Netherlands, the eighties showed an increase, while the nineties can be characterized as a consolidation of the existing value orientation. Also in the second set of countries (Figure 2.9) the trends from 1990 to 1999/2000 are different. Rising levels of autonomy and socio-liberal orientation are clearly observed in Austria, Bulgaria, and the Czech Re­ public. In Finland, Hungary, and Portugal, and to a lesser extent Latvia and Malta, the values go down, while in Poland hardly any shift oc­ curred. 1.5

–2.0

Slovakia

Portugal

Poland

Latvia

–1.5

Bulgaria

–1.0

East Germany

–0.5

Czech Republic

0

Austria

0.5

Hungary

Finland

1.0

–2.5

Figure 2.9 Countries’ scores on ‘autonomy - socio-liberalism’ in 1990 and 1999

It is hard to find any general explanations for these trends. Most proba­ bly, specific, idiosyncratic historical circumstances determine these dif­ ferent trajectories. Moreover, it should be remembered that the period ef­ fects are rather small and almost everywhere within the range of .5 stan­ dard deviations. The general conclusion is justified that major period shifts in autonomy and socio-liberalism are not observed since the eight­ ies. Consequently, the main differences among the countries are valid for

EXPLORING EUROPE’S BASIC VALUES MAP

45

all three points of time. The information with regard to the changes on the second normative dimension is presented in Figures 2.10 and 2.11. As appears from Figure 2.10, the overall change between 1981 and 1990 is true for all individual Western European countries, except West Ger­ many. Further, in Figure 2.7 no aggregate period effects were found be­ tween 1990 and 1999. This seems to be an artifact of the data pooling. Denmark, Iceland, the Netherlands, Spain, and especially Sweden showed decreases on the normative dimension between 1990 and 1999, while increases are found in Belgium, United Kingdom, France, West Germany, Ireland, Italy and a very small one in Northern Ireland.

2.5 2.0

Sweden

Spain

Netherlands

Northern Ireland

Italy

West Germany

–1.5

France

–1.0

Denmark

–0.5

United Kingdom

0

Belgium

0.5

Iceland

1.0

Ireland

1.5

Figure 2.10 Countries’ mean scores on ‘normative - religious’ in 1981, 1990 and 1999

Among the (mainly Eastern) European countries that entered EVS later (Figure 2.11), we also see two groups. There is a decrease in normative values for the Czech Republic, East Germany, Slovakia, and, a big drop in Portugal and a small one in Austria. At the same time an increase in normative values can be found in Bulgaria, Finland, Hungary, Poland and especially Latvia. For several countries the differences are rather small, for others the changes are substantial but not always in the same direction. Most pro­ bably we need again knowledge of the particular historical circum­ stances of these countries to find good explanations of their trajectories

46

CHAPTER TWO

of change. All these results with regard to period-, cohort-, and countrydifferences at least indicate more precisely what is there to explain. 1.5

Portugal

Hungary Latvia

–2.0

Finland

–1.5

East Germany

–1.0

Czech Republic

–0.5

Bulgaria

0

Austria

0.5

Slovakia

Poland

1.0

–2.5

Figure 2.11 Countries’ mean scores on ‘normative - religious’ in 1990 and 1999

5 Conclusion The main purpose of this chapter has been to explore Europe’s basic values and to draw the cultural map of Europe using the survey data from the European Values Study. In order to circumvent the usual arti­ facts connected to the use of single items that are mainly due to particu­ lar different or similar item formats (dichotomous answering categories, five or seven point scales, position in the questionnaire, accidental iden­ tical single words, etcetera), our starting point has been a number of in­ dices constructed from items that together represented substantive, theo­ retical concepts. At least to the surprise of some of us, the whole set of indices could be meaningfully reduced to just two basic value dimen­ sions, labeled the autonomous/socio-liberal dimension and the normative/religious dimension. The first dimension refers to non-materialism, autonomy, the right for women to develop themselves socially, tolerance toward ethnic groups and personal-sexual permissiveness. It also refers to political values such as postmaterialism, protest proness, favoring democracy and environmentalism. Typical ‘normative’ indices were: re­ ligiosity, emphasis on marriage and the family, the rejection of abortion,

EXPLORING EUROPE’S BASIC VALUES MAP

47

the importance of work, the emphasis on authority and general trust in international organizations. Normative values also imply civil morality (rejecting illegal actions and self-interest), stressing the need for solidar­ ity within society. In identifying these ‘fundamental’ value dimensions, the analyses outcomes were essentially the same for the individual and the aggregate level analysis. The position of the countries on this two-dimensional map can be partly understood in general terms. The autonomous/socio-liberal di­ mension correlates especially with the East-West distinction and with economic welfare: the higher a country’s welfare, the more liberal it is. With regard to the changes of the position on this dimension over time, the most important source of change seems to be generation replacement as each following cohort scores a bit higher on this dimension (although a certain degree of stabilization can be expected as there is less differ­ ence between the post war generations in this respect). The ‘pure’ period effects are generally rather small. Regarding the normative/religious dimension cultural background, especially the distinction between traditionally Roman-Catholic and non-Catholic countries seems to be crucial. There is a high correlation between the scores on this dimension and gender stratification defined as the relative number of women having paid work. The cohort differences reflect here probably to some extent age differences: the older, the more religious/normatively oriented. However, additional analyses are neces­ sary to confirm this interpretation. ‘Pure’ period effects are generally small. The fact that overall in Europe the trend of the eighties towards more normative/religious values seems to have stopped in the nineties, reflect the circumstance that in some countries the trend did continue while it reverses in other countries. We are now able to place the countries more precisely on our map. The map reveals that Europe is not a homogenous part of the world. On the contrary, despite a common Christian tradition, European unity seems to be a unity of diversity; there appear important differences be­ tween Europe’s societies that have to do with different levels of eco­ nomic development but also with varieties in cultural heritages, lan­ guages, religious and ideological traditions, and differences in political and educational systems. The general trends and explanations offered above are certainly not sufficient to understand all outcomes. The results of our analyses seem to suggest that there is no unique trajectory of val­ ues change. As Van Deth and Scarbrough (1995) have argued, the dy­ namics of values change cannot be summarized in a single and straight­ forward way by referring to a single theory of change. Value orientations appear dependent upon specific national contexts and a nation’s histori­

48

CHAPTER TWO

cal development. A much more precise look at the (changing) map and digging deeper into the country-specific circumstances and historical de­ velopments is necessary to arrive at really satisfactory socio-cultural ex­ planations. This chapter only made the very first attempt. The same is true for what happened over time to the countries’ positions. From a methodological point of view more extended age-period-cohort analyses using much more than just three waves are necessary for an adequate de­ scription, let alone an explanation of what went on in Europe these last twenty years. References Ester, P., L. Halman & R. de Moor (eds.) 1993, 1994. The Individualizing Society. Tilburg: Til­ burg University Press. Etzioni, A. 1993. The Spirit of Community. Rights, Responsibilities, and the Communitarian Agenda. New York: Crown. Hagenaars, J.A.P. 1990. Categorical Longitudinal Data: Loglinear Panel, Trend, and Cohort Analysis. Newbury Park: Sage. Halman, L. 1991. Waarden in de Westerse wereld. Tilburg: Tilburg University Press. —— 2001. The European Values Study: A Third Wave. Sourcebook of the 1999/2000 Euro­ pean Values Studies. Tilburg: EVS, WORC, Tilburg University. ——, F. Heunks, R. de Moor & H. Zanders 1987. Traditie, secularisatie en individualisering. Tilburg: Tilburg University Press. —— & A. Vloet 1994. Measuring and Comparing Values in 16 Countries of the Western World. Tilburg: WORC. Hofstede, G. 1980. Culture’s Consequences: International Differences in Work-related Values. Beverly Hills: Sage. —— 2001. Culture’s Consequences. Comparing Values, Behaviors, Institutions, and Organiza­ tions Across Nations. Thousand Oaks: Sage. Inglehart, R. 1977. The Silent Revolution. Princeton: Princeton University Press. —— 1990. Culture Shift in Advanced Industrial Society. Princeton: Princeton University Press. —— 1997. Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies. Princeton: Princeton University Press. Pétursson, P. & F.H. Johnsson 1994. Religion and Family Values. Attitudes of Modern Icelan­ ders in a Comparative Perspective. Pp. 151–166 in T. Pettersson & O. Riis (eds.), Scandi­ navian Values. Religion and Morality in the Nordic Countries. Stockholm: Gotab. Schwartz, S.H. 1994. Studying Human Values. Pp. 239–254 in A.M. Bouvy, F. van de Vijver, P. Boski & P. Schmitz (eds.), Journeys into Cross-Cultural Psychology. Lisse: Swets & Zeitlinger. Tomka, M. 1998. Coping with Persecution. Religious Change in Communism and in PostCommunist Reconstruction in Central Europe. International Sociology 13: 229–248. Triandis, H. 2000. Culture and Conflict. International Journal of Psychology 35: 145–152. —— 2001. Individualism-Collectivism and Personality. Journal of Personality 69: 907–924. Turner, F.C. 1988. Basic Values: Religion, Patriotism, and Equality. Pp. 181–200 in M. Dogan (ed.), Comparing Pluralist Democracies. Boulder & London: Westview Press. UNDP (United Nations Development Programme) 2000. Human Development Report (HDR) 2000. Oxford: Oxford University Press. www.undp.org/hdro. Van Deth J.W. & E. Scarbrough. 1995. Perspectives on Value Change. Pp. 527–540 in J.W. van Deth & E. Scarbrough (eds.), The Impact of Values. Oxford: Oxford University Press.

APPENDIX

THE ITEMS INCLUDED IN THE ANALYSES

50

CHAPTER TWO

In this appendix we describe the procedures we followed in constructing the variables used in the analyses. The sequence of this description is more or less the same as the order of the questions and items in the ques­ tionnaire. For more details about the questionnaire we can refer to the recent sourcebook (Halman, 2001: 299–334) and the information pro­ vided on EVS website (www.europeanvalues.nl). The data of each coun­ try has been weighed, if necessary, by gender, and age and in some countries also by other characteristics, e.g., region, to correct for devia­ tions from national population parameters (see Halman, 2001: 384–386). The factor analyses are based on data that are weighed in such a way that each country has an equal weight. This weight variable was calculated as the product of the original weight variable multiplied by a value that gives each country a weighted sample of 1,000. Environment Readiness to sacrifice for the environment was measured by three state­ ments. The respondents were asked whether they strongly agreed (= 1), agreed (= 2), disagreed (= 3), or strongly disagreed (= 4) with the following statements:

– I would give part of my income if I were certain that the money would be used to prevent environmental pollution. – I would agree to an increase in taxes if the extra money is used to prevent environmental pollution. – The Government has to reduce environmental pollution but it should not cost me any money. Missing values were substituted with 2,5. Factor analysis was applied to calculate scores. These factor scores were multiplied by –1, so that high score on the construct revealed offering willingness. Intolerance Three kinds of intolerance have been distinguished based on the question ‘Which groups of people one would not like to have as neighbors’. Counting the number of groups the respondents do not want to have as neighbors revealed the level of intolerance.

– Intolerance toward ethnic groups includes not accepting as neigh­ bors: people of different race, Muslims, immigrants, Jews, people with large families.

EXPLORING EUROPE’S BASIC VALUES MAP

51

– Intolerance toward disturbing groups refers to not liking to have as neighbors: people with criminal record, heavy drinkers, emotionally unstable people, people who have AIDS, drug addicts, homosexuals. – Intolerance toward extremists includes two groups only: left wing extremists, right wing extremists. Reasons of poverty Respondents were asked to choose the two most important reasons of why there are people in this country who live in need? The reasons pro­ vided were: 1. Because they are unlucky; 2. Because of laziness and lack of willpower; 3. Because of injustice in our society; 4 It’s an inevitable part of modern progress. Four variables were computed: unluck, laziness, injust and progress. In case a reason was considered most important, the score on the vari­ able was 3; in case it was second most important it was given score 2; in case it was not mentioned it was given score 1. Missing values were not substituted. Work A number of work qualities were presented and the respondents were asked to indicate which of these he/she considered important. A factor analysis revealed two dimensions, one referring to expressive qualities; the other to instrumental qualities. Missing values were included as not mentioned category. Expressive qualities are: an opportunity to use ini­ tiative, a responsible job, a job in which you feel you can achieve some­ thing, a useful job for society, a job that meets one' s abilities, meeting people, a job respected by people in general, a job that is interesting. Scores were calculated by applying factor analysis (orthogonal rotation) on all work qualities. The instrumental qualities did not load, or only modestly, on this expressive dimension. A high score indicates the im­ portance of expressive work qualities. Work ethos combines the degree to which respondents agreed with statements like: ‘To fully develop your talents, you need to have a job’; ‘It is humiliating to receive money without having to work for it’; ‘Peo­ ple who don’t work turn lazy’; ‘Work is a duty towards society’; ‘Work should always come first, even if it means less spare time’. Response categories were: 1. agree strongly; 2. agree; 3. neither agree nor disa­ gree; 4. disagree; 5. disagree strongly. Missing values were recoded as 3. Factor analysis was applied to calculate scores that are multiplied by –1 so that high scores indicate high levels of work ethos.

52

CHAPTER TWO

Religiosity Although various dimensions can be distinguished (see, e.g., Halman & de Moor, 1994; Halman & Vloet, 1994) they appeared to be strongly correlated. Therefore we included only one dimension of religiosity in­ stead of these distinctive dimensions. The factor analyses to calculate scores on religiosity included the following items:

– The importance of religious services at the occasions of birth, mar­ riage, and death. ‘Yes’ answers were coded one; ‘no’ answers as well as ‘missing values’ were coded 0. – The importance of God is revealed from the question ‘How impor­ tant is God in your life?’: 1 = not at all; 10 = very important. Missing values were recoded as 5. – The importance of religion appears in the responses to the question ‘Do you find comfort and strength from religion?’. The responses (yes, no, don’t know) were dichotomized: 1 = yes; else = 0. – Personal religiosity is tapped with the question ‘Independently of whether you go to church or not would you say you are: 1. a reli­ gious person; 2. not a religious person; 3. a convinced atheist’. This variable was recoded simply as religious person (= 1) versus the rest (= 0). – Functional differentiation appears from the responses to the question ‘How much do you agree with the following statements: politicians who do not believe in God are unfit for public office; it would be better for (country) if more people with strong religious beliefs held public office’. Answer categories were 1. strongly agree; 2. agree; 3. neither agree nor disagree; 4. disagree; 4. disagree strongly. Missing values were recoded as 3. – Traditional belief or religious orthodoxy is measured by the belief in God, life after death, hell, heaven, sin. Response categories (yes, no, don’t know) were dichotomized: 1= yes versus the rest = 0. – Finally church attendance was included. The question was ‘Apart from weddings, funerals and christenings, about how often do you attend religious services these days?’ Response categories were: 1. more than once a week; 2. one a week; 3. once a month; 4. Christmas/Easter day; 5. Other specific holy day; 6. once a year; 7. less of­ ten; 8. never, practically never. Factor analyses yielded one factor and has been used to calculate scores. A high score indicates high levels of religiosity.

EXPLORING EUROPE’S BASIC VALUES MAP

53

Marriage Four marital dimensions have been distinguished based on the question what is important for a successful marriage:

– Emotional marital qualities are ‘spending as much time together as possible’, ‘talking a lot about mutual interests’, and ‘sharing house­ hold chores’. – Cultural homogeneity qualities refer to the importance of ‘shared re­ ligious beliefs’, ‘being of the same social background’, ‘agreement on politics’. – Mutual bond qualities are ‘mutual respect and appreciation’, ‘under­ standing and tolerance’, ‘faithfulness’. – Material conditions are ‘an adequate income’, ‘good housing’. Scores were calculated by simply counting the number of qualities re­ spondents considered important. Parental values A number of qualities which children can be encouraged to learn at home were presented and it was asked to choose up to five that one con­ siders important. Qualities that referred to authority, such as ‘hard work’, ‘thrift, saving money and things’, ‘religious faith’, and ‘obedien­ ce’, were added and qualities that referred to autonomy, such as ‘inde­ pendence’, ‘feeling of responsibility’, ‘imagination’, ‘determination, perseverance’, were subtracted. A low score on the resulting scale thus indicates preference of autonomy above authority while a high score in­ dicated a preference of authority above autonomy. Children The importance of marriage and children appears from the responses to two items: ‘Do you think that a woman has to have children in order to be fulfilled or not’ (1 = needs children; 2 = not necessary) and ‘A mar­ riage or long-term stable relationship is necessary to be happy’ (1 = agree strongly; 2 = agree; 3 = neither agree nor disagree; 4 = disagree; 5 = disagree strongly). We simply added the responses so that a high score indicates that children and marriage are not necessary. Women roles Attitudes towards gender (or more specific female) roles in society were measured by the question whether or not people agreed or not with statements like:

54

CHAPTER TWO

– A working mother can establish just as warm and secure a relation­ ship with her children as a mother who does not work. – A pre-school child is likely to suffer if his or her mother works. – A job is alright but what most women really want is a home and children. – Being a housewife is just as fulfilling as working for pay. – Having a job is the best way for a woman to be an independent per­ son. – Both the husband and wife should contribute to household income. Response categories were: 1. agree strongly; 2. agree; 3. neither agree nor disagree; 4. disagree; 5. disagree strongly. Missing values were re­ coded as 2,5. A factor analysis yielded three dimensions and factor analysis was applied to calculate scores on these dimensions. The first factor is la­ beled working woman is ok because it includes the first two items; the third and fourth items express that it is bad for the child if women work while the idea that women should NOT have a job or contribute to in­ come is indicated by the last two items. Abortion The view on abortion appears from a count on the two reasons for abor­ tion one approves (= 1). The reasons are: ‘Where the woman is not mar­ ried’; ‘Where a married couple does not want to have any more chil­ dren’. Protest proness Respondents were presented different forms of political action and it was asked whether they actually done any of these activities, whether they might do it or would never, under any circumstance do it. The actions included were: ‘signing a petition’, ‘joining in boycotts’, ‘attending law­ ful demonstrations’, ‘joining unofficial strikes’, ‘occupying buildings or factories’. We counted the numbers of ‘would never do’ answers so that a high score on the composite variable indicates a low level of protest proness. Liberal versus socialist views Respondents were presented a number of opposite statements and they were asked to place their views on the scale from 1 (one statement) to 10 (the opposite statement). The statements were:

EXPLORING EUROPE’S BASIC VALUES MAP

55

– Individuals should take more responsibility for providing for them selves versus the state should take more responsibility to ensure that everyone is provided for. – People who are unemployed should have to take any job available or lose their unemployment benefits versus people who are unemployed should have the right to refuse a job they do not want. – Competition is good. It stimulates people to work hard and develop new ideas versus competition is harmful, it brings out the worst in people. – The state should give more freedom to firms versus the state should control firms more effectively. Missing values were recoded as 5.5. Factor analysis was applied to cal­ culate scores and the dimension was labeled as state responsibility above individual responsibility. High scores refer to socialist views of state re­ sponsibility whereas low scores on this factor represent the liberal views of personal responsibility. Postmaterialism The index of materialism versus postmaterialism follows earlier work of Inglehart (1977, 1990, 1997) and is based upon one question and four items. Respondents were asked to choose the two most important of the following political goals (postmaterialist items are indicated as pm):

– – – –

Maintain order in the country. Give people more say in the decisions of government (pm). Fight rising prices. Protect freedom of speech (pm).

The index is created as follows: in case the first and second most impor­ tant goals ware materialist, the respondent is considered to be a pure ma­ terialist. In case two postmaterialist items are most important, the re­ spondent is a pure postmaterialist. In case a materialist items is most im­ portant and one of the postmaterialist items is second most important, the respondent is mixed with materialist preference; in case a postmate­ rialist goal is most important and materialist goal second most important, the respondent is also mixed, but now with postmaterialist preference. Future views A list of changes in our way of life that might take place in the near fu­ ture was presented and it was asked to tell for each one, if it were to happen whether people thought it would be a good thing (= 1), a bad

56

CHAPTER TWO

thing (= 2), or whether the respondent did not mind (= 3)? Two future views were distinguished and we simply counted the number of changes the respondents considered a good change. Emphasis on authorities is good appears from ‘greater respect for au­ thority’, ‘more emphasis on family life’, ‘more power to local authori­ ties’. Less emphasis on material aspects is good appears from ‘less empha­ sis on money and material possessions’, ‘decrease in the importance of work in our lives’. Confidence in institutions Confidence in legitimacy institutions appears from the degree to which respondents have confidence in the education system, the press, trade unions, the police, parliament, civil service, the social security system, health care system, and the justice system. Respondents were asked to indicate if they had a great deal (=1), quite a lot (=2) not very much (=3) none at all (=4) confidence in these. Missing values were recoded as 2.5. Scores were calculated using factor analysis. After multiplying these fac­ tor scores with –1, high scores refer to high levels of confidence, and low levels indicate lower levels of confidence in these institutions that can be regarded legitimacy institutions. Confidence in international organizations is a simple sum score (di­ vided by 2) of levels of confidence in the European Union and United Nations Organization. Original scores were first recoded in such a way that code 4 indicated high levels of trust, and code 1 low levels. Missing values were recoded as 2.5. Confidence in authority institutions is tapped by two items only: the church and armed forces. Again answer categories were recoded and summed, divide by two to reveal the levels of confidence in these institutions. Democratic attitude To measure pro-democratic attitudes it was asked to indicate if various types of political systems that were described were very good (=1), fairly good (=2), fairly bad (=3), or very bad (=4) ways of governing a coun­ try. The political systems described were:

– Having a strong leader who does not have to bother with parliament and elections. – Having experts, not government, make decisions according to what they think is best for the country. – Having the army rule the country.

EXPLORING EUROPE’S BASIC VALUES MAP

57

– Having a democratic political system. Missing codes were substituted with 2,5. Further, a number of statement about a democratic political system were read off and it was asked to indicate if one agreed strongly (=1), agreed (=2), disagreed (=3) or disagreed strongly (=4) with these:

– Democracy may have problems but it’s better than any other form of government. – In democracy, the economic system runs badly. – Democracies are indecisive and have too much squabbling. – Democracies aren’t good at maintaining order. Missing values were recoded into 2,5. A factor analysis yielded one factor and was used to calculate scores. High scores on this factor indicate being in favor of democracy. Moral values With regard to morality we have made a distinction in three moral orien­ tations, based on the degree to which behaviors could be justified (or not). A scale from (1 =) never justified to (10 =) always justified was presented and the respondents were asked to indicate his/her position. Missing values were recoded as 5.5. Legal-illegal permissiveness appears from the justification of ‘taking and driving away a car belonging to someone else (joyriding)’; ‘throw­ ing away litter in a public place’; ‘driving under the influence of alco­ hol’; ‘speeding over the limit in built-up areas’; Personal sexual permissiveness appears from the acceptability of ‘homosexuality’; ‘abortion’; ‘divorce’; ‘euthanasia (terminating the life of the incurably sick)’; ‘suicide’; ‘having casual sex’. Self-interest permissiveness refers to the acceptance of ‘claiming state benefits which you are not entitled to’; ‘cheating on tax if you have the chance’; ‘lying in your own interest’; ‘paying cash for services to avoid taxes’. For each of these moral orientations we have applied different factor analyses, yielding one factor solutions. Factor scores were calculated and high scores indicate high levels of permissiveness. Regional belonging For regional belong we computed 5 new variables based on the question which of the geographical groups the respondent belonged to. The geo­ graphical groups presented were: locality or town where you live; region

58

CHAPTER TWO

of country where you live; country as a whole; Europe; the world as a whole. The 5 new variables are town, area, nation, Europe, world, and these have been computed as follows. If a geographical groups was men­ tioned first, a score 3 was give; if chosen secondly, a score 2 was given. If not chosen, a score 1 was given. Justice dimension A number of items yield information on just society. It was asked how important it is to:

– Eliminating big inequalities in income between citizens. – Guaranteeing that basic needs are met for all, in terms of food, hou­ sing, cloths, education, health. – Recognizing people on their merits. – Giving young people equal opportunity to pursue their education ir­ respective of family income. Response categories ranged from 1 = very important to 5 = not at all im­ portant. Missing values were recoded in 3. Scores are based on factor analysis. A high score on the dimension reflects the acceptance of injus­ tice in society. Solidarity The measurement of solidarity consisted of two questions. One question asked about the degree to which people are concerned with the living conditions of ‘people in the neighborhood’, ‘people in the region’, ‘fel­ low countrymen’, and ‘Europeans’. The other question asked about con­ cern with ‘elderly in country’, ‘unemployed in country’, ‘immigrants’ and ‘sick and disabled’. Response categories ranged from 1 = very much to 5 not at all. Missing values were recoded as 3. A factor analysis yielded one factor on which, after being multiplied with –1, high scores indicate high levels of solidarity.

INTERMEZZO

RUUD LUBBERS ON EUROPEAN PRIDE AND PRESIDENTS: ‘A UNITED EUROPE IS AN ANTIDOTE TO GLOBALIZATION’ INTERVIEW BY MARGA VAN ZUNDERT

The UN High Commissioner for Refugees, Ruud Lubbers, perceives a strong paradox in Europe. When serious problems arise, each nation wants to handle them in its own way – even if an adequate answer is only possible through European measures. Nevertheless, Lubbers feels a European federation is inevitable and is a good thing: ‘When I get homesick, it’s for Europe, not for the Netherlands.’ By nature Ruud Lubbers is a businessman, a doer. With his nononsense politics and negotiation skills he set a record in Dutch history: over twelve years in the office as a prime-minister. He may be known as a post-ideologist, but he sees the decline of the Muses as the greatest threat to our modern society, and is one of the initiators of Earth Char­ ter, a worldwide initiative for global ethics. Lubbers foresees a federal European Union in the long run, complete with president, flag and na­ tional pride, a union that can act as an antitoxin to globalization. Lub­ bers’ Europe, however, is not all about power: ‘My Europe is a valuable continent thanks to its development and civilization, thanks to the values it puts first and its way of life.’

Mister Lubbers, do you believe that there are European values? In my function as high commissioner for the refugees, I travel a lot. Af­ ter a while, I get a little homesick. However, I’m always longing more for Europe than for my home country, the Netherlands. To me, that pro­ ves that Europe has an identity, a certain individuality. In daily practice, that European identity is perhaps most recognizable in our culture, but culture is based on common values; values translate into culture through ideas, ideals, and ideologies. How would you describe these European values? A common history has marked the different European peoples; they share a great deal of cultural inheritance. I think that the values of Euro­

60

INTERMEZZO

pe are still largely those of humanism and Christianity, in which the Christian values include both Catholic influences, a focus on ‘we,’ the community, and Protestant influences: a focus on ‘I’ and individual res­ ponsibility. The first millennium was perhaps the most prominent stage in Euro­ pean history for creating European values: individualization is coming of age and humanism breaks through. From that moment on the human be­ ing has been put first. Putting the individual first like Europeans do, is, however, also a weakness. We show very little respect for nature. During the industrial revolution we degraded nature, and excelled in exploiting natural re­ sources. Respect for nature has not been lost, it can be corrected. I see signs that we recognize it as a problem. Another true European characteristic is that fundamentalism doesn’t thrive here. There has always been openness for other ideas and dissi­ dents. And it is evident that Europeans have problems with fundamenta­ lism in Islamic countries—they find these ideas too radical—but they also have problems with the right-wing fundamentalism in the US. One may argue that Europe has also had its waves of fundamental­ ism: Luther who nailed his propositions to the church doors, and the Ca­ thari and the Albigensians in an earlier period. Perhaps I’m an optimist, but I believe that these movements were resistances against those who seek power, against regents who misappropriate values for power and reign rather than fundamentalism. Europe has a strong tradition of the people opposing to those who confuse ideals and power; consider the Reformation and the French Revolution. If I were to summarize European values, I would say: ‘Europe is not prone to fundamentalism, but is social minded and has clear individua­ listic elements: the human being is put first.’ The Reformation and the Enlightenment have been clear successes on this continent. But maybe five hundred years ago these values were even more European. After the peace treaty of Westphalia in 1648, the various states were formed, and that divided Europe to a certain extent. Borders were erected that made patterns of values more difficult to recognize. Has your opinion about Europe changed since you have become High Commissioner for the UN? Not really, but being more and more a global citizen now, a certain para­ dox in Europe has struck me. European countries consider the problem of refugees and immigration highly important, so important that they prefer to follow their own approach in handling it. But that is simply im­ possible. In a Europe without borders one only shifts the problem to its

RUUD LUBBERS ON EUROPEAN PRIDE AND PRESIDENTS

61

neighbors. Immigrants, illegal strangers and asylum seekers are per defi­ nition European problems. Before I became high commissioner, I hadn’t realized just how big that paradox is. Serious threats from outside lead to small-scale thinking in Europe instead of large-scale thinking, even if an adequate solution is only possible through large-scale measures. Are European values changing? Like all people, Europeans are undergoing globalization at the moment. Globalization means an enormous intensification of international con­ tacts and this also evokes resistance. As a reflex against globalization, people turn to their own familiar culture and values. The threat that glo­ balization gives rise to mobilizes the protection of and desire for one’s own culture, traditions, rites and way of life. Regretfully, it also results in opposition to immigrants and foreigners. This trend of antigloba­ lization is very strong; I think that if you assess the European values to­ day, you will witness the inclination towards tradition clearly. People have become more traditional, more conservative. You participate in Earth Charter, an initiative to formulate global val­ ues. Why would we, Europeans, need global values? As I mentioned before, a weak point in European values is our relation­ ship to nature and the environment. Although Earth Charter encompas­ ses far more, it started out as an initiative to improve this relationship. That also explains the name of the initiative. However, Earth Charter is perhaps even more about intergenerational thinking. Our world is highly focused on today. We reason that if one receives something in ten years, it will only be worth a quarter of its value today. We discount everything on the basis of X percent interest. This way of reasoning does not com­ ply with nature; nature has no interest rate. On the contrary, nature’s beauty will perhaps be worth more in ten years than it is now; environ­ mental stress can accumulate to a breaking point. Earth Charter seeks to promote intergenerational thinking, seeks to oppose horizontalism and horizontal thinking. But isn’t Europe, of all states, a stimulator of sustainability? If you look at Kyoto that’s true. In a way horizontalism is un-European, but we have surrendered to the economization of society. There are so­ cieties that have surrendered even more, the USA for instance and the new economies. They pursue the optimal return of investments even more grimly. For that matter, there is something peculiar if you analyze the role of the US in Kyoto. American values are all about democracy

62

INTERMEZZO

and freedom, perhaps best symbolized by Thanksgiving Day. But when it comes to American politics, I dare say that ‘oil talks’ prevails over val­ ues. Why does Kyoto evoke so much resistance? Because it affects the oil industry. Why is villain state Iraq under pressure and North Korea not? Because there are oil reserves in Iraq and not in North Korea. Globalization and Europeanization, it’s not a competition, but which will win? Globalization will continue, that’s a fact. The question is: How do you position yourself in a globalized world? From that perspective I have a strong preference for stimulating Europeanization. Not because of some kind of romantic feeling, but because we will be overwhelmed by globalization if we retreat to our Dutch, Belgian, French or German ways. A united Europe is needed, almost as an antidote, an antitoxin, to globalization. And perhaps I’m too much a politician, but a united Euro­ pe also means power. Together, you have a certain influence. An in­ fluence that you can use to formulate a good answer towards global de­ velopments. Thus, we are heading towards having a European president? In the long run, yes. However, it still may take many decades. That is perhaps a blessing in disguise, the process is possibly more important than the result. But Europe will form a federation, a federation with a good balance between top and bottom, because individual states should have ample room for their own identities – that will work better. It will be a federation with its own president, flag and pride, but not a Europe in which all differences are wiped out; if Texas perseveres in the USA, the Netherlands can persevere in Europe. What role will the European federation play in the world? The answer to that question is, by definition, subjective. My Europe is centered around the individual, an individual who cannot flourish with­ out intergenerational thinking. He or she needs an orientation, a focus on something absolute. To me that focus, that goal, is religious. It may ma­ ture into humanism à la Erasmus, but even he continued to be religious. To me the orientation also includes a dialogue with reality; that is ac­ cepting reality as it is, but at the same time being convinced that your ac­ tion and that of society really can make a difference; not detaching your­ self from reality in a kind of romantic perception or delivering yourself to reality because you cannot change it. I sincerely believe mankind can make a difference: we can ruin Europe, but we can also make it a beauti­ ful place. Europe should spread this notion by not giving in to horizon­

RUUD LUBBERS ON EUROPEAN PRIDE AND PRESIDENTS

63

talism, to shallowness. I think Europe has the potential to do this— thanks to its history—and should do it. My Europe is not a superpower, but a valuable continent thanks to its development, its civilization, the values it puts first and its way of life. Your orientation is religious, but Europe is secularizing fast … That secularization goes on and on is not true. Certainly there is seculari­ zation, but religion is deeply rooted in our genes. I’m actually more con­ cerned about the risk of those genes mutating into fundamentalism than I am about secularization. You once said that the diminishing knowledge of the muses is the grea­ test danger to society. Do you still feel that way? Yes, horizontalism, economization, science and technology result in the decline of the arts. And culture, being the celebration and expression of values, this presents the greatest danger of our times. However, I want to add another large danger: fundamentalism on the rise. On the one hand we have the diabolical situation of terrorists who commit horrible crimes with an appeal to values, on the other hand there is the pitfall of fighting them with all our means. There is rightly a lot of attention on beating ter­ rorism, but we must take care not to set aside all human values in doing so, because it is not the case that when terrorism has been wiped out, all the world’s problems will have been solved. Consider the man who says: ‘I can’t take care of my family, because I have to go to war.’ That may be true, one has to set priorities. But if you don’t cultivate your land and you neglect your family, what’s left when you have won the war? I feel that the view in the battle against terrorism is narrowing, with the risk of neglecting values. The EVS-studies show that people have low confidence in official bodies. Does that surprise you? It doesn’t surprise me, but it’s very alarming. It is just plain shocking to realize that trust in the parliament is so low. I think democracy in itself is still considered as a high value, but people have lost faith in its social manifestation: in parties, politics and politicians. The distrust is even no­ ticeable in our language. When someone is called ‘political,’ it is defini­ tely a negative characterization. While political should mean that a per­ son stands up for the public interest, today it refers to a person who ma­ neuvers, who may even cheat and deceive, who is likely to manipulate and to corrupt. The concept of democracy seems to have eroded and that is truly scary. Because what is the alternative? No democracy? Churchill

64

INTERMEZZO

said: ‘Democracy is a damn bad system, but it is the best we have.’ He suggests that politics as we know it is a bad system and I think many people feel the same: politics is a game for politicians who are only looking out for their own interests. Globalization also changes the confidence in administrations. For many people globalization is experienced as a threat, predominantly as a threat to safety and cultural identity. In reaction people put their trust in organizations that stand for law and order: the police and the army. I think these authorities are more appreciated than education, and that is dramatic. The lack of trust in politics could also be the result of the long period of peace we have experienced. After the Second World War men of state were highly valued because of their ability to prevent war, to bring safety and welfare. What can a politician accomplish today? People have heard it all before, many, many times. How can trust in institutions being restored? That’s a question to which I have, unfortunately, no good answer. May­ be the answer is to give politics less power. That sounds anti-democratic, but representative democracy is not the highest good. The highest good is a society that celebrates, lives, its values. That kind of a society cannot exist without some kind of order. Representative democracy is a way of achieving this, but fills only part of the need. Perhaps the best way to protect that democracy against disappointment, and maybe also against corruption and exploitation, is to make politics less important, to dimi­ nish its role. Decision-making should be decentralized; local authorities should gain power. Also, politicians should be involved in fewer mat­ ters. Euthanasia, for example, could better be a responsibility of the medical profession, it is a medical problem. And the more schools and universities organize themselves the better. Capitals and Brussels should steer less, should leave matters more to the people and authorities di­ rectly involved. This demands, however, that all of society should de­ velop and apply their values.

RUUD LUBBERS ON EUROPEAN PRIDE AND PRESIDENTS

65

Prof. dr. Ruud R.F.M. Lubbers (1939) has been High Commissioner for Refugees at the United Nations since October 2000. He was prime min­ ister of the Netherlands for almost thirteen years (1982-1994), a record in Dutch history. Lubbers studied economics at the Erasmus University in Rotterdam. In 1962, immediately after his graduation (cum laude) he joined the management of Hollandia, his family’s engineering works. In the sixties, he was an active member of the Dutch Christian employer’s organization and the Katholieke Volkspartij (a party that later merged into the Dutch Christian Democrats). In 1973, at the age of 34, Lubbers became Minister for Economic Affairs, and in 1982 he was installed as Prime Minister. During his almost thirteen years in the office, he was characterized as a doer, a post-ideologist and a no-nonsense politician. He is renowned for his large dossier-knowledge and negotiation skills. After leaving politics, he became a part-time professor of globalization at Tilburg University. Lubbers is a member of the Club of Rome and one of the initiators of the Earth Charter. He is married and has three chil­ dren and eight grandchildren.

This page intentionally left blank

CHAPTER THREE

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE WIL ARTS, ARNOUD-JAN BIJSTERVELD & KAREL VERAGHTERT Zusammengestohlen aus Verschiedenem, Diesem und Jenen. Karl Popper, Auf der Suche nach einer besseren Welt.

1 Introduction Tradition has it that Jean Monnet, one of the architects of a united Europe, remarked at the end of his life that if he had to start building the ‘European House’ anew, he would begin with culture instead of eco­ nomic and political life. In a certain sense, however, he had started with culture. He was after all the principal draftsman of the ‘Declaration on Franco-British Union’ of the 16th of June 1940. Winston Churchill and Charles de Gaulle, speaking on behalf of the British and ‘free’ French governments, declared in that document their unyielding resolution to defend justice and freedom against subjection to a system which reduces mankind to a life of robots and slaves. Are justice and freedom not the most fundamental and genuine European values? And are these values not part and parcel of European culture? Soon, however, Monnet turned his attention to economic and political issues instead of cultural ones. In the ‘Algiers Memorandum’ of the 5th of August 1943 he reflected on what would follow in Europe after the end of World War II. He argued that there would be no peace in Europe if its states were reconstituted on a basis of national sovereignty with all that implies in terms of prestige politics and economic protectionism. This turn from culture to economics and politics was also reflected in the Treaty establishing the European Coal and Steel Community signed in Paris on the 18th of April 1951. The six countries involved (Belgium, France, Italy, Luxembourg, the Netherlands, and the German Federal Republic) recognized that a peaceful Europe could only be built through practical economic and political achievements. European integration was, however, also at the beginning not simply about coal and steel. The member states were committed to ‘ever closer union’, as it was said in the 1957 Treaty of Rome. The European Economic Community and

68

CHAPTER THREE

Euratom were the result of this treaty. In 1967 the executive organs of the three communities merged and the European Community governed by the European Commission and the European Council was erected. After years of negotiation, the United Kingdom, Ireland, and Denmark joined in January 1973. The poorer countries of Southern Europe took longer to prepare and had to shed first their authoritarian governments. Greece was admitted in January 1981 and Spain and Portugal in January 1986. Much later, in January 1995, the richer but at first reluctant coun­ tries of Austria, Finland, and Sweden joined the European Union. Nor­ way and Switzerland and several islands at the fringe of the continent such as Cyprus, Iceland, and Malta were the only non-communist Euro­ pean countries that remained outside the European Union. With the Treaty of Maastricht of 1991, some important new steps to­ wards further unification were made. The European Community was re­ baptized to European Union and in 1999 the plan for one European cur­ rency was adopted. The European unification clearly went further than opening the economic borders. The Treaty formally empowered the European Union to intervene in some aspects of policy-making in the fields of education, youth, and culture. The EU policy-makers responded to this challenge by promoting awareness of European integration through its education and training programmes, such as Erasmus, Leo­ nardo, Socrates, and Tempus. In a paragraph dealing with culture, it was noted explicitly that ‘the EU wants to contribute towards cultural educa­ tion and especially the common cultural heritage, whilst respecting na­ tional and regional variety. The EU attempts to stimulate collaboration in order to improve the knowledge of culture and history of the Euro­ pean peoples, to protect cultural heritage of European importance, to make stronger the cultural exchanges, to stimulate art and culture, in­ cluding the audiovisual sector’. The unexpectedly fast dissipation of the Soviet Empire at the end of the 1980s and beginning of the 1990s surprised the Western political world and had important consequences for European collaboration. The disappearance of the Iron Curtain in 1989 changed the perspective of the European unification. Suddenly Central and Eastern Europe came within the reach of the European Union. Instead of a geographically limited and more or less tacit anti-communist Western European integration, the European Union could now aim towards a complete and genuine Euro­ pean integration. In 2004 the European Union will probably not only be enlarged further southerly with Cyprus and Malta, but also in an east­ ward direction with several post-communist countries, namely the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia. Other eastern and southern countries such as Bulgaria, Roma­

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE

69

nia, and Turkey are queuing up for membership at a later time. In spite of many beautifully phrased speeches by the leaders of both the new would-be and old actual member states, one cannot escape the conclu­ sion that enlightened self-interest is a more important motivation to join than the attraction of the dream of European cultural unity. There seems to be nothing new under the sun. The founders of the European dream before, during, and after the Second World War were after all mostly in­ tent on rebuilding their own states from the ruins of several decades of world conflict. Theodor Heuss, the first president of the German Federal Republic once said: ‘You cannot use politics to make a culture. You might, how­ ever, use a culture to create a political system’. From the end of the Se­ cond World War to this day most politicians have considered the integration of Europe to be a political, military, and economic necessity. Does this mean that European unity got off to a bad start because eco­ nomic expediency and military fear sooner dictated developments than the desire to reinforce the cultural identity of Europe? Some politicians have answered this question in the affirmative. In line with the remarks by Monnet and Heuss they have argued that successful integration pre­ supposes a common cultural foundation. The pivotal question of the European Values Study is whether there is a common cultural founda­ tion to be found in the public opinions of the populations of Europe at large. The empirical answer to that question is a conditional one. If one means by that question that such a common cultural foundation must re­ veal itself in a unified, idiosyncratic whole, then the answer is in the negative. If one does not place such stiff requirements, then the answer is in the affirmative. If there is cultural unity in Europe, it is unity in di­ versity. Assuming that this answer is an empirically correct one, the follow-up question is how cultural unity and diversity emerged in the course of time. This is the question we will address in this chapter. In answering this question we do not lay any claim to originality. For the cognoscenti, it will simply be a rehearsal of well-known facts, with­ out adding anything really new. For those who are unfamiliar with the cultural history of Europe we hope to provide a birds-eye view of what happened in Europe in more than twenty-five centuries without them getting lost in a blizzard of undifferentiated facts. For the sake of reada­ bility we have decided, in defiance of scholarly decency, to refrain from direct references. There would be too many of them. From the extensive literature we have borrowed a vast amount of insights and tried to weave them together into a short and concise essay. The sources we used can be found in the bibliography at the end of this chapter. It goes without say­

70

CHAPTER THREE

ing that we are greatly indebted to the authors that supplied the insights that we reproduce in this essay. One of the essential parts of European history is that it produced unity out of diversity, while at the same time allowing diversity to continue. In the next section we will start our search for signs of this cultural unity in diversity and the cultural diversity of this unity by a one-sided quest for the emergence of cultural unity in the history of Europe. In the following section we will especially look out for cultural diversity, but without los­ ing unity completely out of sight. In the concluding section we will dis­ cuss, as a bonus, the implication of cultural unity and diversity for the enlargement of the European Union. Have the successive enlargements watered down European cultural identity and will it get even thinner by the coming expansion? Will this diminish the probability that the Euro­ pean Union will become an historical success story? And if so, what will be the further unintended consequences of diluting the European cultural identity? 2 The unity of European diversity Providing an answer to the question of whether, and if so, when, where and how a more or less common cultural European identity emerged from history has to start with defining the term Europe. The word Europe refers both to a region and an idea, but sooner to a region than an idea. The word Europe has always been and still is first and foremost used as a geographical term. Geographical Europe has, however, always had to compete with notions of Europe as an idea, an image of a set of societies that can be seen as sharing a more or less general cultural pat­ tern. The societies and cultures that have existed in this western extrem­ ity of the Eurasian landmass have, however, always been highly diverse, and the case for grouping them together culturally as ‘European’ has va­ ried from period to period. The question of whether the appellation ‘European’ designates more than a mere location on the map of the world is a pivotal one for the European Values Study. European borders have fluctuated, institutions have waxed and waned, nations have not only been formed and reformed, but also disappeared. Europe as a geographical-cum-cultural entity has in the course of history been a rather diffuse phenomenon. Geographically it has been a phenomenon difficult to mark off through the ages and historical periods. Europe ended in the Germanic forests at one time and in the Urals at another. Culturally the case is even more problematic. It began as a myth, the abduction of Eu­ ropa, daughter of king Agenor of Tyre, and it evolved into a way of life based on the sense of belonging to a common civilization. This mentalité

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE

71

collective developed through the civilizing process itself, the shared ex­ perience of living under Roman law, Christian religion, and the secular culture developed in the Age of Enlightenment. European civilization, however, has not shaped the values of every inhabitant in the same way and with the same depth in the course of time. Still, if there is at least some unity in European diversity it has to be found in the shared experi­ ences of Europeans in each of the great epochs of Europe’s past. From the perspective of the quest for a specific European cultural identity Europe can best be conceived of as a pseudo-geographical con­ cept made up of a number of traditions. These traditions can be isolated in their chronological appearance on the European scene. They origi­ nated most of the time in the south or the west of the continent to be dif­ fused gradually to the north and the east. Although each one of them may not be specifically, uniquely European, together they constitute a coherent culture, a heritage. The word Europe first came into use in ancient Greece in the seventh century B.C. Many see the primeval roots of the term Europe in the Ak­ kadian word erebu, which means twilight or sunset. Asia, on the other hand, is believed to have derived its name from the Akkadian term asu meaning daybreak. Europe at that time was conceived of as an unspeci­ fied area around the Peloponnese and the Greek islands. By that time, the coasts of Southern Europe, Asia Minor and North Africa, together with the neighboring island groups in the Atlantic, were known in their general configuration, and some progress had been made in the task of fixing these places to the sphere. The geographical outlines had, how­ ever, not accurately been ascertained, and the longitude of the united terra firma of Europe and Asia was greatly over-estimated. What later in geographical terms became the European continent got its cultural iden­ tity by way of trial and error, i.e., via learning by doing, but also through the diffusion of values from the cultures that were dominant on the coastline of the eastern Mediterranean in Antiquity. The cultural legacy of Antiquity has come from three treasure houses: the ancient Near East, Greece and Rome. In Israel the religious founda­ tions were laid for what later became Christian civilization. Christianity absorbed much of this Judaic heritage. Not only monotheism, but also the values embedded in the Old Testament. Ancient Greece has influ­ enced European civilization especially by way of its philosophy and ra­ tionalism. Rome had an impact via both political principles and syste­ matic law. In Greece in the fifth century B.C. during the Persian Wars, Europe as a geographical term acquired for the first time a cultural connotation. It was at that time mainly identified with freedom and democracy. The

72

CHAPTER THREE

Greeks thought that Europe—bounded by the Mediterranean and the Atlantic—was strong because its civilization was concentrated in free, in­ dependent, and democratic city-states. This identification of Europe with freedom and democracy, however, remained an incidental relationship. For centuries afterwards, the term Europe was rarely used except do de­ note a geographical landmark. The second time that a European cultural identity in statu nascendi made its appearance was in classical Rome. The Roman Empire was an intensely legal society. In the course of several centuries Roman jurists worked out a system of legal principles devised to protect both life and property. The decline and fall of the Roman Empire obstructed, how­ ever, the emergence of a real European cultural consciousness. This did not prevent Roman law to provide, since its rebirth in the Middle Ages, the basis of the law in large parts of Europe to date. Its legal principles are still parts of European cultural identity. From the perspective of Ro­ man law the pivotal values to be cherished are citizenship and individual rights, equity and the common good. It was in late Antiquity that European history ceased to be an assort­ ment of unrelated events within the given Peninsula and began to take on the characteristics of a more coherent civilizational process. Central to this process was the merging of the classical and the Judaic legacies with acquisitions from the barbarian worlds. A unifying cultural framework emerged, which appeared to be amendable to further differentiation and remained in the long run open to new interpretations. The result was not only an assertion of a consciously Christian community—in other words the founding and crystallization of Christendom—but also a reassertion of internal pluralism. The privileged role of Christendom did not exclude inputs from more peripheral or subaltern sources. The Middle Ages are the era of the gradual expansion of Christen­ dom. At the same time, however, the legacy of Greek and Roman Antiq­ uity was preserved. Thanks to Muslim thinkers and their stewardship of the inquisitive spirit of Greek science and philosophy, and also thanks to the pivotal contribution of early Muslim librarians in preserving the an­ cient texts. Not only a culture shift, but also a geographical shift oc­ curred in the Middle Ages. Western Europe had till the eleventh and twelfth centuries been a peripheral region of the Mediterranean-based culture of classical Europe. Now a new-cum-old civilization emerged in those parts. A resurgence of interest in the culture of the classical Greek and Roman worlds occurred at the court of Charlemagne and can be noted even more markedly from the twelfth century onwards. By 1300 Europe existed for the first time in history as an identifiable cultural en­ tity. By the fourteenth century a large part of Europe, including England,

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE

73

France, Germany, the Low Countries, Scandinavia, and the northern part of Italy and Spain, had come to possess a relatively high degree of cul­ tural homogeneity. The church played a key role in establishing this homogeneity by way of a cultural ecumene, self-defined as Christendom and strengthened but never fully controlled by the papal monarchy that coexisted with a plu­ rality of political centers. Especially Christian philosophy, resting on Greek foundations, and Canon law, resting on Roman foundations, played an important unifying role. A distinct feature of European society during the Middle Ages was the close alliance of ecclesiastical and secu­ lar power. The close unity between the Pope on the one hand and Em­ peror and kings on the other was regarded as the foundation of both spiritual and worldly authority. Although successive clashes between church and state resulted in the recurrent redefinition of their interde­ pendence, the alliance was never completely challenged until the Age of the Enlightenment. Both cultural unity and political pluralism were cru­ cial to the rise in the Middle Ages of autonomous urban communities as a new civilizational force. These communities gave inter alia birth to the universities and rebirth to Roman law. The universities made admirable attempts to unite the Christian-Judaic theological tradition with the Greek tradition of rational strife for knowledge. The rise of merchant capitalism in the urban communities, the emergence of worldly and ec­ clesiastical bureaucracy, and the expansion of the written word promoted the use of Roman law for worldly and ecclesiastical use. It was not until the fifteenth century that Europe came to be really identified with Christendom. In this century, Europe completed a pro­ cess, which integrated strong Christian moral beliefs with the GrecoRoman values of Antiquity. Christianity became not only the heir of the Roman Empire’s Greco-Hellenistic and Latin culture, but also added to this the moral value that protecting only oneself would not result in a humane society. The biblical values—faith, hope, and charity—were merged with the cardinal virtues of Greco-Roman culture. This process of cultural ‘transformation’ and ‘convergence’ caused a dramatic change in what was shared in Europe and how widely this was shared. The spread of the Christian culture took place not only through influence and cultural diffusion, but also through conquest and forced assimilation. Whereas in the Middle Ages ‘Europe’ was equated with ‘Christen­ dom’ in Modern Times ‘Europe’ was put on a par with ‘Civilization’. ‘Europe’ in the sense of ‘Civilization’ is therefore a relatively new idea. It gradually replaced the earlier concept of ‘Christendom’ in a complex intellectual process lasting from Renaissance humanism to Enlighten­ ment rationalism. In the Renaissance with its passion with Antiquity a

74

CHAPTER THREE

new vision of man was being created: sovereign in the world and, with his reason and creative powers, also able to penetrate any secret, make anything he invented. This vision was broadened and gained more indepth knowledge through the scientific revolutions of the seventeenth and the Enlightenment of the eighteenth centuries. Humanism added to the biblical virtues of Christianity the singular importance of religious tolerance. Without losing their beliefs in the basic tenets of Christianity humanist thinkers emphasized more individualistic and more secular values than their scholastic predecessors did. The Treaty of Utrecht of 1713 provided the last major occasion on which public reference was made to the Respublica Christiana, the ‘Christian commonwealth’. After that, the awareness of a European as opposed to a Christian community gained the upper hand. During the Enlightenment the term Europe was for the first time no longer treated as a quasi-geographical concept but became more and more identified with civilization. Voltaire, the most successful author of the eighteenth century, stated very emphatically in 1751, that Europe shared common morals and customs. He described Europe as a kind of great republic divided into several states, some mo­ narchical, some republican, the others mixed, but all corresponding with one another. They all had the same religious foundation, even if divided into several confessions. They all had the same principle of public law and politics, unknown in other parts of the world. So he argued. Two vast upheavals catapulted Europe into modernity: the French and the Industrial revolutions. These not only led to the rise of industrial ca­ pitalism and parliamentary democracy, but also to the transformation of religious beliefs into economic and political ethics. They also gave birth to new secular ideologies such as liberalism, socialism, and nationalism. The most important of these ideologies, also acting as an umbrella ideo­ logy, was that of democracy. The democratic ideals (Liberté, Egalité, Fraternité) of the French Revolution revived and strengthened not only the ideals of the Dutch and English Revolutions of the sixteenth and se­ venteenth century, but also those of Greek Antiquity. The ‘Declaration des droits de l’homme’ on the 26th of August 1789 stressed the equal rights for all citizens, the right of well-being and happiness, the right of education, the right of a dignified existence. The French Revolution brought not only a new secular ideology, but also a new political culture, a practice of efficient and rational governing of the institutions in the ar­ eas that Napoleon had conquered. In this new governing practice, mo­ dernization, and rationalization were priorities. The feudalistic legal structure made way for democratic structures wherein justice was done based on general laws that applied to all people (Code Civil 1804, Code Commerce 1807, Code Pénal 1810). A systematic and efficiently organ­

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE

75

ized administration supported this system. The abolition of privileges and obligations (taxes) certainly helped to make the reforms effective. In some parts of Europe, citizens, sometimes helped by the French army, cleared away the old regime. In other countries, the hierarchical ordered society of the old regime was not so easily deposed and democratic prin­ ciples would seep through to national legislation only slowly and par­ tially. After the French revolution the term Europe was used much more consciously than before. Europe had a familiar ring for centuries, but now came to be endowed with historical interpretations and political ideas. An elaborate consciousness of the history of Europe emerged in which freedom and liberty were the orienting ideas. One of the reasons for this increased historical awareness of cultural unity was a growing sense of nationhood and therewith a growing sense of cultural diversity. The nineteenth century was a period in which increasing seculariza­ tion and religious indifference battled with revival of religion. A deci­ sive victory of secular over religious ideologies could be observed in the twentieth century. During the nineteenth century, values like constitu­ tionalism, sovereignty of the people, the separation of powers, the sepa­ ration of church and state, legal equity, and civil liberties spread throughout Europe. The ‘modern’ ideal of the equality under law of all people ensued not only from a strengthening of democratic ideas, but also from Christian ideals. The most influential Christian idea in the his­ tory of Europe has probably been that of the equality of every person be­ fore God in heaven. This latter idea gained momentum from the tenth century on especially through the urbanization process and the rebirth of Roman law. In an early stage cities were open, relatively egalitarian so­ cieties. Citizenship gave rise to ideas of social solidarity and revived the ideas of individual rights, liberties, and citizenship. In the nineteenth century a specifically national sense of citizenship emerged. Not only civil and political rights were more and more emphasized in the context of the European nation states of the nineteenth century, but also social rights. The result was the rise of the welfare state of the twentieth cen­ tury, which came to embody not only the idea of social solidarity, but also that of social justice. A European needs not be a Christian him or herself to accept the fact that all he or she does, makes and says depends to a high degree on the Christian heritage. It is also through Christianity that Europeans can trace back their common descent to Greece and Rome. From the moment that the call for political democracy became stronger the history of Eu­ rope was carried back also directly to Athenian democracy and Greek freedom as a memorable starting point. The various identifications of

76

CHAPTER THREE

Europe with liberty, with Christendom, and with civilization were not only rediscovered in the nineteenth century, but also reassessed and gi­ ven unprecedented prominence. The concept of a common European cul­ ture endured, whereas Europe broke up into nation-states and national cultures. In the period 1880 to 1914 nationalism took a dramatic leap forward. A novel tendency was discernable to define a nation in terms of ethnicity and especially in terms of language. People proved increasingly ready to identify themselves emotionally with their nation and to be po­ litically mobilized. Nationalism and patriotism became increasingly val­ ues of the right. They found their most extreme expression in the fascist and national-socialist ideologies of the twentieth century. Nationalism appeared to be a very destructive force in this century, mobilizing masses and hurling them against each other. Europe tore itself in pieces. Yet surprisingly, European culture was still regarded as a shared entity. Ortega y Gasset, for example, claimed that in spite of its national diffe­ rences, Europe was really one large family. It was not until the First World War that the simple equation of Europe with civilization ceased to exist as a dominant idea. The great edifice of nineteenth-century civilization crumpled in the flames of the First World War, as its pillars collapsed. Europe went through a deep crisis that dragged on through the depression of the 1930s and the hor­ rors of the Second World War. In this period of prolonged crisis, of con­ frontation, of war the need was felt by many to reflect again on the cul­ tural identity of Europe. After a war borders are often redrawn for the benefit of the winners as happened after the First World War. Changing the European map using the principle of nationality meant that Europe became like a puzzle with ever more pieces. As a result of the repositioning, Europe was more di­ vided than ever. Finland, Estonia, Latvia, Lithuania, Poland, Czechoslo­ vakia, and Yugoslavia were born or reborn as nations. International commissions of experts had to resolve discussions about borderlines. Endless discussions about areas inhabited by several minorities had star­ ted whilst Russia, Germany, and Austria were waiting for revenge be­ cause they had had to give up parts of their territories. The League of Nations was erected which held national sovereignty as an untouchable principle. During the 1930s under the auspices of this powerless League it was tried in vain to take a stand against rising nationalism. The Euro­ pean optimism that existed before the war changed during the inter-war period into negativity about the ‘Untergang des Abendlandes’. Just like the League of Nations, the initiative of the French homme d’etat Aristide Briand, who had been working since the Treaty of Ver­ sailles to reconcile France and Germany, produced no results. In 1929,

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE

77

he proposed to all European member-states of the League of Nations to discuss the creation of a European Federation. The memorandum that Briand designed following the request of the League of Nations in 1930 came too late. Against the backdrop of the 1929 world economic crisis there was an increasing flood of fascism and nazism while the Commu­ nists strived to create a world-wide revolution. Thus, the initiative of Briand quietly died in the years leading to the Second World War. After the Second World War, Briand’s dream of European unity see­ med crushed forever. For years, European states had been fighting a bit­ ter war destroying the lives of millions of soldiers and civilians. The war left terrible scars between European states and their populations and fu­ ture generations did not seem able to cooperate any longer. On top of this the Cold War started and an Iron Curtain descended splitting Europe from North to South. Churchill stated on the 5th of March 1946 at West­ minster College in Fulton (Missouri) of the Iron Curtain: ‘From Stettin on the Baltic to Trieste on the Adriatic, an iron curtain has descended across the continent. Behind that line, lie all the capitals of the ancient states of Central and Eastern Europe – Warsaw, Berlin, Prague, Vienna, Budapest, Belgrade, Bucharest, and Sofia. This is certainly not the libe­ rated Europe which we fought to build up’. At the same time the Iron Curtain also meant a challenge for non­ communist Europe to partially set aside the fatal nationalism. The split of Europe stimulated a concrete and successful supranational collabora­ tion. In Zürich on the 19th of September 1946, Churchill appealed ‘If we are to form a United States of Europe (...) we must begin now’. Chur­ chill called the Christian beliefs and the Christian ethics the foundation of Europe and pointed out the importance of the common cultural heri­ tage. The movement for European unity was fired by an idealism that aimed to remove the welter of ultra-nationalistic attitudes, which had fu­ elled the conflicts of the past. As a result of the European Congress of The Hague (7-10 May 1948), strongly influenced by Churchill, the Council of Europe was established on the 5th of May 1949. The Council of Europe provided the forum within which most early discussions took place. As an organization supported by twenty-four governments in Western Europe, it was never constrained by the political horizons of ei­ ther the EU or NATO; and in the cultural field it gained the cooperation of four non-member countries from the Soviet bloc, Poland, Czechoslo­ vakia, Hungary, and the USSR itself. From this beginning the Council of Europe has expanded to 43 member states, and now stretches from Ice­ land and the Republic of Ireland in the West to the Russian Federation and Azerbaidzhan in the East. The role of the Council of Europe is to prepare opinions and resolutions on issues relating generally to the func­

78

CHAPTER THREE

tioning of democracy, the ‘cultural identity’ of Europe and the rule of law, and to ensure that member states abide the agreements reached. Mainly because of British influence, the Council of Europe did not get any supranational allure. This institution thus developed into a useful but powerless forum for exchanges of thoughts on European issues. While the Council of Europe may not have the broad public profile of the European Union, it has nevertheless made some significant contributions to social and political life in Europe. Arguably the Council of Europe’s most significant contribution was the European Convention on Human Rights, signed in Rome on 4 November 1950, which paved the way for the establishment of the European Court of Human Rights, based in Strasbourg, in 1959. 3 The diversity of European unity From an historical perspective, it is amazing how much Europe has achieved in the last fifty years. In 1945 Europe was a wrecked and di­ vided continent. In the following decades peace was established, a re­ markable economic recovery took place and the struggle against com­ munism was won. Parts of these successes were due to increasing Euro­ pean unity. It is also amazing how far the European Union has come in less than half a century since the signing of the Treaty of Rome. During each decade since the seventies new countries have joined the European Union and new community institutions were founded. In a few years from now, almost half a billion Europeans will be united in one eco­ nomic and political framework around European symbols such as the s Ninth flag, the Euro-currency and the ‘Ode to Joy’ from Beethoven' Symphony, the unofficial anthem of the Union. If we follow this course still further in the twenty-first century the European Union will inevita­ bly lead to a complete unification of Europe. The European dream has then come true. Dreams are, however, deceiving. In Europe as a continent, as a geo­ graphical unit, there are still many internal divisions. It is a complicated mosaic of old and new natural, political, and religious dividing lines that separate people. Old dividing lines are tenacious: even when they no longer have any physical, political or religious reality, they continue to influence the ideas and identity of Europeans. They create a multiplicity and diversity of culture. They create cultural fault lines, which are the products of historical and contemporary borders between states, worldviews, ideologies, languages and religions. One could therefore argue that not European cultural unity, but diversity is de rigueur. Has there ever been a real common European cultural identity that ultimately

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE

79

rested on a sufficiency of shared values, culture and history or did diver­ sity prevail? In this section we will look into that question. It is impossible, however, to treat the history of European cultural di­ versity satisfactorily within the scope of this essay. It is, nevertheless, necessary to give at least some informative examples of historical devel­ opments that led to cultural diversity. Of prime importance is the rise of nationalism. From the twelfth century onwards some European peoples, for example, the English and the French, started to develop a national consciousness. These national sentiments were founded on a strong sense of their own military, socio-economic and cultural superiority and a deliberate cultivation of hostility towards other ‘nations’ within or out­ side the realm. This process of national identification is closely linked to the process of state formation. It took a very long time to cover the dis­ tance between the ‘barbarian’ kingdoms and chiefdoms of the early Middle Ages through the feudal principalities and city-states to the stri­ king diversity of state-like political forms in the late Middle Ages. Around 1500 Europe’s 80 million people were divided into something like 200 states, would-be states, statelets, and statelike organizations. Despite their variation, these states showed a similar pattern of national identification and internal organization, which would endure for ages. These social entities developed from the ‘imagined communities’ of the Middle Ages into the ‘constructed communities’ of modernity. The French revolution gave birth to a new political order based on the princi­ ple that the people make the nation and that this nation forms the only source of political legitimacy. The result was that the dynastic states of the Middle Ages were replaced by the nation states of Modern Times with their institutions, a central government, and a shared national ideo­ logy. Values such as patriotism and national pride became inextricably bound up with these national ideologies. Within Europe from 1500 till Modern Times a trend has been dis­ cernible towards fewer and more encompassing states. This development was attended with a metamorphosis of nationalism. Never before in European history were cultural divisions so vigorously emphasized as with the birth of modern nationalism in the nineteenth century. On pseudo-historical and pseudo-linguistic grounds nineteenth-century na­ tionalists all over Europe assumed that the peoples of Europe could be distinguished as objectively identifiable social and cultural units on the basis of language, religion, custom, and national character, which were supposed to be unambiguous and immutable. In addition it was taken for granted that ethnic claims demanded the political autonomy of all per­ sons belonging to a particular ethnic group and at the same time the right

80

CHAPTER THREE

of that people to its historic territory. Where no ‘natural’ boundaries could be found, these were invented and imposed. The tide of European nationalism obscured for at least a century the unity of European culture. The nineteenth century gave birth to idiosyn­ cratic national cultures, which under the aegis of strong national govern­ ments even developed antagonistic schools of literature, music, and visu­ al arts, which previously had been the hallmarks of a supranational European culture that could refer to its Greco-Roman and JudeoChristian heritage. Even the most supranational of Europe’s cultural fea­ tures, the academic world of the universities, fell apart in supposedly ir­ reconcilable scientific traditions, which complied with linguistic and na­ tional boundaries. Another example of increasing diversity in European history is that of religious life. Although Europe at a certain point in time became identi­ fied with Christendom, there are strands in European cultures that are definitely not Christian but Islamic or pagan and within Christianity there have been many schisms. The Great Schism of 1054, the formal separation of Latin and Orthodox Christendom, resulted in the first ma­ jor religious dividing line within Europe after the conquests of large parts of the Mediterranean by the Islam. This schism had enduring po­ litical and cultural implications. During the sixteenth and seventeenth century, another religious watershed, a product of the confessionaliza­ tion process in the aftermath of the Reformation, added up to the diversi­ fication of Europe’s political entities. While Europe in the Middle Ages can be characterized as deeply religious, between 1560 and 1650 Europe’s history was shaped by ‘confession’: the modern variant of Christendom defined by an explicit statement or doctrine. In the ensuing process of confessionalization the four confessions in Western and Southern Europe—Lutheran, Calvinist, Anglican, and Roman Catholic—were moving on parallel but separate tracks, though at different speeds, towards the same goal: the regulation of religion and society in a Christian sense. Based on their respective confessions of faith, the prin­ cipal denominations developed into internally coherent and externally exclusive communities distinct in institutions, membership, and belief. Each in its own way, these churches entered into alliances with the early modern states, the coercive power of which enhanced the churches’ abi­ lity to manage religion, though with results that usually lagged far be­ hind their aims. In most European states, the pursuit of religious unity was explicitly regarded a legitimate tool of achieving social and political cohesion (cuius regio eius religio). The absolutism and the enlightened despotism of the seventeenth and eighteenth centuries presupposed a

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE

81

one-religion people under a king who was regarded as the defender of the faith. Although political and religious developments, i.e., state-formation and schisms, strengthened each other and led to an increasing cultural diversity of Europe, there is at least still some unity in this diversity dis­ cernable. A rich reservoir of ideas is linked to Europe—Greco-Roman, Judeo-Christian, Humanist, Rationalist, and so forth—but there is no such thing as an unchanging core of European culture, no fixed Euro­ pean identity. Still there is a mysterious thin thread, which connects these ideas, which we might call in default of a better description Euro­ pean culture and identity in flux. To gain insight into the unity of European cultural diversity and the diversity of European unity we need to attain a good understanding of the complex history of shifting fault lines. Such an understanding is also crucial for understanding the results of the European Values Study sur­ veys. Fault lines should be seen as the borderlines on the map of Euro­ pean values. They are the marks of cultural diversity. The most obvious dividing line is the one that separates Western Europe from Eastern Europe, with a wide transitional zone, sometimes called Central Europe, stretching from the Baltic to the Balkans. Yet one has to insist that the West-East division has never been fixed or permanent. Probably the most durable is the line between Catholic (Latin) Christianity and Or­ thodox (Greek) Christianity. There is also the line between the Western Roman Empire and the Eastern Roman Empire and there is the Ottoman line, which marked off the Balkan lands, which lived for centuries under Muslim rule. Economic historians see a line separating the industrialized countries of the West from the peasant societies of the East. Historical anthropologists have identified a St. Petersburg-Trieste line, which sup­ posedly separates the zone of nuclear families from that of the extended family. Most recently there was the Iron Curtain. All these lines have more or less profoundly affected the values of Europeans. The division of Europe into two opposing halves, therefore, is not entirely fanciful. It rides, however, roughshod over many other lines of division of equal importance. It ignores serious differences both within the West and within the East and it ignores the strong historic division between North and South. There is, for example, the line of the Roman limes, dividing Europe into one area with a Roman past and another area without it. There is a division between the region where the Romance languages were spoken, and the part where Germanic or Slavic was the basis of the languages. Legal historians trace a line separating the lands which adopted forms of Roman law and those that did not. In the fifteenth and early sixteenth centuries the Church of Rome lost its monopoly of power

82

CHAPTER THREE

in Latin Christendom. The actions of Luther, Calvin, and other reformers created a fundamental split between Catholicism and Protestantism crea­ ting different cultures. The Reformation, Calvinism in particular, did re­ emphasize Judaism with its strong focus on the Old Testament. Constitu­ tional historians emphasize a line dividing countries with a liberal, de­ mocratic tradition from those without. Political scientists have found a line dividing ‘Western’ and ‘non-Western’ forms of nationalism. Taking the full range of factors into consideration one can only con­ clude that to see Europe’s cultural fault lines one should not divide Europe into two regions, but at least into four or five overlapping ones. Despite their differences all the regions of Europe still hold a great deal in common. They are co-heirs of Christendom and Antiquity, of the Enlightenment and Nationalism. Every sort of political, economic, and cultural overlap and interaction connects them. Despite their own an­ tagonisms, they share fears and anxieties about influences from outside – whether from America, from Africa, or from Asia. Fundamental unity is no less obvious than manifest diversity. 4 Conclusion and discussion One could ask whether the past identifications of Europe—the geo­ graphical definition, the equation with political freedom, the association with Christendom, the connection with civilization—still have a con­ temporary relevance now that we have entered the twenty-first century. Is it really a precondition for the stability of a European Union that it is rooted in a cultural unity? If so, is the European project then destined to come to grief on the rocks of increased and increasing national and re­ gional diversity? Is it not better to gear the legitimacy of the European Union, as the latest manifestation of a European project, rather to the fu­ ture than to the past? The ideals and values that catched on in the history of Europe, such as freedom and tolerance, have often been lacking and neither have been nor are yet fully realized in practice. On the contrary, Europe has often been untrue to its own heritage. Therefore, it would be wise to reassess the idea of European unity in the light of its cultural history, with due regard for both Western and East-Central Europe. This means that justice should be done not only to Europe’s Judeo-Christian origins but also to its Orthodox (Byzantine) variants and to Islamic (Ottoman) strands. They all share the conscious­ ness of humanist values such as the emphasis on the word as exponent of truth, morality (guilt) and (work) ethics, and remembrance as the basis of culture and identity. This historical reassessment will bring about tol­ erance and respect for cultural diversity. These are the only means to es­

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE

83

cape from the pendulum movement which seems to have determined a millennium of Europe’s history, that is a between an open-minded, ‘uni­ versalist’ or international approach versus a conceited nationalistic or re­ gionalistic orientation. Two centuries of nationalism and its exclusive focus on national or even regional values have obscured the apparent pluriformity in every culture, be it on regional, national, or international scale, as well as the values shared across national and other boundaries. In 1932 Joseph Roth described the pre-First World War Austrian Dual Monarchy as ‘a big house with many doors and rooms, for diffe­ rent kinds of people.’ Our continent, in a way, was and still is like Roth’s big house, with many doors and rooms, in which everybody is talking at the same time in fifty languages. No economic development, even no external foe and even less a centralist government will ever be able to transform the European house into one big space with no divid­ ing walls. Instead, knowledge of and due respect for cultural, linguistic, religious, and historical differences will have to stand at the heart of any dialogue or debate on European values and cultural unity. This debate, based on the completely free exchange and traffic of people and ideas, will bring to light the foundations of cultural unity on the one hand and foster a new basis for accepting and respecting cultural diversity on the other. As history has shown, the pretensions of European civilization and its values have not prevented the horrors of war and the systematic oppres­ sion and even extermination of fellow human beings. This happened not only because boiling passions and worldly greed eclipsed morality, but also because European values were overzealously put into practice. One could, for example, refer to the crusades, the inquisition, religious wars, the persecution of heretics, witch hunts, auto-da-fés, the holocaust, the gulag archipelago and so on and so forth. Therefore, without diminishing the great value of Europe’s cultural inheritance, it is necessary to evalu­ ate past and present practices thoroughly to preserve the heritage in as pristine a condition as possible for the future. To provide Europe with a set of common values, with shared ideals, is one of the greatest chal­ lenges for the European Union in the years to come. Another challenge is that in this common value pattern cultural, political, and religious dif­ ferences are allowed their place and that it is open, that it is a culture that learns from others. Otherwise those cynics who argue that the continuing debate on European values is a continuation of war by different means will be proven right. Still there is a long way to go. Clearly European Union citizens’ iden­ tification with ‘Europe’ lags considerably behind attachments to their re­ spective national states. Data from EU surveys, such as the Eurobarome­

84

CHAPTER THREE

ter, and the European Values Study show that the European Union has yet to emerge as contender for the loyalties of EU citizens. The notion of ‘l'Europe des états’, a phrase first employed by Charles de Gaulle, ar­ guably remains the preferred vision of ‘Europe’ for most Europeans. It seems unlikely that a European identity will generate the sorts of pas­ sions and loyalty that people feel towards their nation state. For the mo­ ment Europeans are citizens of the European Union because they are first and foremost citizens of one of its member states. The European Union is not the end of the nation state but membership of the European Union is no longer without consequences and has a growing importance in the daily lives of its citizens. In the future, one might start to wonder what justifies the existence of a national state within a European Union with its own legislation, its own government, its own currency, and its own army. For the time being, however, the European Union does not threaten the existence of the national states. Rather than an integrated Europe, the future could be a multi-layered Europe with a range of inter­ nal conflicts and strains embedded within it. Not only myopian selfinterest is to blame for those tensions, but also cultural differences. The final question is, of course, which of its shared values will prove strong enough to give Europe cohesion in the years to come. It is impor­ tant to realize the historical selfishness and resistance of the individual states. Will the values and ideological fundamentals that Europeans are so proud of such as freedom, tolerance, parliamentary democracy, de­ fense of human rights, freedom of speech, and equal opportunity be strong enough to peacefully unite European countries that have fought two bitter and bloody wars over the last century? Will the European Un­ ion overcome the oppositions between Catholic Christians, Protestant Christians, Orthodox Christians, and the Islam? Entanglement occurs daily. Borders are being opened to a free exchange of goods, capital, and labor. Not only custom booths are being torn down, mental barriers as well are being removed. After joining the European Union, it seems im­ possible not to stay within it. Apart from Greenland, no member state or part of it has ever left the European Union. On the contrary, the territo­ ries of four out of the five big European powers of the eighteenth cen­ tury, France, Great Britain, Prussia, and Austria have been united at the start of the twenty-first century. One might wonder whether, and if so when Russia will follow and even whether Europe will then halt at the Ural. Even further than that, may the old Ottoman Empire, that was for a long time so important to the Balkans, find its way back into Europe with the possible membership of Turkey? Or will Europe then enter Asia via Turkey and the Bosphorus as once Alexander the Great and the Ro­ man army did?

EUROPE AND ITS VALUES IN AN HISTORICAL PERSPECTIVE

85

Nobody can tell, for the future is shrouded in mystery. One thing, however, is certain and that is something the ancient Greeks already knew: panta rhei, everything is in motion, comes and goes in the course of centuries. There is no such thing as the end of history resulting from processes of ongoing modernization and convergence. Although mod­ ernization tends to push value patterns in a common direction, the spe­ cific cultural heritage of a society leaves an imprint on value patterns that endures despite modernization. So the cultural diversity of European unity seems to be a never-ending story. There is, however, always such a thing as the rise and decline of cultures. Nothing, not even Europe and its values, is forever. References Barnard, B. 2002. Tegen de draad van de tijd. Over de ware aard van Europa. Huizinga-lezing. Rotterdam-Leiden: NRC-Handelsblad-Universiteit Leiden-Maatschappij der Nederlandse Letterkunde. Bartlett, R. 1993. The Making of Europe. Conquest, Colonization and Cultural Change 950­ 1350. London: Allen Lane. Blockmans, W. & P. Hoppenbrouwers 2002. Eeuwen des onderscheids. Een geschiedenis van middeleeuws Europa, Amsterdam: Prometheus. Bosch, A. & L.H.M. Wessels 1992. Veranderende grenzen, 1919-1989. Nationalisme in Euro­ pa. Nijmegen: SUN. Brady, Th.A., H.A. Oberman & J.D. Tracy (eds.) 1994 & 1995. Handbook of European History 1400-1600. Late Middle Ages, Renaissance, and Reformation. Vol. 1 & 2. Leiden: Brill. Davies, N. 1996. Europe. A History. London: BCA. Den Boer, P. 1997. Europa. De geschiedenis van een idee. Amsterdam: Prometheus. Dunkerley, D., L. Hodgson, S. Konopacki, T. Spybey & A. Thompson 2002. Changing Europe. Identities, Nations and Citizens. London: Routledge. Eisenstadt, S.N. 1987. European Civilization in a Comparative Perspective. Oslo: Norwegian University Press. Geary, P.J. 2002. The Myth of Nations. The Medieval Origins of Europe. Princeton: Princeton University Press. Hobsbawm, E. 1962. The Age of Revolution 1789-1848. London: Weidenfeld & Nicolson. —— 1975. The Age of Capital 1848-1875. London: Weidenfeld & Nicolson. —— 1987. The Age of Empire 1875-1914. London: Weidenfeld & Nicolson. —— 1994. The Age of Extremes 1914-1991. London: Michael Joseph. Huntington, S.P. 1997. The Clash of Civilizations and the Remaking of World Order. London: Simon & Schuster. Leerssen, J. 1999. Nationaal denken in Europa. Een cultuurhistorische schets. Amsterdam: Amsterdam University Press. Moore, R.J. 1977. The Origins of European Dissent. London: Allen Lane. Paling, K & V. Veldheer (eds.) 1998. The European Challenge. Essays on Culture, Values and Policy in a Changing Continent. The Hague: Vuga. Rietbergen, P. 1998. Europe. A Cultural History. London & New York: Routledge. Salmon, T. & W. Nicoll 1997. Building European Union. A documentary history and analysis. Manchester & New York: Manchester University Press. Tilly, Ch. 1992. Coercion, Capital, and European States, AD 990-1992. Cambridge MA: Blackwell. Wilson, K. & J. van den Dussen 1995. The History of the Idea of Europe, Vol. I-IV, rev. ed.. Milton Keynes & London: The Open University & Routledge.

This page intentionally left blank

INTERMEZZO

HERMAN WIJFFELS ON THE VOCATION OF EUROPE: ‘WE ARE GIVING BIRTH TO THE ORGANICALLY ORDERED SOCIETY’ INTERVIEW BY MARGA VAN ZUNDERT

Europe is leaving the age of industrialization and its hierarchical insti­ tutional structures behind, affirms SER-chairman Herman Wijffels. In its place an organic Europe is growing, ‘a biotope,’ that only acts when its citizens need coordination for a higher quality of life. ‘Those who want to shape Europe into a US look-alike have it all wrong.’ Herman Wijffels is the farmer’s eldest son who didn’t follow his ‘des­ tiny’. He left the family farm to study economics and made a splendid career in management. Four years ago, after heading the Rabobank for almost fifteen years, he became the main advisor of the Dutch govern­ ment as chairman of the national Social and Economic Council. Wijffels is also a EU veteran, and forty years of closely watching Europe form have not turned him into a skeptic. On the contrary, Wijffels thinks Eu­ rope is setting a unique example on how to live together on this planet in peace, prosperity and cultural multiformity – ‘maybe without actually realizing it.’

Mister Wijffels, do you believe that there are European values? ‘Yes, I do. Where we come from is clear: classical antiquity, Christian­ ity, Renaissance, the Enlightenment (…). But I also consider it from the result: What is Europe’s role in the world? Or as the French would say vocation? I’m a European veteran; after my graduation in 1966 I fol­ lowed a traineeship at the European Commission in Brussels and later I was a EU negotiator for the Dutch government for eight years. In the past forty years, Europe has not done badly, all in all. I see a new Europe forming, a society on this old continent that strives towards peace, pros­ perity and cultural multiformity. The latter is especially important, as Europe has to be a concept in which every citizen can feel at home. This new Europe that is developing lacks a classical, central power structure; instead it is an intelligent combination of large and small-scale organiza­ tions. I hope and believe that this concept proves viable: cooperating on

88

INTERMEZZO

a large scale when needed, but leaving ample room for regional and local structures, ideas and values. Maybe without actually realizing it, Europe is setting an example for the world on how to live together on this planet in harmony. It is a social project of which the essence is that we move on from the quantitative, material development of the industrial age, to­ wards sustainable development with a strong focus on quality of life; towards better relationships between people, less competitive and less hostile, and through this to a better relationship between people and the earth. I see this happening in Europe and believe that it is the result of no­ tions that have their ancient roots in our European society. We draw from the depths of our own nature, our values and moral principles in forming Europe. In these roots we find our identity and proceed along the lines I sketched. Can you sketch in the lines in more detail? In the industrial age man defined himself as being outside of nature: I think, therefore I am. Mankind was thereby reduced to thinking; people were put above nature, in a position of control. The machine is a good metaphor for this period: a few people at the buttons and in control, the others mere extensions of the machine. Paintings from the early Russian revolution depict this very well: large machines, angular workers laugh­ ing broadly in front of them. Of course, industrialization has brought us a lot of good: welfare, well-being, health and emancipation, but it has also resulted in very stressful relationships between people and between mankind and earth. It has led to colonization, slavery, exploitative capitalism and hierarchical constitutions. Now, we need to move on, towards sustainable development. We come to a phase of quality orientation in social development, and also in the sense of personal responsibility. In my opinion we are witnessing— as is currently clearly visible in the Netherlands—a clash of the industri­ ally shaped state and the social-cultural reality of the emancipated man, who wants to have an influence on the matters that are truly important for his own quality of life, that is security, health care, education and a pension that isn’t dependent on Wall street dynamics. We must develop new, smart, small-scale social structures that are part of bigger ones, but in which people can have an influence or a vote. A good example is edu­ cation. After the Second World War, in most nation states education de­ veloped into state run systems. We will go the opposite way in the com­ ing years. The primary legitimacy of education will not be vested in the nation, but in the parties concerned: parents, teachers, students and com­

HERMAN WIJFFELS ON THE VOCATION OF EUROPE

89

panies in the region. The rapidly emerging single-issue organizations will also play an important role in this civil society. This implies major organizational changes, and a crisis in representational democracy as we know it. Nation states will loose influence to these small-scale struc­ tures, to their people, but also to larger structures, to Europe. A good metaphor for this post-industrial European society is a biotope: different species living together in harmony, influencing each other, and as in any biotope, the richer it is, the better. And this European biotope is already developing? Yes, we are currently witnessing the throes of this process. We take steps forward, but also backwards. But don’t misunderstand me; this Europe is not a plan. It is not a question of twenty years of hard work along well thought out pathways; it is a logical, dialectic reaction on what has preceded in the past fifty years. Shortly after the Second World War, the strong notion of ‘never again’ arose in Europe. We developed a buffer against war by merging a few crucial economic functions of na­ tions: steel and coal production, and later agriculture as well. The en­ twining of interests worked: it made war counterproductive. Today, we live on the spurs of this development, moving on to the next phase. I see that as an organic process, unrolling itself further, steered by our com­ mon values and beliefs in the sense that we act when the time is ripe. Europe is a biotope that just needs to be fed; we need to optimize the conditions. Can you give a specific example of how the biotope works? Due to the growth of the financial markets, European currencies became vulnerable in the eighties and nineties. For stability and prosperity, the biotope Europe reacted in the only way that was logical, by turning to cooperation. It sensed that a central bank was necessary for the stability of the currency, and thus it arose. But let me comment on some controversial subjects to clarify the European biotope. If you ask me whether the democratic legitimacy of Europe stems from the European parliament or from the national parlia­ ments, I say the latter. Five hundred people in Strasbourg who are cho­ sen by one third of the electorate do not provide full democratic legiti­ macy. The European Parliament is useful as a independent, strong con­ trol on the working European Commission, but cannot function as a full representative of the people; it is just too far away from people’s day-today lives. Actual legitimacy for Europe must come from member states and the national parliaments.

90

INTERMEZZO

I also prefer the often-applied, open coordination strategies over European legislation. When necessary and logical, national or regional representatives come together to discuss how transnational affairs should be dealt with. This has happened in the case of the Euro, the European market and in agriculture, and is happening now in education and pen­ sion laws. These discussions can of course result in European legislation, regulations and institutions. But these need to come from below, not from above. Europe must act as a server instead of as a booster. Maybe such discussions will result in a European defense force. However, not as an instrument to compete with the United States, but as a tool to be used when Europe wants to act in a conflict on its territories or near its borders. Europeans just do not want a European government, a European president or a European military. Many regard this as a shortcoming; they see Europe as a copy of the United States. Well, I do not. Europe is unique, forming along lines that result from ancient values, and resulting in a totally new concept on how to live together peacefully on this planet. Instead of following the US, we will set our own example. You foresee a Europe without a president. But what about the recent call for a strong man? We need inspiring leaders, not a strong man, that will soon be an arche­ type. We need leaders like Nelson Mandela and Kofi Annan, leaders that again and again confront people with the truly important aspects of life. Inspiring, calm leaders like Ghandi; soft powers with a tremendous im­ pact. Which values form the basis of the non-pyramidal European? Let me use a Biblical image to answer that question, the image of one’s neighbor. In ancient writings the word neighbor refers not just to the person next to you, but to someone sharing your territory. Literally translated neighbors are meadow-fellows. This concept dates back to the ages in which we lived as nomads, moving from meadow to meadow depending on season and weather. Thus, the Biblical image of one’s neighbors includes the important notion that people share this earth. It comprises the social message that one must treat his territory well, as someone will come behind him. The message is treat other people and places as you would treat your own. My feeling is, and in this perhaps I’m anticipatory and optimistic, that we are picking up that ancient no­ tion again in a modern way. The Bruntland report was in fact the starting point of that process. During industrialization people’s lives have be­

HERMAN WIJFFELS ON THE VOCATION OF EUROPE

91

come segregated to such an extent that it makes them feel uncomfor­ table. In the years to come man and life will be regarded once again as a whole. That old notion will be rediscovered. Your vision includes a very positive image of people. EVS-studies, how­ ever, show that only a mere thirty percent of the Europeans have true faith in their neighbors. That doesn’t surprise me. If you ask me whether I believed that the next person I meet will sincerely wish me all the best, I would also be very skeptical. And be honest, our society is to a great extent based on rivalry. The European competition policy explicitly states that economic actors should be opponents. Still, I believe that humanity in the long, long run wants to put civilization on a higher plain. If not, what is our purpose? This also brings us again to the fundamental value intrinsic in our soci­ ety: that we are members of a long, long line of passers-by on earth; that we are part of a long line of life, development and creation. In our secu­ larized world this notion has largely been lost. Society is highly indi­ vidualized; we live staccato lives. Sustainability can help bring the longterm notion back into our lives, also in our daily lives. I think that real polarity in this century will be much more along the axis of short-term thinking versus sustainability, rather than on the old axes of capital and labor. So sustainability is the new religion? No, sustainability is not a religion. But sustainability can rebuild a value pattern that connects man with nature and long-term thinking; a vertical dimension in our horizontalized world. Sustainability draws on various cultures and religions; it is a kind of common value. However, it is also the new source of polarity in world politics; consider, for example, the differences in opinions between Europe and the US as manifested in Jo­ hannesburg. Europe does some pioneering work in sustainability, especially if it comes to shaping institutional structures. However, Europe’s role is not quite that coherent. At the European summit in Lisbon, it was declared that Europe is striving to beat the US in twenty years. But that is not the point at all. We don’t need to do better than the US, Asia or anyone else. We need to achieve goals that improve the life of Europeans – that is the only legitimate orientation for Europe. It is all about our needs, our cul­ ture and our values.

92

INTERMEZZO

Dr. Herman H.F. Wijffels (1942) is chairman of the Social and Eco­ nomic Council (SER), the main advisory body of the Dutch government on national and international social and economic policy. Wijffels, a farmer’s son, studied economics at Tilburg University. He graduated in 1966 and calls himself a member of the Cobbenhagen school. After gra­ duation, he followed a traineeship at the European Commission in Brus­ sels, where he was soon ‘scouted’ by the Minister of Agriculture. He ac­ cepted a job at the Dutch Ministry of Agriculture and shortly afterwards became a director. During those years he was a civil servant at the Euro­ pean Commission. In 1977 he was appointed Secretary General of the National Federation of Christian Employers (NCW), and in 1982 be­ came a member of the board of executive directors of the Rabobank, a cooperative bank that sprang out of agricultural communities in the Ne­ therlands. He had been its CEO for almost fifteen years when he trans­ ferred to the SER in 1999. Wijffels is also chairman of the Society for Preservation of Nature in the Netherlands and chairman of the Founda­ tion Board of Tilburg University. Wijffels is married and has three grown children.

PART TWO

EUROPEAN VALUES IN DEPTH

This page intentionally left blank

CHAPTER FOUR

OPENING THE BLACK-BOX OF REGIONAL CULTURE. ENTREPRENEURIAL ATTITUDE AND ECONOMIC GROWTH IN 54 EUROPEAN REGIONS SJOERD BEUGELSDIJK & NIELS NOORDERHAVEN 1 Introduction Entrepreneurship is ‘at the heart of national advantage’ (Porter, 1990: 125). Especially in the field of economic geography and regional eco­ nomics there has been a recent upswing in the interest in the influence of regional culture on regional economic development. The literature on regional clusters increasingly stresses the role of entrepreneurship and an entrepreneurial culture in explaining the economic success of regions. In an analysis of U.S. biotechnology clusters Audretsch (2001) argues that the existence of an entrepreneurial culture is an important factor in fostering the start-up and growth processes of biotech firms. But also in related literature stemming from theoretical concepts like ‘industrial dis­ tricts’ (Marshall, 1920; Markusen, 1996; Ottati, 1994; Rabellotti, 1998; Storper, 1992), ‘regional innovation systems’ (Cooke et al., 1997; Mal­ ecki, 1997) and ‘the learning region’ (Florida, 1995; Morgan, 1997) terms like ‘regional innovative capacity’ (Lawson & Lorenz, 1999), ‘en­ terprise culture’ (Amin & Tomaney, 1991), ‘entrepreneurial ability’ (Kangasharju, 2000), ‘entrepreneurial human capital’ (Georgellis & Wall, 2000) and ‘regional cultures of innovation’ (Thomas, 2000) are frequently used. It is argued that local social conditions play an impor­ tant role in the genesis and assimilation of innovation and its transforma­ tion into economic growth. More specific, entrepreneurial attitude is seen as an important element of a regional culture facilitating the success of regional clusters and regional economies in general1. Still, empirical research on entrepreneurship as a driving force of economic develop­ ment is not well developed (Wennekers & Thurik, 1999). The measurement of entrepreneurial attitude is difficult and espe­ cially on the regional level it is hard to obtain data. The scarce empirical studies that explicitly take regional culture into account only measure it ______________ 1

For a critical reflection on this literature on regional clusters see Hospers and Beu­ gelsdijk (2002).

96

CHAPTER FOUR

in an indirect way, either by allowing for region-specific effects (e.g., Georgellis & Wall, 2000) or using a proxy for regional culture (e.g., Kangasharju, 2000)2. This chapter is an attempt to empirically test if certain societal cha­ racteristics are related to regional economic growth. In specific, we test if regions, the culture of which can be characterized as ‘entrepreneurial’, grow faster than regions that score lower on our measure of entrepreneu­ rial attitude. Despite the growing literature in the field of economic geo­ graphy and regional economics in which the role of an entrepreneurial culture is stressed, to our knowledge nobody has explicated the values that make up this entrepreneurial attitude at the regional level. It is in most cases a black box, which is commonly referred to, but never de­ mystified. The contribution of the chapter is twofold. Firstly, we show that en­ trepreneurs differ from the rest of the population in several ways. Our analysis shows that entrepreneurs are more individually oriented. Indi­ vidual responsibility and effort are distinguishing characteristics. Secondly, based on these entrepreneurial characteristics, we construct a regional aggregate of ‘entrepreneurial attitude’. We study 54 regions in Europe and show that regions that score higher on these entrepreneurial characteristics grow faster. By unraveling the soft factors influencing economic growth we open the black box of regional entrepreneurial cul­ ture. Doing so, we shed empirical light on the relationship between en­ trepreneurship and growth. In this chapter we start with a discussion why regional culture mat­ ters. Then, we study the self-employed and compare their personality characteristics with the general working population. Based on a sample ______________ 2

Though we sometimes seem to use the terms culture, value and attitude inter­ changeably, it is important to note that these concepts are closely related but not the same. It is beyond the scope of this chapter to extensively discuss these different concepts and meanings. Numerous definitions of culture exist, and most include elements like mea­ nings, values and religion or ideology. One of the most accepted and extensive definitions is the one proposed by Clifford Geertz. He defines culture as ‘an historically transmitted pattern of meanings embodied in symbols, a system of inherited conceptions expressed in symbolic forms by means of which men communicate, perpetuate, and develop their own knowledge about and attitudes toward life’ (1973: 89). Hofstede’s definition is in a way a summary of Geertz, as Hofstede (2001: 1) defines culture as the ‘collective programming of the mind’. Hofstede adds to that culture does not only manifests itself in values but also in more superficial ways, in symbols, heroes and rituals (ibid.). A central element in most definitions of culture is the concept of values. Similar to culture, numerous definitions ex­ ist but here we follow Hofstede. Building on a large body of literature he defines a value as a ‘broad tendency to prefer certain states of affairs over others’ (Hofstede, 2001: 5). The concept of attitude comes close to this definition of a value and may perhaps be seen as a specific pattern of values. For an excellent overview of definitions of ‘culture’ and ‘values’ we refer to the first chapter ‘Values and Culture’ of Hofstede’s 2001 revised se­ cond edition of Culture’s Consequences.

OPENING THE BLACK-BOX OF REGIONAL CULTURE

97

of 8,332 individuals we find 5 distinguishing characteristics of entrepre­ neurs. Building on these characteristics, the next step consists of con­ structing a score on entrepreneurial attitude for 54 regions in Europe. By using principal components analysis, we construct a measure of entre­ preneurial attitude for each region. Based on standard growth analyses we test if regions that have more entrepreneurial attitude grow faster. We conclude with suggestions for further research and discuss the policy implications of our findings. 2 Why would entrepreneurial culture matter? Wennekers and Thurik (1999) investigate the relationship between en­ trepreneurship and economic growth extensively. Building on various perspectives like macro-economic growth theory, historical views on en­ trepreneurship, industrial economics (mainly Porter’s view), and evolu­ tionary economics they try to synthesize these insights to provide a broad picture of how economic growth is linked to entrepreneurship. In their view, entrepreneurship is a behavioral characteristic of persons. Therefore, ‘linking entrepreneurship to economic growth means linking the individual level to the aggregate level’ (Wennekers & Thurik, 1999: 46). When describing the function of entrepreneurship in relation to eco­ nomic growth, Wennekers and Thurik (1999) single out two major roles. The first has to do with the start-up rate of new firms. The second has to do with, what they call ‘newness’ in general. In the first role, the entrepreneur is seen as the founder of a new business. In the second case we think of enterprising individuals (intrapreneurs or corporate entrepre­ neurs) in large existing firms, who undertake entrepreneurial action. Nations and regions that are characterized by a culture that is condu­ cive to entrepreneurship may have higher start-up rates. This may, in turn influence economic growth in a way that is in the eyes of many re­ searchers what entrepreneurship is all about. In an analysis of the effects of regional characteristics on gross firm formation in Finland, Kan­ gasharju (2000) argues there are a number of local characteristics. Be­ sides local market growth, agglomeration and urbanization effects, and government policies, he argues that entrepreneurial ability is an impor­ tant factor in explaining the profitability of firm formation. According to Kangasharju (2000) this entrepreneurial ability in a region depends on both the stochastic distribution of entrepreneurial talent among the in­ habitants of a region and on region specific factors that enhance this ability. Georgellis and Wall (2000) study levels of entrepreneurship in terms of rates of self-employed across regions in Britain for the period 1983-1995. Besides labor market conditions, labor force characteristics

98

CHAPTER FOUR

and industry composition they find that the ‘entrepreneurial human capi­ tal’ of a region is an important explanatory factor. However, entrepreneurship not only occurs through the formation of new small firms but also in the form of corporate entrepreneurship. Stopford and Baden-Fuller (1994) identify three types of corporate entrepre­ neurship. The first type is what they call corporate venturing. This im­ plies the creation of new business units of businesses within the existing organization. The second type relates to the transformation of strategic renewal of existing organizations. The third type is where the firm changes the ‘rules of competition’ for its industry. We can, for example, think of an innovation that fundamentally changes the industry. Intrapre­ neurship plays an important role in the process of strategic renewal of existing firms. It can be associated with alertness, finding new productmarket combinations and innovation (Wennekers & Thurik, 1999). In the long run, it is expected to positively affect firms’ competitiveness. According to Penrose (1959), entrepreneurs are important for the growth of firms since they provide the vision and imagination necessary to carry out opportunistic expansion. In sum, this intra-preneurial activity may yield efficiency advantages within firms, which on the aggregate level results in higher growth rates. But besides intra-preneurial activity or corporate entrepreneurship, other authors have focused on technological development in relation to the role of social conditions. In a historical overview of growth differen­ tials between countries Abramowitz (1986) has emphasized the role of social capability. Although he does not provide us with a clear defini­ tion, he argues that ‘tenacious societal characteristics normally account for a portion, perhaps a substantial portion, of a country’s past failure to achieve as high a level of productivity as economically more advanced countries. The same deficiencies, perhaps in attenuated form, normally remain to keep a backward country from making the full technological leap, envisaged by the simple hypothesis (of catching up)’ (1986: 387). Abramowitz argues that a country’s potential for rapid economic growth partly depends on societal characteristics, which he refers to as ‘social capability’. A crucial element of Abramovitz’s concept of social capabi­ lity is adaptability. Some countries may be more fitted to adapt to the re­ quirements of changing circumstances. He assumes that there is a link between technological advancement and social capability and that that link is established through the capacity to adapt to change, i.e., adapta­ bility. Together, social capability and technological gap define a country’s potential for productivity advance by way of catching-up. Or, as he puts it very clearly himself (1986: 390): Countries that are technologi­ cally backward have a potential for generating growth more rapid than

OPENING THE BLACK-BOX OF REGIONAL CULTURE

99

that of more advanced countries, provided their social capabilities are sufficiently developed to permit successful exploitation of technologies already employed by the technological leaders. In an analysis of Euro­ pean regions Pose (1999) uses a similar argument to explain the regional variance in innovativeness. He introduces so-called ‘innovation-prone’ and ‘innovation-averse’ societies. Innovation-prone regions are those featured by a weak social filter, which facilitates the transformation of innovation into growth. Though it can be questioned whether the term weak or strong social filter captures the issue correctly, it is clear what Pose (1999) means. The social structure may hamper or promote the re­ gional economic growth process through its impact on technological de­ velopment. Pose focuses on innovative capacity and social filters. Be­ sides other factors like the amount of local resources devoted to R&D, the nature of the type of R&D, the local economic structure and the na­ ture of local production factors, the capacity of a region to assimilate and transform its own or foreign R&D into economic activity depends on so­ cial factors. The social settings in which economic activity takes place play a crucial role in determining the passage from R&D to innovation and growth. Local social conditions act as a social filter. In sum, entrepreneurial culture influences (regional) economic growth in several ways. First, value patterns conducive to entrepreneur­ ship may increase the start-up rate of new firms. Second, intra-preneurial activities may yield efficiency advantages within firms. Finally, social structures may influence the absorptive capacity and promote the degree to which countries or regions are able to adopt and adapt to new tech­ nologies. Social conditions may serve as a social filter, making societies innovation-prone or innovation-averse. Hence, ‘wherever entrepreneurial employees reap the benefits of their abilities, within the firm or in a spin­ off, their activities are likely to enhance growth at a macro-level’ (Wen­ nekers & Thurik, 1999: 45). We proceed as follows. First we study the value patterns of entrepreneurs and compare them with those of nonentrepreneurs. The analysis is at the individual level. After finding the distinguishing personality characteristics of entrepreneurs we aggregate these characteristics at the regional level. Finally, we test whether entre­ preneurial attitude is related to economic growth. The three steps in our analysis are depicted in Figure 4.1. 3 Entrepreneurial characteristics Reviewing the literature on entrepreneurial trait research, Brockhaus (1982) identified three attributes consistently associated with entrepre­ neurial behavior: need for achievement, internal locus of control, and a

100 SStep 1

Step 2

CHAPTER FOUR

What are the distinguishing characteristics of entrepreneurs?

Analysis at the individual level

Can we create regional scores that reflect the entrepreneurial attitude of the average population? Analysis at the regional level

Step 3

Can we explain regional increase differentials by regional differences in entrepreneurial attitude (conditio­ nnal on other factors)?

Figure 4.1 Structure of the chapter

risk-taking propensity. More recent research on entrepreneurial trait re­ search comes to similar personality characteristics (Thomas & Mueller, 2000). The first attribute, ‘need for achievement’, can be traced back to McCelland’s study (1961), whereas the second attribute, ‘locus of con­ trol’, dates back to Rotter (1966). The concept of locus of control refers to the perceived control over events. Internal locus of control implies the individual’s believe that he or she has influence over outcomes through ability, effort or skills. On the other side of the spectrum, external locus of control means the individual believes that forces outside the control of him or herself determine the outcome. It is clear that individuals with an internal locus of control are more likely to be entrepreneurs. The third at­ tribute, risk-taking propensity refers to the acceptance of risk and failure. Moreover, as extensively described by Mueller and Thomas (2000), there appears to be strong evidence that entrepreneurs have a more inno­ vative attitude than non-entrepreneurs. In sum, achievement motivation, locus of control, risk-taking propensity and preference for innovation are

OPENING THE BLACK-BOX OF REGIONAL CULTURE

101

seen as the classic themes in the entrepreneurial trait research (Stewart et. al., 1998) 3. 3.1 Data and method In order to operationalize the theoretical constructs that were discussed in the previous section, we now turn to the data we have used. The dataset we use to find distinguishing characteristics of entrepreneurs is the European Values Study (EVS) survey. By now, the survey comprises three waves (1981/1990/1999), of which we use the second one. In order to obtain regional scores on our indicators of entrepreneurial attitude in the second step of our analysis (see Figure 4.1), we had to regroup the original individual data. We did not use the first wave that was carried out in 1981, because we could not trace the individual scores in terms of regions. The latest wave, 1999/2000 was not completed by the time we finished this chapter. Therefore we use the 1990 data4. In the following section, we discuss the details and operationalization of our dependent and independent variables, as well as the control variables we included in our analysis for the first step of our analysis, i.e., the empirical dis­ tinction between entrepreneurs and non-entrepreneurs. 3.2 Dependent variables Entrepreneurship is an ill-defined concept (OECD, 1998). Measurement of entrepreneurship is therefore difficult. Nevertheless, there are at least two basic ways in which entrepreneurship can be measured. Firstly, it can be operationalized as ‘self-employment’ or ‘business ownership’. By measuring it this way, it serves as a static indicator. However, selfemployment is a broader concept than the strict definition of entrepre­ neurs. Especially in the agricultural sector a large fraction of the total working population is self-employed, but it can be questioned if these are entrepreneurs in the true Schumpeterian sense, i.e., fit the idea of a process of creative destruction. The same holds for small retail shops or the category of firms that are known as ‘mom-and-dad’-shops. It is im­ portant to control for these factors in empirical research (Gartner & ______________ 3

It goes beyond the scope of this chapter to extensively review the existing studies on entrepreneurial trait research. Our only aim is to provide theoretical ground for the choice of our questions by means we measure entrepreneurial attitude. Our goal in this chapter is not to add insights to the literature on entrepreneurial trait research, but to open the blackbox of regional culture. For an extensive overview of the entrepreneurial trait research we refer to Stewart et. al. (1998), Mueller and Thomas (2000), and Beugelsdijk and Noorder­ haven (2002). 4 For the detailed and extensively described background of the EVS project and the corresponding dataset, we refer to www.uvt.nl/evs.

102

CHAPTER FOUR

Shane, 1995). Secondly, to capture the dynamic aspect of entrepreneur­ ship, it is often measured as nascent and start-up activity, also referred to as turbulence rate (total of entry and exit). In the EVS entrepreneurship was measured by first asking whether the respondent was employed, and if the answer was positive, if he or she was self-employed. Thus our dependent variable is self-employment as indicated by the respondent him- or herself. This corresponds to what is common in entrepreneurship studies of a cross-sectional nature. We estimate two different regression equations. In the first analysis we com­ pare self-employed with the rest of the population, including unem­ ployed, retired people, students, and housewives. The number of obser­ vations equals 14,846 of which 888 are self-employed (6 percent). In our second analysis the reference category in the self-employment equation is the wage- and salary earners. Here the number of observations is 8332 of which again 888 are self-employed (10.6 percent). 3.3 Independent variables In order to test for personality characteristics of entrepreneurs, we se­ lected a number of questions from the EVS, based on existing literature on entrepreneurial trait research. These questions pertained to ascribed reasons for personal success or failure, values instilled in children, atti­ tudes towards future developments, preference for equality versus free­ dom, and the attitude towards a number of social issues. In the EVS respondents are asked to rate the importance of a number of explanations of why people are living in need, which is related to the earlier discussed concept of locus of control. Four possible answers are given, of which the respondents are asked to rate the importance: ‘be­ cause they are unlucky’; ‘because of laziness and lack of willpower’; ‘because of injustice in our society’; and ‘because it’s an inevitable part of modern progress’. We recoded the four answer categories as dum­ mies, with 1 if this reason was indicated to be important, and 0 if not. We think the second reason, referring to the individual responsibility, may be assumed to correlate positively with entrepreneurship, and the other reasons, referring to external factors, negatively. Respondents were also asked to indicate which values they consi­ dered important qualities to teach children. Related to the characteristic of innovative, frame-breaking behavior we selected qualities like ‘inde­ pendence’, ‘imagination’ and ‘obedience’. Other qualities selected were ‘thrift’, ‘hard work’, and ‘determination, perseverance’, of which the lat­ ter two can be seen as indicators of achievement motivation. Thrift can be seen as an indicator of internal locus of control, assuming that savings can be used for later investments to better one’s condition. All these

OPENING THE BLACK-BOX OF REGIONAL CULTURE

103

questions were also recoded as dummies. We expect all values, except ‘obedience’, to correlate positively with entrepreneurship. Another question in the EVS asked respondents whether they evalu­ ated positively or negatively various future changes in the way of life. We selected two possible changes as potentially positively related to en­ trepreneurship. As an indicator for innovativeness we selected ‘more emphasis on the development of technology’. Locus of control was proxied by the evaluation ‘greater emphasis on the development of the individual’. We also selected a question in which the importance of freedom and equality was rated. A preference for freedom can be seen as an indica­ tion of an innovative attitude. Choosing freedom above equality suggests an interest in frame-breaking behavior. We constructed a dummy vari­ able, coded as 1 if freedom was considered more important than equal­ ity, and as 0 otherwise. Then we chose a number of questions pertaining to the attitude of the respondent towards a number of social issues. In these questions respon­ dents were asked to place their views on ten-point Likert-type scales with as anchors, respectively: — Incomes should be made more equal. — Private ownership of business and industry should be increased. — Individuals should take more responsibility for pro­ viding for themselves. — People who are unemployed should have to take any job available or lose their unemployment benefits. — Competition is good. It stimulates people to work hard and develop new ideas. — In the long run, hard work usually brings a better life.

Versus There should be greater incentives for individual effort. Versus Government ownership of business and industry should be increased. Versus The state should take more responsibility to ensure that everyone is provided for. Versus People who are unemployed should have the right to refuse a job they do not want. Versus Competition is harmful. It brings out the worst in peo­ ple. Versus Hard work doesn’t generally bring success – it’s more a matter of luck and connec­ tions.

104

CHAPTER FOUR

In the scales, low values are associated with the statement on the left hand, and high values with that on the right hand. All statements refer to risk-taking, except for the first statement that refers to achievement mo­ tivation and the last that reflects locus of control. We expect a negative correlation with entrepreneurship of all these variables, except for the first, where we expect a positive relationship. 3.4 Control variables We included the GDP per capita (in 1990) to control for level of welfare (taken from Penn World Tables). Countries with a higher level of GDP and a corresponding lower share of the agricultural sector (Chenery, 1960) have lower levels of self-employed, as the number of selfemployed in the agricultural sector is relatively high and the number of small-scale retail and craft establishments (‘mom-and-dad’ shops) de­ creases with the rise of the GDP. Furthermore we included a number of controls in the selfemployment equation. Both self-employment and personality character­ istics are most probably related to factors such as age, wealth, sex, labor market experience and human capital. The dataset allows us to control for sex, age, income and socio-economic status. With respect to sex, we take females as the reference group. Female self-employment rates are generally lower than those of men (OECD, 1998). These lower self-employment rates of women are caused by dif­ ferent factors (see Verheul et al., 2001). An important factor limiting female entrepreneurship is the combination of household and family re­ sponsibilities. Though there are arguments favoring female selfemployment, for example, flexible time schemes (Cowling & Taylor, 2001), we expect more male respondents to be self-employed. Income is only measured in an indirect way. For reasons of privacy, income is not measured in a direct way by asking the gross or net monthly income in the EVS. Instead, income is measured on a 10-point scale, which leaves room for perception and thus results in a rather sub­ jective measure of income. Nevertheless, we decided to include it as a control variable. Age is measured in number of years. For age we expect a curvilinear relationship, as young and old people are less likely to be self-employed. Other studies have also suggested this curvilinear effect (Evans & Leighton, 1989; Storey, 1994; Cowling & Taylor, 2001). Entrepreneurs tend to start a business when they are between 30 and 40 years old (Co­ lombo & Delmastro, 2001). On the one hand, risk aversion and the costs of leaving an employment position are positively related to age, which decreases the age to be self-employed. On the other hand, young people

OPENING THE BLACK-BOX OF REGIONAL CULTURE

105

may lack professional experience and relations and experience liquidity constraints, which have an upward effect on the age to start a business. As our data does not allow us to test when people have started their own business, we are not able to estimate the average age of a starting entre­ preneur. Nevertheless, we still expect this curvilinear effect, as older people might have sold their business. We also control for level of education or human capital. Lack of data does not allow us to use a direct measure of educational background. However, EVS contains information on socio-economic status. Inter­ viewees are categorized in four groups. If the individual interviewed be­ longs to the upper or upper-middle class it is coded 1. People belonging to the middle class (non-manual workers) form the second group and the third class consists of manual workers (skilled or semi-skilled). The last group, coded 4, consists of unskilled manual workers. Finally, we included country dummies to control for country-specific effects other than GDP. All kind of country specific effects may lead to national differences in the probability to become self-employed. For in­ stance, Colombo and Delmastro (2001) find that the educational system in Italy lowers the percentage of self-employed. The institutional setting may also influence the decision to become self-employed. The national bankruptcy and antitrust law are important factors in this respect (Golodner, 2001). 3.5 Method To empirically test for personal characteristics associated with entrepre­ neurship we used a logit equation. We estimate two models. In the first model we estimate the probability of self-employed versus the general population. The second model uses wage and salary-earners as a refer­ ence group. When a variable is statistically significant, it implies that en­ trepreneurs are different from non-entrepreneurs. In case a value is sig­ nificantly positive (negative), it means that entrepreneurs score higher (lower) on this variable. 3.6 Findings Results are well interpretable. The self-employed distinguish themselves both from the general population as well as from wage- and salary ear­ ners in their stronger preference for greater incentives for individual ef­ fort and their opinion that the state should not take more responsibility. Moreover, they feel that private ownership should be increased, that un­ employed should not have the right to refuse a job and that success is not

106

CHAPTER FOUR

Table 4.1

Probability of being self-employed Model 1 Self-employed versus general population

People live in need …

Because they are unlucky Because of laziness and lack of willpower

Because of injustice in our society

Because it’s an inevitable part of modern progress

Important qualities to teach

children ...

Independence Hard work Imagination Thrift Determination, perseverance Obedience Evaluation of future developments ... More emphasis on the development of technology

Greater emphasis on the development of the individual

Freedom is more important than equality

Attitude towards social issues ... There should be greater incentives for individual effort Government ownership of business should be increased The state should take more responsibility Unemployed should have the right to refuse a job

Competition is harmful. It brings out the worst in people

Success is a matter of luck and connections

Model 2 Self-employed versus wage- and salary earners

–.23 .13

(–1.22) (.73)

–0.22 0.12

(1.10)

(0.69)

–.12

(–.70)

–0.11

(0.59)

–.15

(–.82)

–0.16

(0.84)

–.03 .17 .85 –.006 –.05 –.12

(–.37) (2.03) ** (1.01) (–.07) (–.63) (–1.46)

–.05 .10 .11 .04 –.025 –.11

.05

(.63)

.05

(.60)

–.073

(–.69)

–.16

(1.53)

.12

(1.58)

.089

(1.16)

.02

(3.34) ***

.02*

(3.05) **

(.63)

(1.18) (1.24) (.45) (.30) (1.24)

–.02

(1.88) *

–.027

(2.22) **

–.03

(2.36) **

–.029

(2.03) **

–.03

(2.50) **

–.027

(1.91) *

.005

(.75)

.007 –.04

(1.07) (2.85) ***

–.04*

(.02) ***

OPENING THE BLACK-BOX OF REGIONAL CULTURE

107

Table 4.1 (continued) Control variables GDP per capita 1990 Age Age squared Sex Income Socio–economic status N Chi-square Log Likelihood

–.1 .18 –.002 .71 .03 –.096 14846 658 –3032.95

(8.79) *** (11.28) *** (11.25) *** (9.30) *** (1.74) * (2.73) ***

–.1 .027 .00005 .25 –.03 –.028 8332 512 –2570.89

(8.84) *** (1.55) (.25) (3.15) *** (1.79) * (.81)

The dependent variable is 1 if self-employed. The reference group in model 1 is the gene­ ral population, whereas the reference group in model 2 are the wage- and salary earners. Key-words in variable names in italics. T-statistics are in parentheses. *** p < .01, ** p < .05, * p < .1. GDP per capita in 1000 USD. Country dummies not re­ ported. Estimation is logit in STATA. For the exact formulation of the questions see www.uvt.nl/evs.

a matter of luck and having connections but of hard work. All these fin­ dings fit in a picture of self-employed attaching more value to individual freedom and responsibility, and by nurturing values consistent with the frame-breaking creative destruction associated with Schumpeterian en­ trepreneurs. We also find that self-employed differ from the general population with respect to values that the self-employed think are impor­ tant in raising children. Self-employed attach significantly more impor­ tance to hard work than the rest of the population. The non-significant finding in model 2 suggests that this characteristic is not a distinguishing factor between wage and salary earners and self-employed. In other words, our results suggest that hard work as a quality to teach children does not have to do with being self-employed, but with having a job, ei­ ther as wage or salary earner, or as an entrepreneur. As expected, the coefficient for GDP per capita is significantly nega­ tive. The predicted curvy-linear relationship between age and selfemployment holds for the comparison of self-employed and the general population (model 1), but does not yield significant differences between self-employed and wage and salary earners. The reasons for the invertedU shape in model 1 is that individuals tend to become self-employed at the end of their twenties or in their thirties and probably sell their firm or retire when they reach their sixties. If we compare self-employed with wage and salary earners we use a reference group that also retires at a certain age. As predicted, both models show that men are more frequently selfemployed than women. The income effect is significant in both models

108

CHAPTER FOUR

(though only at 10%), with one crucial difference. In model 1 it is posi­ tively related to self-employment, whereas in model 2 it is negatively re­ lated. If we compare self-employed with the general population includ­ ing retired people, students, and housewives, as we do in model 1, it can be expected that there is positive relationship between income and selfemployed. The negative effect in model 2 is more surprising in this re­ spect. It suggests that given our subjective measure of income, selfemployed perceive their income as being lower than wage-and salary earners. This does not imply their actual income is lower, but means that given the risk entrepreneurs take—compared to wage and salary-earners — they feel their income is relatively low. Socio-economic status is a significantly distinguishing factor between self-employed and the general population. Self-employed have a higher socio-economic status. Recall that socio-economic status indirectly re­ flects the educational profile of an individual (skilled-unskilled). The positive relationship between socio-economic status and being selfemployed is logical if we compare this group with the general popula­ tion. If we compare self-employed with wage and salary earners socio­ economic status is not significant. The reason for this is that the varia­ tion in socio-economic status among the general population (including, for example, unemployed) is higher than among wage- and salary earn­ ers5. The next step in our analysis consists of constructing a regional ag­ gregate that captures the characteristics we distinguished. In order to construct one measure for entrepreneurial attitude that is internally con­ sistent and stable, we applied principal components analysis on the items included in this measure. The items we used in our analysis are the five items that were significant in both model 1 and model 2, i.e., the general population as well as the wage and salary-earners. We estimated the principal components by making use of the interval-scaled items ‘indi­ vidual effort’, ‘government ownership’, ‘state responsibility’, ‘unem­ ployed’ and ‘success’ (0-10). For the matrix see Table 4.2. The output shows that the 5 items can be grouped in 1 component (groups of items). For the analyses in the following sections we have calculated regional scores on entrepreneurial attitude on the basis of this five-item-based factor score. Following Porter (2001) our regional aggregate reflects the entrepreneurial ‘spirit’ at the regional level. We think of entrepreneurial attitude as ‘taking initiative, being innovative, shaping the environment according to one’s ideas and goals, etc’ (Brandstätter, 1997: 160). ______________ 5

In the first model the standard deviation of socio-economic status equals 1.32, whereas this standard deviation is 1.20 in the second model, which confirms our reaso­ ning.

OPENING THE BLACK-BOX OF REGIONAL CULTURE

Table 4.2

109

Component matrix

1 component extracted Individual effort Government ownership State responsibility Unemployed Success

.384 –.679 –.720 –.577 –.628

4 Empirical test In order to test if entrepreneurial attitude is related to economic growth, we have taken a standard growth framework, in which economic growth is explained by a number of key economic variables (Baumol, 1986; Barro, 1991; Mankiw et. al., 1992). These type of empirical growth re­ gressions typically include initial level of welfare, and proxies for hu­ man and physical capital (mostly the school enrolment ratio and the in­ vestment ratio). These empirical growth models are also referred to as Barro-regression (after Barro, 1991). We analyze the period 1950-1998. The number of regions equals 54. The set contains 7 European countries: France, Belgium, Italy, Germany, Spain, The Netherlands and the United Kingdom. The regional level is the NUTS1 level, which means that France is divided in 8 regions, Bel­ gium 3, Italy 11, Germany 11 (former German Democratic Republic ex­ cluded), Spain 7, The Netherlands 4 and the UK 10. Similar to Barro and Sala-I-Martin (1995), we have computed the re­ gional growth figures by relating the regional GDP per capita informa­ tion to the country mean6. There are two reasons to use the country mean as a correction factor. First of all we do not have regional price data. Secondly, the figures on regional GDP are provided in an index form that is not comparable across countries. Hence, we have used Gross Re­ gional Product (GRP) figures that are expressed as deviations from the means from the respective countries. The 1950 data is based on Molle, Van Holst and Smit (1980), except for the data for Spain which refers to 1955 and are based on Barro and Sala-I-Martin’s (1995) calculations. The 1998 data on GRP is based on Eurostat information. Hence, our basic regression analysis includes initial level of welfare, school enrolment rate, investment ratio, spatial auto-correlation and a ______________ 6

Gross Regional Product of a region in 1950 is divided by the mean of the Gross Re­ gional Products of all regions belonging to a certain country. A similar formula is applied to calculate the 1998 relative regional product. Regional growth over the period 1950­ 1998 is subsequently based on these two indices.

110

CHAPTER FOUR

variable that captures the concentration of human capital in major ag­ glomerations (for the details of the growth regression we refer to the contribution of Beugelsdijk and Van Schaik, chapter five in this vol­ ume). Table 4.3a provides an overview of the descriptive statistics. Ta­ ble 4.3b plots the correlation coefficients between the variables used. Table 4.3a Descriptive statistics Mean 24.30 .51 .15 .92 .06

Investment Schooling Entrepreneurial attitude Spatial spillover Agglomeration

Std. Dev. 3.740 .067 .27 .30 .16

N=54; investment data is national.

Table 4.3b Correlation table

Growth 19501998 Growth 1950-1998 Schooling Investment Spatial spillover Agglomeration Entrepreneurial attitude Initial level of welfare

1

Schooling

Investment

–.149

.13

1

Spatial Agglo- Entrespillover meration preneurial attitude .051 –.072 .43*

Initial level of welfare –.55*

–.31*

–.049

–.098

–.28*

.29*

1

–.189

–.028

.39*

–.006

1

–.189

.19

.169

1

.02

.35*

1

–.02 1

* p < .1.

The first model we estimated is the standard model, only including basic economic variables. As the results show, all variables except for the school enrolment rate are significant at the 5% level. Schooling is sig­ nificant at the 10% level. Initial level of welfare is strongly negatively

OPENING THE BLACK-BOX OF REGIONAL CULTURE

111

related to economic growth, which corresponds with the convergence hypothesis. Table 4.4

Regression results. Entrepreneurial attitude and regional economic performance, 1950-1998

Dependent variable Method Constant Initial level of welfare Investment Schooling Agglomeration Spatial spillover Entrepreneurial attitude R-square VIF factor (maximum) CW test

Model 1 Model 2 Regional Economic Growth OLS –1.44 (.62) –.11 (.62) –.97 (.20)*** –.93 (.17)*** .48 (.20)** .14 (.18) .53 (.32)* .65 (.30)** ** .53 (.20) .44 (.18)** *** .31 (.09) .18 (.10)* .49 (.13)*** .41 1.49 .69

.53 1.49 .95

Standard errors (White corrected) between parentheses. N = 54. *** p < .01, ** p < .05, * p < .1%. We have tested for heteroskedasticity (residual plots and Cook-Weisburg (CW test) and multi-collinearity (Variance Inflation Factors) and found no indications of a possible bias. If we observe the period 1970-1998 or 1984-1998, the conclusion on entrepreneurial attitude does not change.

In the second step we included our construct of entrepreneurial attitude. The result is shown in Table 4.4. Initial level of welfare remains strongly negatively related to economic growth. Schooling becomes significant at 5% level. The investment ratio is insignificant and the spillover variable is only significant at 10% level. Our variable that measures entrepreneu­ rial attitude is positively significant at 1%. This means that a value sys­ tem reflecting an entrepreneurial attitude is positively related to eco­ nomic success, measured as regional economic growth. The question is if our finding on entrepreneurial attitude is robust. We applied several robustness tests. Firstly, we tested for heteroske­ dasticity and multi-collinearity. As shown in Table 4.4, the tests for hete­ roskedasticity show that this is not a problem. The Variance Inflation Factor (VIF) should not exceed values of 10 (Neter et al., 1996), and given the maximum value of 1.49 this indicates multi-collinearity is not a problem. In the next step we have tested for country-specific effects. We have tested for country-specific effects in two ways. First we in­ cluded country dummies. Second we have used cluster-based corrected standard errors where the clusters are defined on the basis of countries.

112 Step 1

Step 2

CHAPTER FOUR

Estimation of personality characteristics of entrepre­ neurs vs. general population and wage- and salary earners

Logit regression: 888 self-employed

Do these distinguishing char­ acteristics fit into one scale? …Yes.

Factor analysis

New variable: Entrepreneurial attitude

Step 3

European Regional Analysis:

Q: Does an entrepreneurial culture have an economic pay off?

Calculate scores on entrepreneurial attitude for 54 European regions Regional Economic Growth 1950-1998

Other ‘standard’ economic variables: - Investment - Initial level of welfare - Human capital - Spatial spillovers

Robustness analysis: - time period 1984-1998 - country-specific effects

Figure 4.2 Structure of the empirical part of the chapter

Answer: YES

OPENING THE BLACK-BOX OF REGIONAL CULTURE

113

When controlling for country specific effects, investment ratio is no longer significant. This is according to expectation, as the investment ra­ tio is measured at the national level. In case country specific effects are included, the country effects pick up the variance in the investment ratio. More important is that entrepreneurial attitude remains significant at the 5% level. We conclude that our robustness tests all indicate the persistent sig­ nificance of entrepreneurial attitude on economic growth in the Euro­ pean regions. Figure 4.2 summarizes the main steps we have taken in our empirical analysis. 5 Implications and limitations We have shown that local social conditions contribute to regional eco­ nomic growth. The main theoretical implication of our analysis thus is that regional cultural differences can be linked in a meaningful way to regional economic outcomes. Even controlling for national characteris­ tics, regional variations are important enough to have a significant im­ pact on economic growth. As the delimitation of regions is based on an administrative criterion (NUTS), rather than on substantive social or economic criteria, the regional effects we found are likely to be underes­ timations of the real effects. Our findings raise the question what factors within regions lead to the formation and persistence of cultural charac­ teristics inducive to economic growth. An important question that remains is, through what mechanisms entrepreneurial attitude influences regional economic growth. On the one hand, it can be argued that regions with a higher score on entrepreneurial attitude may be expected to have higher start-up rates, which results in a relatively high share of self-employed. This may influence economic growth in a traditional Schumpeterian way. On the other hand, higher scores on entrepreneurial attitude do not necessarily imply a higher start­ up rate, but may also be expressed in intra-preneurial activity. This intra-preneurial activity may increase the innovative capabilities of firms, which on the aggregate level results in higher growth rates. Future re­ search might focus on the intermediating mechanisms between entrepre­ neurial attitude and regional economic growth. A logical next step would be to test if high scores on entrepreneurial attitude go together with a high level of entrepreneurship (number of self-employed). It is interes­ ting to test if, for example, rate and level of technological development of firms in regions is related to entrepreneurial attitude. Another ques­ tion is if the success or failure of regional development programs is re­ lated to entrepreneurial attitude in a certain region. It might be that re­

114

CHAPTER FOUR

gions in the process of structural change are better able to cope with the necessary re­ structuring of the regional economy, if they have a higher ‘amount’ of entrepreneurial attitude. However, lack of regional data on European re­ gions will probably be a problem. Our findings also have consequences for the current trend among policymakers to create regional innovation systems. The capacity of each region to build a successful regional innovation infrastructure is re­ lated to social conditions. It has been argued that especially the cultural uniqueness of successful examples like Silicon Valley and Third Italy makes copying of these successful regions difficult if not impossible (Hospers & Beugelsdijk, 2002). Our results suggest that lack of entre­ preneurial attitude may be an important reason for the failure to create regional innovation systems in certain regions. Policy makers should be aware that entrepreneurial attitude differs from place to place and initia­ tives in the field of regional cluster policy may end up unsuccessfully for lack of entrepreneurial attitude. Hence, in promoting high-tech regions, governments may not only develop R&D programs but also initiatives that aim at increasing the entrepreneurial attitude. In line with the find­ ings of Kangasharju (2000), the results of our study call for the encour­ agement of culture and tradition favorable to self-employment. This is a long term project, as it takes time for such a regional culture to be devel­ oped and take root. One of the limitations of our study is the fact that we used data on en­ trepreneurial attitude based on a 1990 survey and estimated regional economic growth for the period 1950-1998. Lack of data concerning re­ gional origin of respondents prevents us from using the 1981 wave of the EVS surveys. However, as cultural characteristics are persistent in time (Hofstede, 2001), the possible lack of internal validity is probably lim­ ited. Moreover, we minimized the possible effect of endogeneity by tes­ ting the effect of entrepreneurial attitude on the regional-economic growth between respectively 1970-1998 and 1984-1998. As described under table 4 the conclusion on entrepreneurial attitude does not change. A logical next step is to identify intermediating mechanisms through which entrepreneurial attitude influences regional economic growth. The existing case studies on regional systems of innovations and clusters provide sufficient conceptual ideas. 6 Conclusion In this chapter we have established an empirical link between entrepre­ neurial attitude and economic growth. Entrepreneurial attitude as a be­

OPENING THE BLACK-BOX OF REGIONAL CULTURE

115

havioral characteristic has been determined by means of an empirical test in which we compare self-employed with respectively the general population and wage-and salary earners. Based on these distinguishing characteristics we calculated a regional aggregate that reflects the aver­ age score of this entrepreneurial attitude of a population in a region. We have estimated postwar economic growth for 54 European regions and we have shown that entrepreneurial attitude is an important explanatory factor for the explanation of growth differentials. We have opened the black box of entrepreneurial culture, which in this literature often is des­ ignated to be important, but rarely empirically analyzed. Using a unique dataset on norms and values in 54 European regions, we have shown that regions do indeed differ in entrepreneurial attitude, and that a relatively high score on entrepreneurial characteristics is correlated with a rela­ tively high rate of regional economic growth. References Abramovitz, M. 1986. Catching up, forging ahead, and falling behind. Journal of Economic History 46: 385–406. Amin, A. & J. Tomaney 1991. Creating an enterprise culture in the North-East? The impact of urban and regional policies of the 1980s. Regional Studies 25: 479–487. Audretsch, D. 2001. The role of small firms in U.S. Biotechnology clusters. Small Business economics 17: 3–15. Barro, R.J. 1991. Economic growth in a cross section of countries. Quarterly Journal of Econo­ mics 106: 407–443. —— & X. Sala-I-Martin 1995. Economic Growth. New York: McGrawhill. Baumol, W. 1986. Productivity growth, convergence, and welfare: what the long run data show. American Economic Review 76: 1072–1085. Begg, I. 1995. Factor mobility and regional disparities in the European Union. Oxford Review of Economic Policy 11: 96–112. Beugelsdijk, S. & N. Noorderhaven 2002. Personality Characteristics of Entrepreneurs; An Empirical Study. Working Paper. Tilburg: Tilburg University. Beugelsdijk, S. & A.B.T.M. van Schaik 2003. Participation in civil society and European re­ gional economic growth. Chapter five in this volume. Brandstätter, H. 1997. Becoming an entrepreneur – a question of personality structure? Journal of Economic Psychology 18: 157–177. Brockhaus, R.H. 1982. The psychology of an entrepreneur. Pp. 39–56 in C. Kent & D.L. Sex­ ton & K.H. Vesper (eds.), Encyclopedia of entrepreneurship. Englewood Cliffs: Prentice Hall. Chenery, H. B. 1960. Patterns of industrial growth. The American Economic Review 50: 624– 654. Cooke, P., G. Uranga & G. Etxebarria 1997. Regional innovation systems: institutional and or­ ganisational dimensions. Research Policy 26: 475–491. Colombo, M. & M. Delmastro 2001. Technology bases entrepreneurs: does internet make a dif­ ference? Small Business Economics 16: 177–190. Cowling, M. & M. Taylor 2001. Entrepreneurial women and men: two different species? Small Business Economics 16: 167–175. Evans, D.S. & L.S. Leighton 1989. The determinants of changes in US self-employment, 1968­ 1987. Small Business Economics 1: 111–119. Florida, R. 1997. Toward the learning region. Futures 5: 527–536.

116

CHAPTER FOUR

Gartner, W. B. & S. A. Shane 1995. Measuring entrepreneurship over time. Journal of Business Venturing 10: 283–301. Geertz, C. 1973. The Interpretation of Cultures. New York: Basic Books. Georgellis, Y. & H.J. Wall 2000. What makes a region entrepreneurial? Evidence from Britain. The Annals of Regional Science 34: 385–403. Golodner, A. 2001. Antitrust, innovation, entrepreneurship, and small business. Small Business Economics 16: 31–35. Hofstede, G. 2001. Culture’s Consequences; Comparing Values, Behaviors, Institutions and Organizations across Nations. Beverly Hills: Sage. Hospers, G.J. & S. Beugelsdijk 2002. Regional cluster policies; learning by comparing? Kyklos 3: 381–401. Kangasharju, A. 2000. Regional variation in firm formation: panel and cross-section data evi­ dence from Finland. Papers in Regional Science 79: 355–373. Lawson, C. & E. Lorenz 1999. Collective learning, tacit knowledge and regional innovative ca­ pacity. Regional Studies 33: 305–317. Malecki, E. 1997. Technology and economic development: the dynamics of local, regional and national competitiveness. Harlow: Longman. Mankiw, N.G., D. Romer & D. Weil 1992. A contribution to the empirics of economic growth. Quarterly Journal of Economics 107: 407–431. Marshall, A. (1890) 1920. Principles of Economics. London: Macmillan Press. Markusen, A. 1996. Sticky places in slippery space: a typology of industrial districts. Economic Geography 72: 293–314. McCelland, D. 1961. The Achieving Society. Princeton: Van Nostrand Reinhold. Molle, W., B. van Holst & H. Smit 1980. Regional Disparity and Economic Development in the European Community. Westmead: Saxon House. Morgan, K. 1997. The learning region: institutions, innovation and regional renewal. Regional Studies 31: 491–503. Mueller, S.L. & A.S. Thomas 2000. Culture and entrepreneurial potential: a nine country study of locus of control and innovativeness. Journal of Business Venturing 16: 51–75. Nijkamp, P. & J. Poot 1998. Spatial perspectives on new theories of economic growth. The An­ nals of Regional Science 32: 7–38. OECD, 1998. Fostering Entrepreneurship, the OECD Jobs Strategy. OECD, Paris. Ottati, G.D. 1994. Trust, interlinking transactions and credit in the industrial district. Cam­ bridge Journal of Economics 18: 529–546. Paci, R. & S. Usai 2000. Technological enclaves and industrial districts: An analysis of the re­ gional distribution of innovative activity in Europe. Regional Studies 34: 97–114. Penrose, E. 1959. The Theory of the Growth of the Firm. Oxford: Blackwell. Porter, M.E. 1998. On Competition. Cambridge: Harvard University Press. —— 2001. The Netherlands innovation lecture. The Hague. Pose, A.R. 1999. Innovation prone and innovation averse societies: economic performance in Europe. Growth and Change 30: 75–105. Quah, D. 1996. Regional convergence clusters across Europe. European Economic Review 40: 951–958. Rabellotti, R. 1998. Collective effects in Italian and Mexican footwear industrial clusters. Small Business Economics 10: 243–262. Rotter, J.B. 1966. Generalised expectancies for internal versus external control of reinforce­ ment. Psychological monographs: general and applied 80: No 609. Schumpeter, J. 1934. The Theory of Economic Development. Cambridge: Harvard University Press. Stewart, J.R., H. Wayne, W.E. Watson, J.C. Carland & J.W. Carland 1998. A proclivity for en­ trepreneurship: a comparison of entrepreneurs, small business owners, and corporate managers. Journal of Business Venturing 14: 189–214. Stopford, J.M. & C.W.F. Baden-Fuller 1994. Creating corporate entrepreneurship. Strategic management journal 15: 521–536. Storey, D.J. 1994. Understanding the Small Business Sector. Routledge: London. Storper, M. 1992. The limits to globalisation: technology districts and international trade. Eco­ nomic Geography 68: 60–92.

OPENING THE BLACK-BOX OF REGIONAL CULTURE

117

Thomas, K. 2000. Creating regional cultures of innovation? The regional innovation strategies in England and Scotland. Regional Studies 34: 190–198. Thomas, A.S. & S.L. Mueller 2000. A case for comparative entrepreneurship: assessing the relevance of culture. Journal of International Business Studies 31: 287–301. Verheul, I., S. Wennekers, D. Audretsch & R. Thurik 2001. An Eclectic Theory of Entrepre­ neurship: Policies, Institutions and Culture. EIM research report. Wennekers, S. & R. Thurik 1999. Linking entrepreneurship and economic growth. Small Busi­ ness Economics 13: 27–55.

This page intentionally left blank

CHAPTER FIVE

PARTICIPATION IN CIVIL SOCIETY AND EUROPEAN REGIONAL ECONOMIC GROWTH SJOERD BEUGELSDIJK & TON VAN SCHAIK 1 Introduction Recently, economists show an increased interest in the role of culture in relation to economic development. Although Hofstede’s publications (1991, 2001) have contributed to the incorporation of the role of culture and cultural differences in the field of business economics, mainly inter­ national management, this is not the case for general economics. How­ ever, since the mid 1980s there is a revival in the study on the determi­ nants of economic growth. This new or modern growth theory has re­ sulted in a number of empirical studies, in which traditional inputs capi­ tal and labor are complemented with human capital and indicators that proxy institutional and geographical differences between countries. Since the pioneering work of Kormendi and Meguire (1985), Baumol (1986), Grier and Tullock (1989), Barro (1991), and Mankiw, Romer and Weil (1992) growth empirics have become rather popular. However, as Temple (1999) argues, despite this stream of research there is only limited progression in this field. He concludes his impressive survey of empirical growth literature by arguing that there is a role for research on the relation between culture and economics. He (1999: 146) writes: ‘Some of the most interesting thinking on economic growth is to be found on the borders of political science and sociology’. Temple and Johnson (1998: 987) reach a similar conclusion when stating that ‘there are many possible reasons why society might matter, and their investiga­ tion should be a worthwhile direction for further research’. An influential contribution in this respect is the publication of ‘Ma­ king democracy work’ by Putnam, Leonardi and Nanetti in 1993. These authors study Italian regions and find that social capital matters in ex­ plaining the regional differences in economic and institutional (govern­ ment) performance. Putnam et al. (1993: 167) define social capital as those ‘features of social organization, such as trust, norms, and net­ works, that can improve the efficiency of society by facilitating coordi­

120

CHAPTER FIVE

nated actions’. The Worldbank uses a similar definition. According to the Worldbank, social capital refers to the norms and networks that en­ able collective action. It refers to the institutions, relationships and norms that shape the quality and quantity of a society’s social interac­ tions (see: www.worldbank.org/poverty/scapital). In addition to standard economic variables, social capital is conside­ red an important factor in explaining economic success, a statement that we choose to refer to as the Putnam hypothesis. Besides Putnam et al. (1993), Fukuyama (1995) has emphasized the importance of social capi­ tal. He argues that social capital in the form of non-family or generalized trust is of crucial importance for successful performance in advanced economies. As becomes clear in Putnam et al.’s definition of social capi­ tal, trust and networks are seen as dimensions of social capital. Where Putnam et al. (1993) stress the role of networks, Fukuyama (1995) stresses the role of trust. A number of studies has appeared on the concept of social capital since then (Fukuyama, 1995a; Granato et al., 1996; Helliwell, 1996; Swank, 1996; Inglehart, 1997; Fedderke et al., 1999; Paxton, 1999; Van Deth et al., 1999; Inkeles, 2000; Putnam, 2000). However, empirical stu­ dies that focus on the question whether the Putnam hypothesis can be generalized are scarce. Though the concept of social capital is intuitively highly appealing, it is hard to measure it empirically. As Woolcock (1998) puts it, vagueness has plagued social capital scholarship. There are a number of concepts that are used in similar ways as social capital, like social infrastructure (Hall & Jones, 1999) and social capability (Abramowitz, 1986; Temple & Johnson, 1998). The indicators used in the literature on social capital are often trust and social participation. A key empirical paper relating social capital with economic growth is Knack and Keefer’s study (1997). Nevertheless, as Beugelsdijk et al. (2002) have shown, the statistical robustness of their study is limited. The question whether social capital in terms of generalized trust and as­ sociational activity influences economic growth is still not answered. The core question remains whether Putnam et al.’s (1993) study on Ita­ lian regions can be generalized. Besides great academic and journalistic attention, policy makers also show increased interest in the concept of social capital. According to the European Committee and the European Investment Bank (EIB) the en­ dowment of social capital in the form of business culture and shared norms of behavior is of particular importance for regional development (EIB, 2000; EC, 1999). ‘The need, in sum, is for a long term strategy which addresses simultaneously the many aspects of the problem of a lack of competitiveness and attempt to build up the social capital of a re­

PARTICIPATION IN CIVIL SOCIETY

121

gion in parallel with its physical infrastructure, the skills of its work force and its productive base’ (EIB, 2000: 20). Research on the relation­ ship between social capital and regional economic development in the EU may have consequences for the allocation of the structural funds. At the moment, there is too little known about social capital, its functions and the impact on economic growth to formulate clear policy implica­ tions. From a policy point of view it is therefore important to find em­ pirical evidence for the role of social capital in regional economic devel­ opment. This chapter presents an analysis of the relation between social capi­ tal and economic growth for European regions1. By doing so, we are able to test Putnam et al.’s hypothesis on an analogous sub-national level used in their study2. Our study has two major findings. First, we do not find that on a re­ gional level trust and growth are associated with each other. Second, as­ sociational activity and in specific active—unpaid—voluntary work is positively related to regional economic growth. The outline of the chapter is as follows. First we summarize theory on social capital and how it is perceived in the literature. Thereby, we focus on trust and group membership. We describe several functions of trust and argue that trust fulfils different functions at different stages of eco­ nomic development. Besides as a substitute for a well-functioning insti­ tutional system, trust can be seen as a necessary element in complex transactions with incomplete contracts. The second element of social capital we discuss is group membership. Then we turn to statistical ana­ lyses, and test if trust and group-membership are related to regional eco­ nomic growth. After an extensive robustness analysis, we conclude with suggestions for further research. 2 Trust In general, trust can be seen as the perception and interpretation of the other’s expected dependability. Trust is based upon the expectation that one will find what is expected. It refers to the confidence that a partner will not exploit the vulnerabilities of the other (Gambetta, 1988). Several ______________ 1

We wish to thank the European Values Study (EVS) working group for useful sug­ gestions and comments, especially Niels Noorderhaven, Sjak Smulders, and Lans Boven­ berg. 2 There is small difference however. Putnam et al. analyze regions on a different level than we do. Where we use the NUTS1 level (resulting in 11 regions), Putnam et al. apply another definition resulting in 20 regions. In line with Putnam et al. we study sub-units of a country.

122

CHAPTER FIVE

authors have shown the importance of trust in economic transactions. These studies can be seen as an extension of Williamson’s (1975, 1985, 1998) transaction cost theory. Ring and Van de Ven (1992) have shown that informal, personal connections between and across organizations play an important role in determining the governance structures used to organize their transactions. Gulati (1995) pointed to the fact that both transaction cost elements as well as social factors are relevant and im­ portant in studying inter-firm relationships and cooperation. Repeated ties between firms engender trust that is manifested in the form of the contracts used to organize subsequent alliances. Trust within social net­ works provides options for control through third parties and serves there­ fore as a substitute for a legal system. This function is related to the re­ duction of transaction costs, the costs of running the economic system. Moreover, trust is linked with the facilitation of highly uncertain and complex transactions. It reduces the uncertainty of these kinds of trans­ actions. Uzzi (1996) shows in a study on the apparel industry in New York that trust facilitates the exchange of resources and information that are crucial for high performance but are difficult to value and transfer via market ties. This second function of trust is related to its information function. As Malecki puts it (2000: 195) ‘through the economic and so­ cial relationship in the network, diverse information becomes inexpen­ sive to obtain’. When discussing alliances, Gulati (1998: 308) argues that ‘trust not only enables greater exchange of information, but it also promotes ease of interaction and a flexible orientation on the part of each partner’. It operates as a mechanism that facilitates communication and co-operation between firms. For example, trust relationships can result in a supplier exceeding contractual requirements, whether by early de­ livery, higher quality, or some other means of assuring goodwill (Sako, 1992). Or as Williamson (1985: 62) states: ‘where personal integrity is believed to be operative, individuals (...) may refuse to be part of oppor­ tunistic efforts to take advantage of the letter of the contract when the spirit of the exchange is emasculated’. Nooteboom (1999) even reasons that too detailed and formal contracts may seriously inhibit the growth of trust. Trust and contractual safeguards are to some degree substitutes. Among those who see trust as a substitute for rules and contracts, Ken­ neth Arrow (1971: 22) is perhaps the most explicit: ‘It is useful for indi­ viduals to have some trust in each other’s word. In the absence of trust, it would become very costly to arrange for alternative sanctions and gua­ rantees, and many opportunities for mutually beneficial cooperation would have to be foregone’. (emphasis added). According to Fukuyama (1995), societies endowed with generalized trust enjoy a form of social capital, that—complementary to traditional

PARTICIPATION IN CIVIL SOCIETY

123

factor endowments like labor and capital—contributes at least as much to their success in modern economic competition. Generalized trust is based on a set of ethical habits and reciprocal moral obligations interna­ lized by members of a community (Fukuyama, 1995). High trust socie­ ties can do with fewer regulations and coercive enforcement mecha­ nisms. In this view, trust is seen as a substitute for contracts. But in case an institutional system functions properly, the function of trust should be seen in the light of the facilitation of complex transactions. It lowers transaction costs and moreover, it contributes to flexibility. Fukuyama argues that non-family or generalized trust is therefore of importance for successful performance in advanced economies. First, trust allows for the dis-embedding of social relations and second, trust allows for coope­ ration without the direct influence of power and market. Korczynski (2000) argues that these two functions are of crucial importance to ad­ vanced capitalist economies given their increasingly globalized and tur­ bulent nature. Thus, trust not only serves as a substitute for legal sys­ tems, but also functions as a facilitator of complex transactions that even in case of a well-functioning institutional system cannot be fully ‘ar­ ranged’ in terms of contracts. Hence, in general the economic function of trust refers to the reduc­ tion of transaction costs and its influence on promoting co-operation and reducing the need (costs) for intervention to prevent or correct dishones­ ty. But also from a sociological point of view, trust has several func­ tions. Especially Parsons’ (1969) study and Luhmann’s (1979) study are important in this respect. Parsons places trust in the center of the con­ struction of social order. In Parsons’ view, a common value system based on widely shared norms and values, stabilizes interactions in a so­ cial system. Trust is grounded in pre-existing consensus and is a product of an effective integration of norms and values. Trust fulfils an integra­ tive function in the establishment of social order. The second function of trust in sociological thinking has been put forward by Luhmann in 1979. He views trust as a social mechanism that reduces complexity and en­ ables individuals to deal with the complexity and contingency of modern life. This corresponds with Williamson’s (1985) argument that exchange relations that feature personal trust will survive greater stress and will display greater adaptability. 3 Group membership Regarding the function of associational activity and its link to economic growth, theory is less clear than with respect to trust (Bertrand et al.,

124

CHAPTER FIVE

2000). We distinguish two functions of associational activity or group membership on welfare. Putnam et al. (1993) show that networks relationships improve the ef­ ficiency of society by facilitating coordinated actions. Their study on Italian regions has shown that the critical factor in explaining effective­ ness of regional governments and regional economic performance in Ita­ ly is to be found in regional differences in social structure. Effective go­ vernance hinges critically on traditions of civic engagement and the structure of the civic networks. In regions where social relationships are more horizontal, based on trust and shared values, participation in social organizations is higher and social capital is higher. They conclude that regions, in which the regional government is more successful and the economy is more efficient, are characterized by horizontal relations that both favored and fostered greater networks of civic engagement and le­ vels of organization in society. The reason Putnam et al. (1993: 88) spe­ cifically study the degree of civic community membership is that ‘Citi­ zens in a civic community, though not selfless saints, regard the public domain as more than a battleground for pursuing personal interest’. In this way fewer resources are used incurring transaction costs. Or as Leo­ nardi (1995: 169) writes, high social capital means that citizens accept the positive role played by collective action (organized group behavior) in pursuing collective goods. The second function of associational activity is closely related to the theory of networks and the advantages of being embedded in networks. There are two theoretical approaches for understanding how social rela­ tions and networks create economic and social benefits (Gargiulo & Be­ nassi, 2000; Uzzi, 1999). The weak-tie approach argues that a large net­ work of arm’s-length ties is most advantageous. On the other hand there is the strong-tie approach which claims that a closed tightly knit network of embedded ties is most advantageous. This corresponds with the two opposite views in literature on the optimal structure of networks. Whereas Coleman (1990) argues that closed networks may provide a better basis for cooperation, Burt (1992) stresses cohesive ties as a source of rigidity. However, in both cases the core of the argument re­ lates to the transfer of knowledge between actors. In Burt’s (1992) con­ cept, structural holes are important sources of new information. A fun­ damental idea that inspired Burt’s structural-hole theory is Granovetter’s description of the ‘strength of weak ties’ (Granovetter, 1973). Granovet­ ter reasoned that access to new information is obtained through an ego’s weak ties to nodes at a distance from his own local network. The reaso­ ning is that information within the local network is widely shared lo­

PARTICIPATION IN CIVIL SOCIETY

125

cally, hence most of the local contacts are redundant. New information comes from non-redundant ties. Though Coleman’s closed network approach seems to be opposite to Burt’s view of structural holes (open networks), Coleman states that ex­ actly the closure of the network and the embeddedness of the actors pro­ vide opportunities to obtain information that otherwise would be impos­ sible or too expensive to obtain. In both views, embeddedness in net­ works creates advantages like increased sources of information, and ob­ taining information that is not easily available (spillover effects). In sum, the economic function of associational activity contains two elements. The first refers to the concept of collective action and argues that organized group behavior may lead to the generally shared idea that the pursuit of collective goods is not seen as contradictory to the achievement of personal wealth. Associational activity limits the costs of free riding. Secondly, embeddedness in networks (group membership) promotes the spillover of knowledge and information between the dif­ ferent actors involved. 4 Empirical test In order to test whether social capital influences regional economic growth or not, we investigate 54 European regions. By doing so, we are able to test whether the Putnam hypothesis on social capital based on Italian regions can be generalized. We also test whether Fukuyama’s statement on generalized trust and its importance for economic growth holds for our sample of European regions. In addition, there are other advantages of investigating regions in Europe. First of all, the set of re­ gions is relatively homogeneous compared with studies on culture and economics that incorporate countries like Taiwan and Germany or Japan and the United States in the same regression analysis. Temple’s critical comment (1999) that countries differing widely in social, political and institutional characteristics are unlikely to fall on a common surface, is heeded by taking this relatively homogeneous set of European regions. A second advantage of studying regions is the number of observations. Instead of only 29 countries (Knack & Keefer, 1997), we study 54 re­ gions. Most important, however, is the fact that by comparing national cultures, ‘we risk losing track of the enormous diversity found within many of the major nations of the world’ (Smith & Bond, 1998: 41). By studying regions and regional differences this risk is limited.

126

CHAPTER FIVE

5 Data Data is taken from the European Value Study (EVS), which is a survey on norms and values. The survey comprises three waves (1981/ 1990/1999), of which we use the second one. In order to obtain regional scores on trust we had to regroup the original individual data. We did not use the first wave that was carried out in 1981, because we could not trace the individual scores in terms of regions. The latest wave, 1999/2000 was not completed by the time we finished this chapter. Therefore we use the 1990 data. The set comprises 7 countries, i.e., France, Italy, Germany, Spain, The Netherlands, Belgium, and the United Kingdom. In order to compare the data on norms and values with regional economic data we used the Eurostat definition of regions. The regional level in our analyses is the NUTS1 level. This implies that France consists of 8 regions, Italy 11 (including Sicily and Sardinia), Germany 11 (former eastern regions excluded), Spain 7, The Netherlands 4, Belgium 3, and the UK 10 (including Scotland, excluding Northern Ireland). The total number of regions equals 54 (see Figure 5.1). The numbers of the European regions are defined in Table 5.1.

54 45 14

46

52

47 48 53 51 49 50

7 41 10 43 42 44 11 21 9 24 3 12 22

6

5

4 25

26 28 27 29

17 18

NUTS1 level

included (54) not included (22)

33 34

15 16

European regions

32

31 30

35 36

19

37 38 40

0

0 20

39

21

Figure 5.1 Map of European regions

PARTICIPATION IN CIVIL SOCIETY

Table 5.1 Nr. Region 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

Reg. Bruxelles-Cap. Vlaanderen Wallonie Baden-Württemberg Bayern Berlin Bremen Hamburg Hessen Niedersachsen Nordrhein-Westfalen Rheinland-Pfalz Saarland Schleswig-Holstein Noroeste Noreste Madrid Centro Este Sur Canarias Île de France Bassin Parisien Nord-Pas-de-Calais Est Ouest Sud-Ouest

127

Data for European regions NUTS1 code BE1 BE2 BE3 DE1 DE2 DE3 DE5 DE6 DE7 DE9 DEA DEB DEC DEF ES1 ES2 ES3 ES4 ES5 ES6 ES7 FR1 FR2 FR3 FR4 FR5 FR6

NUTS1 code Centre-Est FR7 Méditerranée FR8 Nord Ovest IT1 Lombardia IT2 Nord Est IT3 Emilia-Romagna IT4 Centro IT5 Lazio IT6 Ambruzzo-Molise IT7 Campania IT8 Sud IT9 Sicilia ITA Sardegna ITB Noord-Nederland NL1 Oost-Nederland NL2 West-Nederland NL3 Zuid-Nederland NL4 North UK1 Yorkshire and Humberside UK2 East Midlands UK3 East Anglia UK4 South East UK5 South West UK6 West Midlands UK7 North West UK8 Wales UK9 Scotland UKA

Nr. Region 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54

6 Trust The question we used to asses the level of trust in a society is: ‘General­ ly speaking, would you say that most people can be trusted, or that you cannot be too careful in dealing with people?’. After deleting the number of respondents that answered ‘don’t know’, we took the fraction of peo­ ple that answered ‘most people can be trusted’. For our sample of 54 regions we have obtained scores on trust. These scores range from 5.5% of the respondents answering that most people can be trusted in Sardegna in Southern Italy to 64.6% in the eastern part

128

CHAPTER FIVE

of the Netherlands. Mean score equals 0.35 with a standard deviation of 0.11. In Figure 5.2 the scores on percentage of people answering that most people can be trusted are shown.

Trust 0.447to 0.646 0.392to 0.447 0.33 to 0.392 0.263to 0.33 0.055to 0.263

(10) (11) (11) (11) (11)

Figure 5.2 Trust scores at NUTS1 level in Europe

As can be seen in Figure 5.2, the regional scores on trust differ considerably within Europe. When looking at countries, we see, for example, that the Netherlands are rather homogeneous in terms of trust, but regions in Italy differ a lot. Putnam et al. (1993) seemed right in the case of Italy, when describing the differences between the Northern and the Southern regions. The North has higher scores on trust than the South. However, at first sight such a picture for Europe as a whole cannot be obtained. While some researchers have suggested that religion, especially Protestantism, correlates with trust (e.g., Inglehart, 1990; Knack & Keefer, 1997: 1283), our regional analysis suggests this is not the case. Traditional Catholic regions in the South of the Netherlands, Flanders, Madrid, and the North of Italy all fall in the group of regions that have the highest scores on trust (0.447-0.646)3, far above average (see Figure 5.2). ______________ 3

The ranges in the figures are based on the equal count criterion in the Mapinfo Geographical Information System (GIS)-application.

PARTICIPATION IN CIVIL SOCIETY

129

7 Group membership Besides interest in general trust, Putnam et al. (1993) explicitly studied memberships of clubs and associations. They suggested that dense hori­ zontal networks positively affect the level of trust and citizenship4. As mentioned earlier, social capital is often perceived in terms of networks and being member of such a network. Knack and Keefer (1997: 1273) find no direct effect of membership of groups on economic growth or investment. Horizontal networks—as measured by membership in groups—are unrelated to trust and civic norms (controlling for education and income) and to economic performance (1997: 1284). Their explana­ tion is the possible counter-effects of Putnam and Olson groups. Whereas Putnam groups may be evoking positive effects, these may be reduced by harmful effects of the Olson groups. Olson (1982) argued that as a society’s networks of associations become highly developed, they strangle economic growth (e.g., lobbying costs). Similar to Knack and Keefer, we measure the average number of groups cited per respondent in each country. However, as the authors ar­ gue, the categories are overly broad (1997: 1274). Second, and more im­ portant, is the fact that level of involvement is not measured, which may reduce the validity of this measure of social capital. As mentioned be­ fore, the hypothesized benefits of network embeddedness may not be captured when taking passive membership of groups and associations. Therefore, we have taken active membership (A-Groups) of a number of associations instead of passive membership. For reasons of clarity, we depicted an overview of the different measures of social capital in Figure 5.3. The question we use to measure group membership, is stated as fol­ lows: ‘Which, if any do you belong to’. The categories are: a. b. c. d. e. f. g. h.

Social welfare services for elderly handicapped or deprived people Religious or church organizations Education, arts, music or cultural activities Trade unions Political parties or groups Local community action Third world development or human rights Conservation, the environment, ecology

______________ 4

This argument is not new. Already in 1835 Tocqueville argued that membership in voluntary associations was conducive to democracy. Putnam (1993) however extends the argument and argues that voluntary associations are not only conducive to democracy, but also to economic development.

130 i. j. k. l. m. n. o.

CHAPTER FIVE

Professional associations Youth work Sports or recreation Women’s groups Peace movement Animal rights Voluntary organizations concerned with health.

The above categories are the same for our measures of passive and ac­ tive group membership. The difference between the two is that in case of active membership respondents are not only a member but also do vo­ luntary work for the particular association. As described earlier and in line with Putnam et al. (1993), we think of the level of doing unpaid vo­ luntary work as an indication of collective feelings of responsibility. As such these moral norms may have positive effects on the provision of public goods. Moreover, as we argued in the previous section on net­ work theory, these networks may provide spillover channels (Oerlemans et al., 2001).

Social Capital

Trust (+)

Networks

Passive group Membership (Groups)

Active group Membership (A-Groups +)

Among others consist of: - Putman groups (P-groups +) - Olson groups (O-Groups –)

Figure 5.3 An overview of the different measures of social capital (+ or – indi­ cates direction of hypothesized relationship with regional economic growth)

The scores are obtained by taking the average score per region of res­ pondents answering yes to the question mentioned above. Furthermore,

131

PARTICIPATION IN CIVIL SOCIETY

we have calculated regional scores on the Putnam and Olson groups corresponding to Knack and Keefer’s analysis at a country level5. The Putnam groups refer to membership of b) religious organizations, c) education, arts, music or cultural activities and j) youth work. The Olson groups consist of membership of d) trade unions, e) political parties of groups, and i) professional associations. Figures 5.4 and 5.5 show the scores on the P-Groups and O-Groups respectively according to the definition of Knack and Keefer (1997).

P-Groups 0.32 to 0.89 0.27 to 0.32 0.17 to 0.27 0.12 to 0.17 0.03 to 0.12

(12) (11) (11) (7) (13)

Figure 5.4 Regional scores on P-Groups in Europe

The mean score on the Putnam groups at the regional level is 0.26 with a standard deviation of 0.18. The highest score is found in the eastern part of the Netherlands, with a score of 0.89. This implies that on average 89% of the people is member of at least one of the organizations included in the Putnam groups. The lowest score can be found in Sardegna, Italy, where only 3% of the people are member of at least one of these organizations. The scores on the Olson groups range from 0 in Sardegna, Italy to 0.55 in the eastern part of the Netherlands. The mean value is 0.22 with a standard deviation from 0.12. Table 5.2 shows the mean scores and the standard deviation for the social capital variables. ______________ 5

Note that Knack and Keefer (1997) have fewer types of associations included in their measure of group membership, because of lack of data. In addition to the associations they analyze, we include items k,l,m,n, and o extra.

132

CHAPTER FIVE

O-Groups 0.36 to 0.55 0.27 to 0.36 0.17 to 0.27 0.12 to 0.17 0 to 0.12

(10) (11) (10) (12) (11)

Figure 5.5 Regional scores on O-Groups in Europe

A-Groups 0.59 to 0.821 0.49 to 0.59 0.38 to 0.49 0.28 to 0.38 0.08 to 0.28

(11) (9) (12) (11) (11)

Figure 5.6 Regional scores on A-Groups in Europe

133

PARTICIPATION IN CIVIL SOCIETY

Table 5.2

Descriptive statistics Mean .35 .26 .22 .41 .62 24.25 .51

Trust P-Groups O-Groups A-Groups Groups Investment Schooling

Std. Dev. .11 .18 .12 .17 .38 3.74 .07

N=54.

Table 5.3a Correlation table of social capital variables Growth Trust 1950-1998 Growth 1950-1998 Trust P-Groups O-Groups A-Groups Groups *

-

.05 -

PGroups .16 .42* -

OGroups .23* .52* .72* -

AGroups

Groups

.29* .21 .70* .69* -

.25* .46* .79* .79* .85* ­

p < .10.

Table 5.3b Correlation table of standard economic variables Growth GRP 1950-1998 1950 Growth 1950-1998 GRP 1950 Schooling Investment Aggledu Spillover *

-

–.55* -

Schooling Investment

Aggledu

–.15 .29* ­

–.07 .35* –.10 –.03 -

.13 –.01 –.31* -

Spillover

.05 .17 –.05 –.19 –.19 -

p < 0.10.

Regarding the question on unpaid voluntary work (A-Groups), we ob­ tained an average score of 0.41 and a standard deviation of 0.17. The highest score is obtained in Bremen, Germany (0.82) and the lowest

134

CHAPTER FIVE

score on active membership can be found in Sardegna (0.08); see Figure 5.6. The correlation between active membership and Putnam groups is 0.7, between active membership and Olson groups 0.69 and between ac­ tive membership and interpersonal trust 0.21. Table 5.3a shows the cor­ relation of the social capital variables. 8 Economic data and testing In order to test whether trust and group membership are related to eco­ nomic growth or not, we have taken a standard growth framework, that corresponds with Knack and Keefer’s empirical test, and which includes initial level of GDP per capita, the investment ratio and the school en­ rolment ratio. As the availability of data on the level of European regions is scarce, the number of empirical studies is relatively limited compared to crosscountry studies. A basic reference is Barro and Sala-i-Martin (1995). Similar to these authors, we have computed the regional growth figures by relating the regional GDP per capita information to the country mean6. There are two reasons to use the country mean as a correction factor. First of all we do not have regional price data. Second, the figures on regional GDP are provided in an index form that is not comparable across countries. Hence, we have used Gross Regional Product (GRP) fi­ gures that are expressed as deviations from the means from the respec­ tive countries. The 1950 data is based on Molle, Van Holst and Smit (1980), whereas the data for Spain refers to 1955 and are based on Barro and Sala-i-Martin’s (1995) calculations. The 1998 data on GRP is drawn from Eurostat information. Regional growth figures are depicted in Fi­ gure 5.7. If we compare Figure 5.7 on Growth with Figure 5.2 on Trust it is hard to observe a clear overlap in the two figures. Regions that grow fast do not seem to be characterized by high scores on trust. If we look at the correlation between Growth and Trust this is confirmed by the correla­ tion of only 0.05 (see Table 5.3a). The picture is less clear with respect to the dimensions of group membership. The correlation between Growth and the different measures of group membership is around 0.25 with the highest correlation of 0.29 between Growth and A-Groups (see Table 5.3a). The figures and correlation tables show that the relationship between our social capital variables and regional economic growth do ______________ 6

Gross Regional Product of a region in 1950 is divided by the mean of the Gross Re­ gional Products of all regions belonging to a certain country. A similar formula is applied to calculate the 1998 relative regional product. Regional growth over the period 1950­ 1998 is then based on these two indices.

135

PARTICIPATION IN CIVIL SOCIETY

not seem to be that strong. However, the question is if the conclusion is right when controlling for other economic variables, like investment in physical capital; see table 5.3b.

growth 1950-1998

0.38 to 0.75 0.13 to 0.38 -0.04 to 0.13 -0.28 to -0.04 -0.75 to -0.28

(10) (11) (11) (10) (12)

Figure 5.7 Regional growth 1950-1998

Investment ratio is measured at country level. Data is taken from the Penn World Tables 5.6. The period for which we have calculated the average of the investment ratio is 1950-19927. Apart from availability of data, another reason to take the country level investment data and not the regional scores, is the underlying assumption of a closed economy. Because of spatial interaction, regional investment figures would only provide a limited understanding of regional economic growth (Nijkamp & Poot, 1998). Therefore we have taken the country level data. School enrolment ratio measures the total number of pupils at first and second level in 1977, divided by total number of people in the corresponding age group. The growth period we analyze is 1950-1998. The school enrolment rate in 1977 falls in between these dates and given the fact that school enrolment rates have increased since 1950, the 1977 information is a reasonable proxy for the average over the entire period. Data comes from Eurostat. Data on school enrolment rates in Spanish regions refers to 1985. We have taken uncorrected regional figures be______________ 7

Penn World Tables 5.6 provides data up to 1992.

136

CHAPTER FIVE

cause it has been shown that migration plays only a minor role in Euro­ pean regions and the relation with per capita GDP is weak (Barro & Sala-i-Martin, 1995; Begg, 1995). The basis for our analyses is the standard ‘Barro’ type of a growth re­ gression, including the investment in physical capital, human capital and the initial level of economic development. In order to control for concen­ tration of human capital in major agglomerations, we included a variable that consists of the score on the school enrolment rate multiplied by a dummy variable for the region in which a major agglomeration is lo­ cated8. Furthermore we tested whether spatial correlation influences our results. Ideally one should use interregional input-output tables to calcu­ late regional multipliers and construct a variable that controls for spatial correlation9. However, this information was not available. In order to control for spatial correlation, we applied Quah’s (1996) approach and calculated the so-called neighbor relative income. This method implies that we use average per capita income of the surrounding, physically contiguous regions to control for spatial auto-correlation. In our sample, however, the 1950 GRP data is related to national average and therefore reflect regional welfare relative to country mean. By using these data we implicitly assume that scores for neighboring regions in foreign coun­ tries influence regional growth if the welfare in this neighboring region is relatively high compared to their own national average. Of the 54 re­ gions in the sample, 19 have neighboring regions in countries other than the region’s own host itself, whereas 4 had no neighboring regions at all10. Hence, our basic regression analysis includes initial level of welfare (GRP1950), school enrolment rate (Schooling), investment ratio (Invest­ ment), spatial correlation (Spillover) and a human capital concentration dummy for agglomerations (Aggledu); see Table 5.3b. We considered log-specifications for the first three variables. The results are shown in Table 5.4. The first model we estimated is the standard model. As the results show, all variables except for the school enrolment rate are significant at the 5% level. Schooling is significant at the 10% level. The initial level of welfare is negatively related to economic growth, which supports the ______________ 8

We selected the Western part of the Netherlands, Greater Paris, Greater Berlin, Greater London, Barcelona area, Brussels, and the Italian region Lazio (Rome). 9 There exist other ways to have a more refined control variable that can be taken into consideration, for example, the physical length of abutting boundaries or the physical characteristics of the border terrain. However, these kinds of extensions go beyond the scope of the current chapter. 10 The average number of physical neighbour regions is 3.3, which corresponds with Quah’s score of 3.3.

137

PARTICIPATION IN CIVIL SOCIETY

convergence hypothesis. This corresponds with other findings on regio­ nal convergence in Europe (Martin & Sunley, 1998). However, if we Table 5.4

Regression results: Trust, group memberships and regional economic performance, 1950-1998

Model Dependent Variable Constant GRP1950 Investment Schooling Aggledu Spillover

1 –1.44** (.623) –.971*** (.201) .476** (.203) .527* (.314) .528*** (.195) .308*** (.093)

Trust P-Groups O-Groups

2

3 Growth 1950-1998 –1.45** –1.49** (.629) (.611) –.968*** –.938*** (.212) (.196) .481** .553*** (.210) (.201) .518 .397 (.329) (.244) .522** .423** (.214) (.204) .301** .213** (.118) (.103) .011 (.086) .007 (.063) .119** (.056)

4 –1.37** (.584) –.942*** (.190) .484** (.188) .449* (.258) .404** (.209) .233** (.101)

–1.01* (582) –.969*** (.196) .422** (.184) .569** (.232) .472** (.197) .244** (.097)

.109** (0.41)

Groups

.175*** (.054)

A-Groups

R-square F-value CW-test VIF (Maximum)

5

0.4089 5.80 .6845 1.49

0.4090 5.06 .6907 1.53

0.4673 5.63 .4543 2.45

0.4641 7.16 .8885 1.50

0.4813 7.56 .8596 1.49

Standard errors (White corrected) between parentheses. N=54. *** p < .01, ** p < .05, * p < .1. CW test refers to the Cook-Weisberg test for heteroskedasticity. Values above 0.05 indicate heteroskedasticity is not problematic. VIF refers to Variance Inflation Factor and values above 10 are indications of multi­ collinearity inflating the R-square. We considered log-specifications in our analysis. In case we do not take the log- specifications, results are not influenced. We also tested for country-specific effects and possible interaction effects. Results indicate that O-Groups are not significant when country-specific effects are included. The overall conclusion on A-Groups is not influenced. An overview of these additional tests can be found in a ‘statistical appendix’, which is available upon request.

138

CHAPTER FIVE

take shorter periods of time (e.g., 1970-1998) we cannot find proof for the convergence hypothesis. This corresponds with findings on country (Levine & Renelt, 1992) and regional level (Fagerberg & Verspagen, 1995). The period in the eighties can be roughly characterized by diver­ gence instead of the observed convergence in the period before (Maur­ seth, 2001). However, based on our sample, we conclude that for the overall after war period, the convergence hypothesis holds, i.e., the growth rate of per capita GDP is negatively related to the starting level of per capita GDP. To test the hypothesized positive relation between social capital and economic growth, we included social capital variables discussed above. First, as shown in the second model specification in Table 5.4, we added the scores on generalized trust. The Trust variable is not significant. This might seem surprising given the results of Knack and Keefer’s study (1997) on country level in which it was found that trust significantly in­ fluenced economic growth between 1980 and 1992 in 29 countries. However, as has been shown by Beugelsdijk et al. (2002), Knack and Keefer’s findings on trust are not statistically robust. In the third model we included group membership. In accordance with Knack and Keefer (1997) we split this variable up in two sub­ groups, namely O and P-Groups. As can be seen in Table 5.4, P-Groups are not significant and O-Groups are significant at the 5% level. In the fourth model we include the measure for passive group membership. Passive group membership has a significant and positive influence on regional growth rate. In the final step we included the variable that indi­ cates active membership. The active membership variable (A-Groups) is highly significant (1% level) and as a consequence, the resulting model has the highest variance explained. Hence, active membership, doing unpaid voluntary work is significantly correlated with regional economic growth. The question is whether these findings are robust. In order to test whether the above findings are robust or not, we per­ formed several tests. First, we tested for multi-collinearity and hetero­ skedasticity. As the results in Table 5.4 indicate, these do not signifi­ cantly influence the results. The Cook-Weisberg (CW) test for hetero­ skedasticity and the Variance Inflation Factor (VIF) for multi­ collinearity both indicate that in the models specified in Table 5.4, these are not problematic and do not influence the results. However, a sensi­ tivity analysis that only consists of test for multi-collinearity and he­ teroskedasticity is not complete. We choose to extend our sensitivity analysis in several ways, among which the recursive method and tests based on the Extreme Bounds Analysis (EBA).

PARTICIPATION IN CIVIL SOCIETY

139

First, we performed several regression analyses in which the different social capital variables are combined. Regarding Trust and the Putnam and Olson groups, results do not change. However, if we include both active and passive membership, passive membership becomes insignifi­ cant. As the correlation between the active and passive membership is 0.85 (see Table 5.3a), this is likely to be due to multi-collinearity pro­ blems. In case we perform a regression analysis in which as well Groups and A-Groups are included, multi-collinearity analysis shows that the variance inflation factor for these two variables is larger than 4.9, whereas the rest of the variables do not exceed 1.5. Though rule of thumb reads that VIF’s exceeding 10 are problematic, we consider the strong correlation and the VIF analysis as an indication of problematic multi-collinearity. Next, we have applied the recursive method to test if the composition of the sample influences our results. First we order the 54 observations according to a certain variable. In this case we chose for regional eco­ nomic growth. This means that the first observation is the region with the lowest growth rate over the period 1950-1998, and observation 54 is the fastest growing region over this period. The recursive method im­ plies that based on the order in which the observations are represented observations are deleted and the coefficients are estimated based on this smaller sample. In Figure 5.8 we have plotted the coefficient of Trust when the order of observations is based on growth, according to the sec­ ond model of Table 5.4. The line in the middle plots the value of the regression coefficient for Trust. The outer lines represent 95% confidence intervals. The horizontal axis represents the observations, where observation 54 is the fastest growing and observation 1 the slowest growing region. The vertical axis represents the value of the Trust coefficient at a certain number of ob­ servations. If 54 observations are included the value of the Trust coeffi­ cient equals 0.011, which can be seen in the figure on the far right of the horizontal axis and corresponds to the results in Table 5.4. Moving from the right to the left on the horizontal axis means deleting regions that are the fastest growing. For example, observation 50 implies that the 4 fast­ est growing regions are deleted. The corresponding value of the Trust coefficient based on the sample of 50 regions is close to 0.011. The sta­ ble line in Figure 5.8 leads us to conclude that the Trust coefficient is in­ dependent of the deletion—or inclusion—of fast growing regions in the sample. We performed similar tests for the other variables, that all behaved in a stable way or according to economic theory. In the latter case we refer to the initial level of welfare. Inclusion of fast growing regions causes

140

CHAPTER FIVE

0.5 0.4 0.3 0.2 0.1 0.0 -0.1 -0.2 -0.3 -0.4 -0.5 -0.6 1

6

11

16

21

26

31

36

41

46

51

54

Observations

Figure 5.8 Coefficient and bands of Trust based on recursive OLS

0.4 0.3 0.2 0.1 0.0 -0.1 -0.2 -0.3

1

6

11

16

21

26

31

36

41

46

51

54

Observations

Figure 5.9 Coefficient and bands of A-groups based on recursive OLS

PARTICIPATION IN CIVIL SOCIETY

141

the coefficient of GRP1950 to decrease (more negative), which corres­ ponds to the convergence hypothesis. Figure 5.9 represents the results of the recursive method for the A-Groups, according to the fifth model of Table 5.4. Observations are again ordered according to regional economic growth. As the figure shows, the coefficient of A-Groups slightly increases when faster growing regions are included. 9 Implications, limitations and suggestions for further research Economists show increased interest in the concept of social capital. An important study in this field of social capital is Putnam’s study on Italian regions. He showed that differences in economic performance and the well functioning of the institutions in Northern and Southern Italy can be traced back to differences in social capital. In this chapter we studied 54 regions in Europe and applied a stan­ dard economic model to test whether the Putnam hypothesis can be ge­ neralized. Social capital is operationalized in terms of generalized trust and associational activity, split up in several elements. Similar to Knack and Keefer (1997) we made a distinction between Putnam groups and Olson groups, and in addition we distinguished active and passive mem­ bership. We have shown several things. First, we found that the convergence hypothesis at the regional level holds. Second, we could not find robust proof for the significant influence of investments and regional school en­ rolment rates on regional economic growth. Third, social capital in terms of trust is not related to economic growth at the regional level in Europe. Fourth, Knack and Keefer’s distinction between P- and O-Groups does not yield additional insights. The findings on O-groups are not robust. However, the main implication of our study is that we found that social capital in terms of (active) group membership contributes to regional economic growth in Europe. We have shown the hypothesis put forward by Putnam et al. (1993) that social capital matters for regional economic success, can be genera­ lized to the extent that it is not only the existence of social networks that contribute to regional economic growth, but also the actual level of in­ volvement in it. In contrast with the country-level analysis of Knack and Keefer (1997), our regional analysis does not support the hypothesis that trust is positively correlated with economic growth. Besides a lack of statistical robustness of Knack and Keefer’s (1997) study (Beugelsdijk et al., 2002), we feel these different results regarding trust are caused by a sample selection bias. In fact, by choosing a sample of nations ranging from a less developed country like Nigeria to wealthy countries like the

142

CHAPTER FIVE

Netherlands or the USA, Knack and Keefer may have overestimated the importance of trust. The cross-country differences in institutional matu­ rity in their sample may lead to this overestimation of trust. A similar line of reasoning holds for the role of the Olson groups. Remember that the O-Groups include membership of trade unions. Given the institu­ tional differences across countries, the probability that someone is a member of a trade union differs considerably across relatively rich and poor countries. In a country like Nigeria membership of a trade union is not as natural, even impossible in some cases, as in a relatively advanced country like Germany. By taking regions in Europe the danger of overes­ timation of the importance of social capital as a result of too large insti­ tutional differences is limited. Our findings regarding active membership may have implications for policymakers. We showed that social capital in terms of active volun­ teering work is positively related to regional economic growth. Does this mean that governments may want to increase active membership of all kinds of associations? Does this imply that policymakers need to take a new look at the relation between labor and leisure? It is clear that a number of factors that policymakers can influence are related to the de­ gree of associational activity. However, as long as we do not exactly know the mechanism between social capital (measured by active mem­ bership) and regional economic growth, it is too early to formulate clear policy implications. Obviously this study suffers from a number of limitations. First of all, lack of proper regional economic data forced us to use country relative regional products. Second, the period of observation is 1950-1998, whereas the social capital data refers to 199011. Ideally, one would prefer social capital data referring to the start of the period of analyses. How­ ever, the earliest period of which we have data on our measurement of social capital (1981) is highly correlated to the 1990 data we used (over .90)12. Future research should focus on the exact mechanisms through which social capital in terms of associational activity influences economic growth in the European regions. As we described in the section on group ______________ 11

We also performed regression analyses on the period 1970-1998. Although we miss data for Spain in 1970 and the number of observations is reduced to 47, results show that overall fit decreases and significance levels generally go down, but overall conclusion holds. 12 Knack & Keefer (1997: 1257) also discuss the validity of the trust measure and conclude that there is no severe noise in this survey-based measure of social capital. They base their conclusion on experiments conducted by the Reader’s Digest and reported in The Economist, June 22, 1996. In an experiment of ‘accidentally’ lost wallets, the per­ centage of wallets returned in each country closely tracks the European Values Study measure of trust.

PARTICIPATION IN CIVIL SOCIETY

143

membership, theory argues that associational activity may promote the spillover of knowledge in networks and, second, may limit the costs of free riding through feelings of collectivity. However, there is no clear understanding how these mechanisms exactly work. More insight in these mechanisms is especially important for policymakers at the re­ gional, national and European level. As we discussed above, the impor­ tance of (active) membership of all kinds of associations for regional economic growth may lead to a re-thinking of the relation between work and spare time. But before actual policy plans in this field are developed, we need to know more about the mechanism between active group membership and regional economic growth. The current attempts and activities of the Worldbank in the field of social capital and developing countries are worth mentioning. The importance of network relation­ ships and the promotion of associational activity have led to a number of successful development projects. Increasingly, the Worldbank acknow­ ledges that social capital may play a crucial role in the reduction of po­ verty and the success of development programs. Social capital is inte­ grated into Worldbank policies in a number of ways13. Nevertheless, these initiatives mainly focus on developing countries and the question remains if the relationship between social capital and economic growth is the same for rich and poor countries. A related question is whether the lack or abundance of social capital influences the success of the regional development programs in the less favored regions of Europe. It would be interesting in future research to relate the degree of success of the Structural Funds of the EU in certain regions to the presence (or absence) of social capital. References Abramovitz, M. 1986. Catching up, forging ahead, and falling behind. Journal of Economic History 46: 385–406. Arrow, K. 1971. Political and economic evaluation of social effects and externalities. Pp. 3–25 in M. Intriligator (ed.), Frontiers of qualitative economics. Amsterdam: North-Holland. Barro, R.J. 1991. Economic growth in a cross section of countries. Quarterly Journal of Eco­ nomics 106: 407–443. —— & X. Sala-I-Martin 1995. Economic growth. New York: McGraw-Hill. Baumol, W. 1986. Productivity growth, convergence, and welfare: what the long run data show. American Economic Review 76: 1072–1085. Begg, I. 1995. Factor mobility and regional disparities in the European Union. Oxford Review of Economic Policy 11: 96–112. Bertrand, M., E.F.P. Luttmer & S. Mullainathan 2000. Network effects and welfare cultures. Quarterly Journal of Economics 65: 1019–1055. ______________ 13

For an overview of the Worldbank social capital initiatives we refer to www.worldbank.org/poverty/scapital/bank2.htm.

144

CHAPTER FIVE

Beugelsdijk, S., H.L.F. de Groot & A.B.T.M. van Schaik 2002. Trust and Economic Growth; a Robustness Test. Tinbergen Discussion Paper 049/3. Burt, R. 1992. The social structure of competition. Pp. 262–287 in N. Nohria & R. Eccles (eds.), Networks and Organizations, Structure, Form and Action. Boston: Harvard Busi­ ness School Press. Coleman, J.S. 1990. Foundations of Social Theory. Cambridge: Harvard University Press. De Tocqueville, A. (1835) 1969. Democracy in America. New York: Scribner Books. European Commission 1999. The European Regions: Sixth Periodic Report on the SocioEconomic Situation in the Regions Of The European Union. Luxembourg: Official Publi­ cation Office. European Investment Bank 2000. Regional convergence in Europe, Theory and empirical evi­ dence. EIB papers 5: 13–24. Fagerberg, J. & B. Verspagen 1995. Heading for Divergence? Regional Growth in Europe Re­ considered. MERIT working paper 2/95–014. Fedderke, J., R. de Kadt & J. Luiz 1999. Economic growth and social capital: a critical reflec­ tion. Theory and Society 28: 709–745. Fukuyama, F. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York: The Free Press. —— 1995a. Social capital and the global economy. Foreign Affairs 74: 89–103. Gambetta, D. 1988. Trust: Making and Breaking Cooperative Relations. Oxford: Blackwell Pu­ blishers. Gargiulo, M. & M. Benassi 2000. Trapped in your own net? Network cohesion, structural holes and the adaptation of social capital. Organisation Science 11: 183–196. Granato, J., R. Inglehart & D. Leblang 1996. The effect of cultural values on economic deve­ lopment: theory, hypotheses, and some empirical tests. American Journal of Political Science 40: 607–631. Granovetter, M. 1973. ‘The strength of weak ties’. American Journal of Sociology 78: 1360– 1380. Grier, K.B. & G. Tullock 1989. An empirical analysis of cross-national economic growth, 1951–1980. Journal of Monetary Economics 24: 259–276. Gulati, R. 1995. Does familiarity breed trust? The implications of repeated ties for contractual choice in alliances. Academy of Management Journal 38: 85–112. —— 1998. Alliances and networks. Strategic Management Journal 19: 293–317. Hall, R.E. & C. Jones 1999. Why do some countries produce so much more output per worker than others? Quarterly Journal of Economics 114: 83–116. Helliwell, J.F. 1996. Economic Growth and Social Capital in Asia. NBER Working Paper 5470. Hofstede, G. 1991. Cultures and Organizations, Software of the Mind. London: McGraw-Hill. —— 2001. Culture’s Consequences; Comparing Values, Behaviors, Institutions and Organiza­ tions across Nations. Beverly Hills: Sage. Inglehart, R. 1990. Culture Shift in Advanced Industrial Society. Princeton: Princeton Univer­ sity Press. —— 1997. Modernization And Postmodernization: Cultural, Economic and Political Change in 43 Societies. Princeton: Princeton University Press. Inkeles, A. 2000. Measuring social capital and its consequences. Policy Sciences 33: 245–268. Knack, S. & P. Keefer 1997. Does social capital have an economic pay-off? A cross country investigation. Quarterly Journal of Economics 112: 1251–1288. Korczynski, M. 2000. The political economy of trust. Journal of Management Studies 37: 1–21. Kormendi, R.C. & P.G. Meguire 1985. Macroeconomic determinants of growth; cross country evidence. Journal of Monetary Economics 16: 141–163. Leonardi, R. 1995. Regional development in Italy, social capital and the Mezzogiorno. Oxford Review of Economic Policy 11: 165–179. Levine, R. & D. Renelt 1992. A sensitivity analysis of cross-country regressions. American Economic Review 82: 942–963. Luhmann, N. 1979. Trust and Power. Chichester: John Wiley & Sons. Malecki, E.J. 2000. Network models for technology-based growth. Pp. 187–204 in Z. Acs (ed.), Regional Innovation, Knowledge and Global Change. London: Pinter.

PARTICIPATION IN CIVIL SOCIETY

145

Mankiw, N.G., D. Romer & D.Weil 1992. A contribution to the empirics of economic growth. Quarterly Journal of Economics 107: 407–431. Martin, R. & P. Sunley 1998. Slow convergence, the new endogenous growth theory and re­ gional development. Economic Geography 74: 201–227. Maurseth, P.B. 2001. Convergence, Geography and technology. Structural Change and Eco­ nomic Dynamics 12: 247–276. Molle, W., B. van Holst & H. Smit 1980. Regional Disparity and Economic Development in the European Community. Westmead: Saxon House. Nijkamp, P. & J. Poot 1998. Spatial perspectives on new theories of economic growth. The An­ nals of Regional Science 32: 7–38. Nooteboom, B. 1999. Innovation, learning and industrial organisation. Cambridge Journal of Economics 23: 127–150. Oerlemans, L.A.G., M.T. Meeus & F. Boekema 2001. On the spatial embeddedness of innova­ tion networks: an exploration of the proximity effect. Tijdschrift voor Sociale en Econo­ mische Geografie 92: 60–75. Olson, M. 1982. The Rise and Decline of Nations. New Haven: Yale University Press. Parsons, T. 1969. Sociological Theory and Modern Society. New York: Free Press. Paxton, P. 1999. Is social capital declining in the United States? A multiple indicator assess­ ment. American Journal of Sociology 105: 88–127. Putnam, R.D. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster. ——, R. Leonardi & R.Y. Nanetti 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press. Quah, D. 1996. Regional convergence clusters across Europe. European Economic Review 40: 951–958. Ring, P.S. & H.A. van de Ven 1992. Structuring cooperative relationships between organiza­ tions. Strategic Management Journal 13: 483–498. Sako, M. 1992. Prices, Quality and Trust: Interfirm Relations in Britain and Japan. Cam­ bridge: Cambridge University Press. Smith, P. & M. Bond 1998. Social Psychology across Cultures. London: Prentice Hall Europe. Swank, D. 1996. Culture, institutions, and economic growth: theory, recent evidence, and the role of communitarian polities. American Journal of Political Science 40: 660–679. Temple, J. 1999. The new growth evidence. Journal of Economic Literature 37: 112–156. —— & P.A. Johnson 1998. Social capability and economic growth. Quarterly Journal of Eco­ nomics 113: 965–990. Uzzi, B. 1996. The sources and consequences of embeddedness for the economic performance of organizations: the network effect. American Sociological Review 61: 674–698. —— 1999. Embeddedness in the making of financial capital: how social relations and networks benefit firms seeking financing. American Sociological Review 64: 481–505. Van Deth, J.W., M. Maraffi, K. Newton & P. Whiteley (eds.) 1999. Social Capital and Euro­ pean Democracy. London: Routledge. Williamson, O.E. 1975. Markets and Hierarchies, Analysis and Antitrust Implications. New York: Free Press. —— 1985. The Economic Institutions of Capitalism. New York: Free Press. —— 1998. Transaction cost economics: how it works; where it is headed. De Economist 146: 23–58. Woolcock, M. 1998. Social capital and economic development: toward a theoretical synthesis and policy framework. Theory and Society 27: 151–208.

This page intentionally left blank

CHAPTER SIX

BRIDGING AND BONDING SOCIAL CAPITAL: WHICH TYPE IS GOOD FOR ECONOMIC GROWTH? SJOERD BEUGELSDIJK & SJAK SMULDERS 1 Introduction Most of our time is spent in the presence of others. We spend our wor­ king time, leisure time, and family hours with others. However, prefer­ ences for socializing differ among individuals and cultures. As Fuku­ yama puts it, ‘Some (societies) show a markedly greater proclivity for association than others, and the preferred form of association differ. In some, family and kinship constitute the primary form of association; in others, voluntary associations are much stronger and serve to draw peo­ ple out of their families’ (1995: 28). Moreover, socializing is time-consuming and may be traded-off against other activities. Participation in the economy and market exchange (working and shopping) compete with social activities, family life and voluntary organizations. We may expect that cultural differences affect the degree to which individuals are more oriented to personal possessions and status. These variations in materialistic attitudes result in different levels of sociali­ zing. What is especially interesting is how these differences in social structure in turn affect economic outcomes. Are countries or regions in which materialistic attitudes dominate characterized by fast economic growth, or does scarcity of socializing somehow hamper growth? Is so­ cializing with family friends and citizens a good in its own, for which some material benefits are happily given up? Or is socializing also in­ strumental in promoting material well-being and increasing economic growth? To study the link between socializing and economic performance, the concept of social capital has been developed, which is often related to trust. Trust and interaction among citizens may stimulate economic growth, when trust facilitates transactions and reduces transaction and monitoring costs in economic exchange. Trust arises mainly within groups with strong social network ties. The repeated interaction among group members prevents opportunistic behavior and cheating in prison­

148

CHAPTER SIX

ers’ dilemma kind of situations. Thus, the formation and maintenance of networks constitutes social capital that works as a productive asset in the economy. However, social interaction and the exploitation of social capital may work in the opposite direction. Closed networks may acts as organiza­ tions that lobby and act against the interests of other groups. Rentseeking behavior reduces overall well-being as a zero-sum (or even negative-sum) game. Corruption often relies on strong personal connec­ tions and extortion practices by mafias may operate through personal connections. In this chapter1 we aim at formalizing and testing the double-sided role of social networks on growth. We model social capital as participa­ tion in two types of social networks: first, closed networks of family and friends, and, second, open networks that bridge different communities. Agents have a preference for socializing, which they trade off against material well-being. Participation in both social networks is timeconsuming and comes at the cost of participation in the formal economic sphere and working time. Through this channel, higher levels of social capital may crowd out economic growth. In addition, participation in intercommunity networks reduces incentives for rent-seeking and chea­ ting. Through this channel, higher level of bridging social capital may enhance economic growth. Testing the model, we find that regional dif­ ferences in materialistic attitudes and the value attached to family life significantly reduce the participation in open networks and that this in turn reduces regional output growth in Europe. The two types of social networks we distinguish correspond to Putnam’s (2000) concepts. He defines ‘bridging social capital’ as bonds of connectedness that are formed across diverse social groups, whereas ‘bonding social capital’ cements only homogeneous groups. The added value of this contribution lies in the formal macroeconomic modeling and the empirical testing of the influence of different types of social capital on economic growth. We show that bridging social capital has a positive effect on growth, whereas bonding social capital has a negative effect on the degree of sociability outside the closed social circle. We find evidence for Fukuyama’s claim that ‘the strength of the family bond implies a certain weakness in ties between individuals not related to one another’ (Fukuyama, 1995: 56). Moreover, we show that an important mechanism that influences the degree to which people are willing to step out of their closed social circle with the associated advantages and build bridging social capital depends on the materialistic attitude of that peo­ ______________ 1

The authors want to thank Jeroen van de Ven and the members of the EVS working group for useful comments on an earlier draft.

BRIDGING AND BONDING SOCIAL CAPITAL

149

ple. People who are more materialistic tend to stick to the type of sociali­ zing that has a direct payoff, whereas less materialistic people are more embedded in social structures that do not directly yield materialistic or worldly advantages. The rest of the chapter is structured as follows. We first review the main ideas in the literature on social capital and materialistic attitudes. Then we present our economic model in which we show the different channels through which the different types of social capital affect eco­ nomic growth and what the role of materialistic attitude is. Readers not interested in the details of economic theoretic modeling can move di­ rectly to Section 4 in which we summarize in a less formal way the test­ able implications of the model and formulate the central hypothesis to be tested. In section 5 we describe the data. Section 6 presents our empirical estimations for 54 European regions using the European Values Study (EVS) surveys. We conclude with a discussion on our findings and sug­ gest future research questions. 2 Background Social capital is the key theme of our chapter. There is not a single uni­ fied or generally accepted theory of social capital. The field of social capital ranges across the whole social sciences, from economics, orga­ nizational sociology to political science. As Fine (2001: 78) states: ‘So­ cial capital provides a technological umbrella for grouping together an extraordinarily diverse range of casually constructed illustrations’. An important reason for the fuzziness of the concept is caused by the fact that researchers from different disciplines use social capital for what at first sight seem to be entirely different objects of study. When organiza­ tion scholars discuss social capital they think of it in terms of the net­ work a firm is embedded in and the resources and limitations this net­ work may provide (e.g., Burt, 1992; Coleman, 1988; Gulati, 1999). When macro-economists and political scientists use social capital, they also think of it in terms of networks, but then referring to networks of associational activity, which is not the same as the previous type of net­ works (e.g., Putnam, 1993, 2000; Knack & Keefer, 1997). For a better understanding of the theoretical concept of social capital and its cause and effect structure, it is necessary to break down the concept of social capital in two levels, i.e., at the level of the firm (micro) and of the na­ tion state (aggregate level); see Beugelsdijk (2002). At the individual level, social capital refers to the network an individ­ ual belongs to. Individuals derive benefits from knowing others with whom they form networks of interconnected agents. The network en­

150

CHAPTER SIX

hances access to and exchange of information, enforcement of contracts, and focusing on a shared vision and collective goals (Nahapiet & Goshal, 1998). At the aggregate level, it is argued that nations or regions can hold different levels of social capital, which affects the level of democracy and economic growth (Fukuyama, 1995; Putnam, 1993, 2000; Beugels­ dijk & Van Schaik, chapter five in this volume). Social capital at this level refers to the social structure that enhances the effectiveness of local governments through traditions of civic engagement and the structure of civic networks. At both levels the effects of social capital can be positive and nega­ tive2. At the micro-level, dense networks may provide useful resources such as improved quality of information, a means for control, influence and power, and also a closed social network may encourage compliance with local—sometimes implicit—rules and customs and reduce the need for formal monitoring. However, the danger of closed social networks lies in the fact that the relation-specific capital that is developed over time may lead to a tendency to stick to existing linkages and networks start to suffocate. This may result in a loss of flexibility and lock-in. At the aggregate level, the effects of social capital are empirically harder to prove and less clear. Although Putnam (1993) claims to have proven that more social capital in Italian regions is positively correlated with effective governance and economic performance, he has been espe­ cially criticized for the method and the lack of a theoretical mechanism between social capital and the other ‘dependent’ variables (Jackman & Miller, 1996; Tarrow, 1996; Dekker et al., 1997; Harris & De Renzio, 1997; Paxton, 1999; Boggs, 2001). Boix and Posner attempt to describe mechanisms through which social capital is translated into better eco­ nomic performance. They argue, among other things, that social capital may reduce the probability of individuals to engage in opportunistic be­ havior. This saves on resources devoted to monitoring agent’s perfor­ mance and makes more resources available for more productive invest­ ments3. In a later work, Putnam (2000) has made a distinction between ‘bridging social capital’ in which bonds of connectedness are formed across diverse social groups, and ‘bonding social capital’ that cements ______________ 2

For a more extensive and in depth discussion of the cause and effect structure of so­ cial capital at the aggregate and the individual (firm) level we refer to Beugelsdijk (2002). 3 Boix and Posner also mention that a) social capital contributes to effective govern­ ance by facilitating the articulation of citizen’s demands, b) social capital reduces the need to secure compliance by creating complex and costly mechanisms of enforcement and reduces transaction costs in the arena of citizen-government relations and c) social capital encourages the articulation of collective demands that are to everyone’s benefit.

BRIDGING AND BONDING SOCIAL CAPITAL

151

only homogeneous groups. Bonding social capital has negative effects for society as a whole, but may have positive effects for the members be­ longing to this closed social group or network. Bridging social capital, hence, making contacts between different groups or networks is positive. At the micro level this is related to Burt’s theory of structural holes, where the optimal position for an individual is between several groups. The literature on social capital has mainly focused on what consti­ tutes social capital, on the differences in its structure, and the conse­ quences, rather than on explanations where social capital comes from. Since social capital is formed through network participation and social interaction in groups, it may well arise as a by-product of social interac­ tion that is initiated mainly for other reasons. As argued above, man simply has a desire for socializing, just like it has a preference for food, shelter and material possessions. Our argument is that there may be a trade-off between satisfying materialistic wants and desires for sociali­ zing. Materialistic attitudes may thus come at the cost of socializing and reduce the accumulation of social capital. Materialism, materialistic attitudes and acquisitive desires are studied in the marketing literature and studies on (business) ethics and economic psychology. Belk (1984, 1985) defines materialism as the importance that possessions play in an individual’s life or the importance one at­ taches to their worldly possessions. Materialism is seen as a personal trait measured along the dimensions of envy, possessiveness and nongenerosity. We are more interested in materialism as a value. Richins and Dawson (1992) approach materialism as beliefs on the value of ma­ terial objects. They measure it along three dimensions: how central is acquisition, how much is it used as the pursuit of happiness, how impor­ tant is possession-defined success. The study by Inglehart (1997) comes closest to our approach to materialism: he is also interested in the con­ nection between economic development and materialism. In his view, high levels of development correlate with postmaterialism, in which ma­ terial consumption becomes less important relative to the consumption of services and civic liberties. In our view, the degree of materialism af­ fects economic development through its effect on social capital. We try to bring together some aspects of social capital at the micro level and the aggregate level and establish links with materialism. In our theoretical model, individuals endogenously choose how much time they spend on closed networks and open networks, depending on their prefer­ ences and the opportunity costs. Both networks provide opportunities for social interaction, for which individuals have a preference. Participation in open networks has the side-effect of protection against opportunistic behavior by others. Each individual also optimally chooses time spent on

152

CHAPTER SIX

rent-seeking activities, on work and on investment and learning. At the aggregate level, participation in open networks (i.e., bridging capital) translates in civic engagement. If the level of civic engagement is high in society, opportunistic behavior becomes less attractive for individuals and a more efficient system of exchange stimulates the economy. We formally link these mechanisms to investment and economic growth and show that more bridging social capital may (but need not) go together with faster growth. The reason why bridging capital is not necessarily good for growth is that it requires the maintenance of networks, which is a time-consuming process and comes at the cost if working time4. Our empirical model follows closely the structure of the theoretical model. Network participation is an endogenous variable so that the ef­ fect of social capital, formed through network participation, on growth requires a careful way of testing. In particular, we need to find relevant exogenous factors that determine simultaneously the level of social capi­ tal and economic growth. In accordance with the model, we use materialistic attitudes as an instrument: a preference for materialistic as­ pects of life relative to the social aspects of life directly affects network participation, and it affects growth only indirectly through network participation. The European Value Study provides the data on materialistic attitudes and social capital. Our results show that mate­ rialism can indeed explain the level of bridging social capital and that bridging social capital is positively correlated with economic growth. 3 The model

3.1 Individuals’ static decision problem Individuals care about produced consumption goods (c) and social inter­ action5 (s). That is, their utility function has both material goods and so­ cial aspects as arguments: ______________ 4

It is important to note that for atomistic agents any form of social interaction—be it either bridging or bonding social capital—yields benefits. The issue is that bridging social capital has a larger (positive) impact on economic growth than bonding social capital. Hence, we do not claim that socializing with family and close friends is a bad activity as such. The crucial point is the distinction between types of socializing; investing in bridg­ ing social capital is better from a growth perspective. In this respect Putnam (2000) makes a relevant distinction between ‘getting by’ (bonding social capital) and ‘getting ahead’ (bridging social capital). 5 From now on, when we write social interaction we mean socializing or sociability. In general, social interaction can also imply the fighting of a war, whereas socializing im­ plies informal friendly social interaction. Nevertheless, in the remainder we restrict social interaction to the process of socializing.

153

BRIDGING AND BONDING SOCIAL CAPITAL

U c ,U s > 0 ,

u = U (c , s ) ,

where subscripts to function symbols denote (partial) derivatives. Social interaction is defined as participating in social networks, so that higher levels of network participation can be labeled as higher levels of social capital. We distinguish two types of networks. First, social in­ teraction takes place with close friends and family (which we categorize as f-networks). Second, networks consist of more remote contacts out­ side the family, within and outside the community one lives, in clubs, pubs and public meeting places, in voluntary organizations (called vnetworks). In Putnam’s (2000) terms, f-networks and v-networks repre­ sent bonding and bridging capital, respectively. Interacting with others is possible in both of these networks, so that they are substitutes to a cer­ tain degree in satisfying the individual’s preference for social interac­ tion. As Fukuyama argues ‘People are embedded in a variety of social groups—families, neighborhoods, networks, businesses churches, and nations—against whose interests they have to balance their own’ (Fuku­ yama, 1995: 21). However, each network type has its own specific type of social interaction: among friends and family feelings of affection and safety can be nurtured; among more remote contacts other interests may be pursued like self-realization, social status seeking, information ex­ change, adventurous contacts with less known ideas and cultures. Hence, on balance the two are imperfect substitutes in the utility function, which is reflected in the sub-utility function for satisfaction from social interac­ tion (s):

S f , Sv > 0

s = S ( f , v) ,

.

Here, f (v) is the intensity of participation in f-networks (v-networks), to be measured by the time devoted to it. A convenient specification of the (sub)utility function is a constant elasticity of substitution (CES) func­ tion, in which two important parameters play a role: one indicating the relative weight of the arguments, and the other indicating how easily the two can be substituted for each other. The specifications are:

(

U (c, s ) = µ1/ σcs c (σcs −1) / σcs + s (σcs −1) / σcs

(

1/ σvf

S ( f , v) = φ

f

( σvf −1) / σvf

+v

)

σcs /(σcs −1)

( σvf −1) / σvf

)

,

σvf /( σvf −1)

.

The relative importance of material consumption is denoted by and will be referred to as the materialism preference parameter. The impor­

154

CHAPTER SIX

tance of f-networks relative to v-networks is denoted by φ and will be re­ ferred to as the family ties preference parameter. The elasticities of sub­ stitution between the two types of social networks is denoted by vf, that between material consumption and social interaction by cs. Individuals choose how much they consume and how much they en­ gage in social interaction. Their choices are constrained by a crucial time (or budget) constraint. Consumption is constrained by income, which is derived from working at a wage w and from transfers x according to the following budget constraint:

c = (n0 − f − v) w + x , where n0 is total time available for working and social interaction, f and v is time devoted to social interaction in f-networks and v-networks re­ spectively, w is the individual’s wage and x is a transfer (which may be negative). Social interaction (that is, maintaining social capital) requires time, which comes at the cost of working time. Thus there is a trade-off between social interaction and material consumption. Through this channel, social interaction crowds out economic activity, so that the so­ cial capital created by social interaction has a negative effect on the economy. To allow for a potential positive effect of social capital on the econ­ omy, we assume that social interaction in certain networks affects the degree of opportunistic behavior. The idea is that agents engage in (time­ consuming) rent-seeking activities, by which we mean corruption and extorting, shirking and distrusting. By doing so they can effectively ex­ tract some of the income of others. However, participation in open net­ works (v) protects agents against rent-seeking: people that are in the same open network never rob each other. As an example consider a shopkeeper, who is left with some products that are below his normal quality standard (say a grocer with some vegetables not so fresh any­ more). He could mix high-quality goods and low-quality goods and sell all of them as high-quality goods. The customers might notice the low quality only when they are at home. But then they do not find it worth­ while to return to the shop and complain. The shopkeeper might also re­ frain from selling the low quality goods, or might sell them at a discount with the explicit warning about the quality. The shopkeeper can be ar­ gued to be less declined to cheat his customers, if he knows he will meet the customer at another occasion, in particular when socializing with the same person. This is because the customer might start to complain if product quality was low.

BRIDGING AND BONDING SOCIAL CAPITAL

155

The idea is that open networks act like bridging capital in connecting different groups. Within these networks, participants build up reputation and show trustworthiness in order to be able to derive value from social interaction across the groups represented in the network. Thus, while the desire for social interaction is the primary reason to join the network, trust and protection against rent-seeking is created as a side-product. To be precise, let z be the time devoted to preparing rent-seeking ac­ tivities. Rent-seeking implies randomly selecting a number of persons and extracting income from them. The most direct interpretation of in­ come extraction is simply theft or robbery. Many more indirect interpre­ tations are possible also: shopkeepers may exploit uninformed customers by selling goods of inferior quality; workers may cheat employers by shirking; one might think of opportunistic behavior in general. You can avoid cheating family or friends. Naturally, we assume rent-seeking is directed at persons outside your own community (f-network). However, you may run the risk that some of your rent-seeking activity affects fel­ lows from your v-network, something you can only avoid after you have already spent the time preparing the rent-seeking activities (that is after choosing z). We also assume you actually want to avoid damage to fel­ lows from your v-networks, because they can exploit the network for some punishment or ostracism strategy. On average a fraction B ( z ) of the average wage of a person you target can be extracted. If all these possible gains would be taken, the expected benefit of rent-seeking would be B( z ) w , where w denotes the average wage. However, only persons with whom one has no ties through vnetworks will be eventually robbed, so the actual benefit is (1 − v) B( z ) w . Rents extracted from others amount to positive transfers to you. However, rents being extracted from you amount to negative transfers. The latter are higher, the higher is your wage, the more rent-seeking pre­ vails in society and the smaller is the number of fellows in your vnetwork who will abstain from robbing you. These (gross) negative transfers can be expressed as (1 − v) D( z ) w , where D( z ) is the damage from being robbed per unit of wage income, and z is the average level of rent-seeking activity in society. Thus, net transfers are:

x = (1 − v)[ B( z ) w − D( z ) w] ,

Bz > 0, Dz > 0 .

Since z is the time devoted to rent-seeking, it comes at the cost of time devoted to labor and social interaction. Normalizing the total time en­ dowment to unity, the following time constraint applies:

156

CHAPTER SIX

n0 = 1 − z . The complete decision problem of the individual agent can now be summarized as: maximize u = U (c, s ) , subject to s = S ( f , v) and

(1)

c = (1 − v − f − z ) w + (1 − v )[ B ( z ) w − D ( z ) w] . The individual takes as given its wage, the average wage and average rent-seeking level in society ( w, w, z respectively). The first-order condi­ tions for a maximum can be written as:

(1 − v) Bz ( z ) w = w , Sv ( f , v) w⎤ ⎡ = 1 − ⎢ D( z ) − B( z ) ⎥ , S f ( f , v) w⎦ ⎣ U c (c, S ( f , v)) S f ( f , v) . = U s (c, S ( f , v)) w

(2)

(3)

(4)

Equation (2) is the condition for optimal rent-seeking: it states that the marginal benefits of rent-seeking (marginal expected gross transfers, left-hand side) should equal the marginal opportunity cost (the wage on foregone labor time, right-hand side). Equation (3) determines the opti­ mal trade-off between the two types of network interaction. The lefthand side represents the amount of time devoted to v-networks an indi­ vidual is maximally willing to give up in exchange for an additional unit of time devoted to v-network participation (marginal rate of substitu­ tion). The right-hand side gives the opportunity cost of engaging in fnetwork participation rather than in f-network participation (marginal rate of transformation). Spending time with friends has a relatively low cost compared to spending time in extra-community networks if the net loss from rent-seeking (term in brackets) is high. Equation (4) deter­ mines the optimal trade-off between the material consumption and social interaction in f-networks.

BRIDGING AND BONDING SOCIAL CAPITAL

157

3.2 Static equilibrium under symmetry The decisions of the individual agent depend on the society-wide vari­ ables like average rent-seeking, which in turn depend on the decisions of others. To solve for the macro-economic levels of the variables, we em­ ploy the simple assumption of complete symmetry: all agents have the same preferences and income and will make the same choices. Hence we have:

z = z, w = w .

(5)

We can now also link the benefits of rent-seeking to the losses. We as­ sume that if all agents engage in the same intensity of rent-seeking, the losses are a constant factor 1 + ζ larger than the benefits:

D ( z ) = (1 + ζ )B ( z ) ,

ζ > 0.

(6)

Thus rent-seeking is a negative sum game: what the extorter gains, is less than the damage to the person being extorted. Part of the transfer may be lost ‘in the battle’ or confiscated by authorities. One might also see this as an implicit way of modeling the costs that the victim has to incur to avoid cheating and shirking (monitoring costs). Parameter ζ captures this externality cost of rent-seeking6. Our main question in this subsection is how economic activity (c) and bridging social capital (v) are related. Note that both variables are en­ dogenous. Therefore, we need to identify how variations in exogenous variables simultaneously affect economic performance and social capi­ tal. The exogenous driving forces in the model are labor productivity (w), preference for family and friends ties (φ), and preference for mate­ rial consumption (materialism ). We reduce the model to two equations in terms of the endogenous variables c and v and the exogenous vari­ ables w,φ� . First note from (2) and (5), that z is a negative function of v:

z = Z (v ) ,

Zv < 0

(7)

Next, substitute this result and (6) into (3) to find that f is a positive function of v and φ:

f = F (v; φ) , ______________ 6

Fv , Fφ > 0 .

(8)

We think that it is realistic to add this negative externality. However, all our qualita­ tive results go through when ζ = 0.

158

CHAPTER SIX

Substituting these results into the budget constraint, we find:

c = w[1 − v − F (v; φ) − Z (v) − (1 − v)ζB( Z (v))] ≡ T (v; φ, w) ,

Tφ < 0, Tw > 0 .

(9)

This is a key result. It reveals that networks have an impact on the econ­ omy through five channels (corresponding to the five places where v shows up in the equation). First, more social interaction in v-networks directly reduces labor time and hence reduces output (see second term in brackets). Second, different types of social networking are positively correlated, so an increase in v-networking also increases time spent with friends and family and further reduces working time (third term in brackets). Together, we call these effects the labor time crowding out ef­ fect. The other three effects stem from the fact that v-capital protects against rent-seeking. In more dense social networks, there is less rentseeking, so that not only time is freed up for production (fourth term in brackets, recall that Z depends negatively on v), but also the negative sum externality is smaller (through lower probability that non-members meet and rob each other and through the smaller rent-seeking effort z). Whether economic activity is positively or negatively related with vnetworks depends on whether the negative labor time crowding out ef­ fect dominates or not the positive protection against rent-seeking effect. Equation (9) also reveals that materialism(µ) has no direct impact on the economy, but can have an indirect impact only through affecting v. Indeed, from (4), we find (after substituting the solutions for f and z) an­ other equation in c and v, which depends on all key exogenous variables, including µ:

c = C (v; φ, µ, w) ,

Cv , Cµ , Cw > 0 .

(10)

In the appendix we derive a more precise solution by linearizing the model. We can prove that C increases in v, µ and w, but that the impact of φ cannot be unambiguously signed. This relationship shows that con­ sumption and v-networks are positively related. The reason is simply that social interaction and material consumption goods are normal goods: richer persons spend more on both. As expected, more materialis­ tic preferences (higher µ) or higher income (w) result in higher con­ sumption for given v. A stronger preference for family ties (higher φ) has two opposite effects: it shifts attention away from material consumption (substitution effect), but it also implies that a given level of interaction

BRIDGING AND BONDING SOCIAL CAPITAL

159

with family and friends generates more utility from social interaction (cf. income effect). The latter effect makes material consumption scarcer relative to social interaction (s) and tends to raise c. The two equations (9) and (10) simultaneously solve for the two en­ dogenous variables, material consumption (c) and bridging capital (v). The two equations can be represented as the lines labeled T and C, re­ spectively, in a simple diagram in the v,c plane (we draw lines instead of curves to stress that results are based on comparative statics, see appen­ dix). The slope of the T-line is ambiguous because of the opposing labor time effect and protection effect. Figure 6.1a and 6.1b represent the two possibilities. We illustrate the working of our model by showing the effects of an increase in the materialism preference parameter (µ), which is a key de­ terminant in our analysis. More materialistic attitudes make the C-line shift to the left. The point of equilibrium moves along the T-line. In Fi­ gure 6.1a, the slope of T is negative since the labor time crowding out effect dominates; then consumption rises and bridging social capital falls. c C(v; µ, φ, w)

T(v; φ, w)

v Labor time crowding out dominates (ε Tv < 0 < εCv)

Figure 6.1a Semi-reduced form of the model (broken lines reflect increase in materialism)

In Figure 6.1b, T slopes upward since the protection effect dominates; then both consumption and social capital fall. Hence, materialism affects the economy (as measured by a change in c) through a change in volun­

160

CHAPTER SIX

tary organizations (a movement along the T-curve), but whether it boosts or hurt the economy depends on the relative strength of the crowding-out effect and protection-against-rent-seeking effect. c

C(v; µ, φ, w)

T(v; φ, w)

v Protection against rent seeking dominates (0 < ε Tv < εCv)

Figure 6.1b Semi-reduced form of the model (broken lines reflect increase in materialism)

Analyzing changes in the other parameters in a similar way (and resor­ ting to the mathematics in appendix A to make results more precise), we find the comparative statics displayed in Table 6.1. Table 6.1

Comparative statics, model without investment* Materialism (µ) c v

Labor time crowding

out dominates (εTv < 0 < εCv) **

Protection against rent-

seeking dominates (0 < εTv < εCv)

Family ties (φ)*** c v

Labor productivity (w) σcs < 1 [σcs = 1] (σcs > 1) c v

+

–

?

–

?[+](+)

+[0](–)

–

–

–

–

+[+](?)

+[0](–)

* We assume εCv – εTv > 0 , which implies that (10) has a steeper slope than (9) in the v,c plane. See appendix A for exact solutions. ** εJi denotes the elasticity of function J with respect to variable i. See appendix A. *** The signs in this column are derived for the assumption: σvf > σcs, σvf > 1.

BRIDGING AND BONDING SOCIAL CAPITAL

161

For example, a rise in the family ties preference parameter (φ) reduces bridging social capital (as measured by v, see the minus signs in the fourth column), and is likely to reduce economic activity (see fourth col­ umn). 3.3

A dynamic version of the model

So far we have not directly studied economic growth. We have consi­ dered only the decision problem at a given moment in time. This section gives a simple extension of the model that allows us to study economic growth. First, we model firm behavior so that we can explain wages and national production levels. In particular, both variables depend on skills (or human capital) per worker and the technology level (or total factor productivity). Second, we model investment and technical change, so that we can explain the growth of production over time. In particular, we allow for investment in human capital and physical capital. Technologi­ cal change arises from development of new technologies and absorption of foreign technologies. The key finding is that the returns to investment are related to the level of social capital (network participation) in a way that is comparable to the relationship between the level of consumption and social capital that was found above in the static version of the model. Through the la­ bor time crowding out effect more network participation reduces the re­ turns to investment. In contrast, the protection against rent-seeking effect creates a positive relationship between social capital and investment re­ turns. Since higher returns to investment stimulate investment and growth in equilibrium, we infer from this that the relationship between growth and social capital is ambiguous: it is positive if the protection ef­ fect dominates. Although our model relies on a specific way of modeling growth through technical change and human capital accumulation (Bils & Klenow, 2000, who build on Nelson & Phelps, 1966; and Mincer, 1974), the results carry over to other standard growth frameworks (e.g., Lucas, 1988; Rebelo, 1991). The time crowding out effect implies lower utiliza­ tion of any kind of capital (or input) in the economic production process (for example, physical capital, human capital). This reduces the incen­ tives to invest in these capital goods (or inputs). Protection against rentseeking implies that the returns from investment can be better appropri­ ated and thus stimulates investment and growth.

162

CHAPTER SIX

Human capital investment We extend the individual’s choice problem for investment in human capital through learning, as well as investment in financial wealth through saving part of wage earnings. The productivity of an individual depends on its human capital, which can be considered as its productive skills, to be denoted by h. Earnings are proportional to this human capi­ tal skill level: doubling h implies doubling earnings. Let wh be the wage of a person with skill level h = 1, nw the time devoted to working. Then earnings of an individual with human capital h are nwhwh. (Note that the wage per unit of time thus equals whh = w, where w was used in the notation above). By devoting l units of time to learning and training activities, the agent acquires a skill level h = eψl (cf. Mincer, 1974; Bils & Klenow, 2000). Learning time is chosen endogenously to maximize utility. From the complete decision problem (see appendix), the following decision rule for learning can be derived:

ψhwh [(1 − v)(1 − D( z )) − f − z − l] = w .

(11)

The left-hand side of this first-order condition represents the marginal benefits of learning: it raises human capital by ψh units, which increases earnings at wage rate wh for each unit of time that human capital is em­ ployed (the term in brackets represents this working time). The righthand side represents the marginal costs of learning, which consist of wages forgone because learning time comes at the cost of working time. The equation can be solved for the time spent on learning:

l = [(1 − v)(1 − D ( z )) − f − z] −1/ ψ . Substituting (5)-(8), we find that in a symmetric equilibrium investment in human capital can be written as a function of v and exogenous varia­ bles:

l = L (v; φ, ψ ) ,

Lφ < 0, Lψ > 0 .

(12)

Similar as in equation (9), five channels can be distinguished through which v-networks affect learning. Also similarly, the elasticity of learn­ ing time with respect to v-networks cannot be unambiguously signed. The labor time crowding out effect implies that more time spent on so­ cial interaction reduces labor time and therefore reduces the incentives to invest in productive skills through learning. The protection against rentseeking effect implies that more (bridging) social capital reduces rent­

163

BRIDGING AND BONDING SOCIAL CAPITAL

seeking, reduces time spent on rent-seeking and reduces the loss of in­ come through rent-seeking, which raise the returns to investment in skills. These two effects are of opposite sign. As in the static model, a budget constraint and the demand for pro­ duced consumption goods complete the model. The model can thus be summarized by three equations in three variables (c, v and l), and all en­ dogenous variables can be solved in terms of the exogenous parameters. We restrict the analysis to the special case in which cs = 1. Then, for example, the solution for bonding social capital can be written as:

Vφ , Vµ ,Vψ < 0 .

v = V (φ, µ, ψ) ,

(13)

The assumption cs = 1 makes the solutions for f, v, and l independent of wh (cf. Table 6.1, bracketed entries). The reason is that income effects (higher wages raise income and raise demand for social interaction) off­ set the substitution effects (higher wages raise the opportunity costs of time spend not working). Table 6.2 summarizes the comparative statics, which are formally de­ rived in the appendix. The key result is that investment is either posi­ tively or negatively affected by materialism, depending on whether the crowding out or the protection against rent-seeking effect dominates. As anticipated before, the results for l in the model with human capital are very similar to the results for c in the model above. Table 6.2

Comparative statics, model with human capital Materialism (µ)

Family ties (φ)

*

Investment productivity (ψ) l v

l

v

l

v

Labor time crowding out dominates (εLv < 0)

+

–

?

–

+

–

Protection against rent-seeking dominates (0 < εLv)

–

–

–

–

+

–

* We assume σcs = 1, εCv > εTv. See appendix B for exact solutions.

Firms and regional output growth Regional production (Y) is a function of physical capital (K), effective labor input (H), and technology (A):

Y = AK β H 1−β .

(14)

164

CHAPTER SIX

Firms produce their output according to this production function. They hire labor and capital up to the point where the marginal products equal the wage (wh) and interest rate (r), respectively.

AβK β−1H 1−β = r ,

(15)

A(1 − β)K β H −β = wh .

(16)

We assume that physical capital is regionally mobile so that the supply of capital is perfectly elastic at the exogenously given international in­ terest rate. Solving for K in (15), and substituting into (14) and (16), we find that output and wages can be expressed as:

Y = (β / r)β /(1−β ) A1/(1−β ) H ,

(17)

wh = (β / r)β /(1−β ) (1 − β) A1/(1−β ) .

(18)

Hence, increases in the technology level (A) and effective labor input (H) drive growth in output and wages. The effective labor supply equals the labor time corrected for their skill level (or human capital) h. As above, we assume symmetry among agents (all have the same level h). Each agent supplies (1 – v – f – z)h units of (effective) labor at the wage wh, but (1- (z))h units ultimately do not result in regional output, because they get lost in the process of rent-seeking7. Aggregate effective labor input is therefore (where N is the number of agents, or population size):

H = N ⋅ [(1 − v)(1 − ζB) − f − z] ⋅ h .

(19)

The level of technology a country can exploit is the result of technologi­ cal innovation and absorption of foreign technologies. As in Nelson and Phelps (1966, cf. Bils & Klenow, 2000), a larger level of human capital per worker facilitates the absorption of foreign technologies. We choose the following specification for the growth rate of technology:

g A = α ln( A / A) + λ ln h , ______________ 7

(20)

We do not subtract learning time. Thus we integrate the learning sector (education, training and consultancy) in our measure of output Y.

BRIDGING AND BONDING SOCIAL CAPITAL

165

where A is the world technology set a region can learn from. Note that this equation captures convergence through catching-up. As long as the technology gap with other regions, A / A , is relatively large, the region has a large pool of technologies to absorb from other regions. As a result it realizes relatively fast technical change. That in itself, however, im­ plies that the technology gap A / A falls over time and technological change slows down. The other way around, countries with high techno­ logy levels relative to other regions experience relatively slow technical change (for similar levels of human capital). Hence, over time, the rates of technical change in the regions tend to converge. Growth of per capita output can now be calculated as:

⎡ ⎛ λ ⎞ ⎤ g y = α ln(Y / H ) − α ln y + ⎢α ln(1 − v − f − z) + ⎜ α + ⎟ ψl ⎥ 1− β ⎠ ⎦ ⎝ ⎣ dl d [(1 − v)(1 − ζB) − f − z] / dt , (21) +ψ + dt [(1 − v)(1 − ζB) − f − z] where per capita output is denoted by y ≡ Y / N , where we have used (17) to eliminate A / A and Y / H is the average income per unit of human capital in rest of the world. In our model we can ignore the last two terms if (due to the assumption σcs = 1) l, v, f, and z are constant over time. In terms of testing the model, these terms are expected to be rela­ tively small. Moreover, no time series data is available for these vari­ ables. We are then left with three relevant terms that explain growth: the foreign income level, own income level, and the term in brackets, which can be written in terms of v and the parameters φ and only (see (12), (8), (7), (6)).

– The first term at the right-hand side of (21) captures spillover effect: rich neighboring regions provide a region with the opportunities to learn from and grow faster. – The second term at the right-hand side of (21) captures betaconvergence. Poor countries grow faster than rich countries, ceteris paribus, due to the technological catch-up effect just described. – The third term at the right-hand side of (21) captures the effect of so­ cial capital on growth. Note that the sign is ambiguous because the labor time allocation effect may or may not be dominated by the pro­ tection against rent-seeking effect. Also the effect of is ambiguous: on the one hand a higher productivity of learning enhances human capital, on the other hand it reduces hours worked.

166

CHAPTER SIX

Of course, v is an endogenous variable, but its solution is already given in (13): materialistic attitudes, investment opportunities and family ties preferences affect the level of bridging social capital. Interesting to note is that materialism may be good or bad for growth. In particular, if the protection against rent-seeking effect dominates, more materialism leads to lower bridging capital and thus to lower growth. 4 The hypotheses In the theoretical model, the following results have been derived about the relationship between growth and social interaction:

– Growth and bridging social capital are endogenous variables, which are simultaneously determined by attitudes towards spending time with friends and family, materialism, and the productivity of invest­ ment. – Controlling for family ties, initial income, and productivity of in­ vestment, an exogenous increase in bridging capital may affect growth negatively or positively. In the former case, the time cost of networking dominates the productive benefits. The latter case arises if the protection of bridging capital against rent-seeking is strong enough (see equations (21) and (12)). – Materialism affects growth only through bridging social capital. – Family ties, investment and materialism negatively affect bridging capital. Initial income does not affect bridging capital (see (13)). Figure 6.2 summarizes the model predictions. Arrows with plus (minus) sign denote positive (negative) relationships between two variables. In the next section we explain the background of the data and test the above formulated hypotheses. 5 Measurement To empirically test the hypotheses we use data from the 1990 wave of the European Values Study. Our economic data comes from Eurostat. In order to calculate regional scores for our ‘cultural’ data that measure so­ cial capital and materialism we regionalized the scores of individual res­ pondents in the 1990 wave. We did not use the first 1981 wave as we could not trace the regional background of the respondents. Figure 5.1 and Table 5.1 in this volume provide an overview of the regions in­ cluded in our analysis.

BRIDGING AND BONDING SOCIAL CAPITAL

?

Economic growth (gy)

–

Bridging – social capital (v)

167

–

Materialism preference parameter ( )

–

Family and friends preference parameter (φ)

–

Investment ( )

– and friends Family preference parameter (φ)

Standard economic variables: + – convergence (ln Y/H – ln y) – investment ( ) Figure 6.2 Graphical representation of the theoretical model

Our theoretical model and its implications summarized above closely guide our empirical model. We can distinguish two main features of the empirical model, the modeling of growth and the modeling of social atti­ tudes and interaction. In this section we first discuss how we measure economic growth and then how we measure social variables. 5.1 Economic growth To test the growth part of our theoretical model, we use the standard growth framework, in which economic growth is explained by a number of key economic variables (Baumol, 1986; Barro, 1991; Mankiw et. al., 1992; Barro & Sala-i-Martin, 1995). These type of empirical growth re­ gressions typically include initial level of income, and proxies for human and physical capital (mostly the school enrolment ratio and the invest­ ment ratio). These empirical growth models are also referred to as Barro­ regression (after Barro, 1991). Similar to Barro and Sala-i-Martin (1995), we have computed the re­ gional growth figures by relating the regional GDP per capita informa­

168

CHAPTER SIX

tion to the country mean8. There are at least two reasons to use the coun­ try mean as a correction factor. First of all we do not have regional price data. Second, the figures on regional GDP are provided in an index form that is not comparable across countries. In addition one could argue that by measuring regional growth this way we directly control for national growth rates that may bias the regional growth rates. Hence, we have used Gross Regional Product (GRP) figures that are expressed as devia­ tions from the means from the respective countries. The 1950 data is based on Molle, Van Holst and Smit (1980), whereas the data for Spain refers to 1955 and are based on Barro and Sala-i-Martin’s (1995) calcu­ lations. Just as the other economic data, the 1998 data on GRP is drawn from Eurostat information. Following standard empirical growth models as developed by Barro (1991) we include a measure that controls for the level of initial income and measures convergence. In addition we include a measure to control for the level of welfare of neighboring regions, as is common in regional growth analyses. Low initial income and large spillovers from other re­ gions may stimulate growth by the convergence measure. Initial per ca­ pita income of the region is included, measured relative to the income of the other regions in the country (cf. Barro & Sala-i-Martin 1995; see Beugelsdijk & Van Schaik, chapter five in this volume, for more de­ tails). Spillovers are measured as the average income of the regions ad­ jacent to the region. Next to initial income, ‘Barro’ regressions typically include measures for human and physical capital. Our proxies for the productivity of in­ vestment are educational attainment, national investment rates, and in addition we use a measure for the concentration of human capital in ag­ glomerations (created by the interaction of a dummy variable indicating the major agglomerations in a country and the school enrolment ratio)9. Regions in which large agglomerations are present may benefit from scale economics, concentration of human capital, the presence of a clus­ ter of specialized suppliers, and a market with a critical mass of consum­ ers (network externalities). Further, the idea is that years of schooling facilitate learning on the job (which was theoretically modeled by vari­ ables l and h). Since regional investment rates are not available, we take the national rates10. Data comes from Eurostat. Schooling is measured in ______________ 8

Gross Regional Product of a region in 1950 is divided by the mean of the Gross Re­ gional Products of all regions belonging to a certain country. A similar formula is applied to calculate the 1998 relative regional product. Regional growth over the period 1950­ 1998 is then based on these two indices. 9 For details see Beugelsdijk and Van Schaik (chapter five in this volume). 10 Eurostat does provide data on Gross Fixed Capital Formation, but these data are in­ complete for some countries.

BRIDGING AND BONDING SOCIAL CAPITAL

169

1977, whereas investment data is based on Penn World Tables 5.6 and are the average investment ratio in the period of analysis. 5.2 Bridging social capital To operationalize bridging social capital we follow Knack and Keefer (1997) and Beugelsdijk and Van Schaik (this volume) by exploiting data on membership of certain voluntary associations. We measure the aver­ age number of groups cited per respondent in each region. However, in­ stead of measuring membership in general as done by Beugelsdijk and Van Schaik, we selected a number of associations that resemble the Put­ nam groups as described and analyzed by Knack and Keefer (1997). We measure bridging social capital by the density of associational activity, or in other words the average per capita membership of an association. We have then taken the logarithm. Of the associations mentioned in EVS we have used membership of the following groups: a. b. c. d. e.

Religious or church organizations Education, arts, music, cultural activities Youth work (e.g., scouts, guides, youth clubs) Sports or recreation Women’s groups.

The groups mentioned under a, b and c were also used by Knack and Keefer (1997) in their analysis of the Putnam groups and the relation with economic growth. We have chosen to add d and e as they also proxy associational activity that is not focused on rent-seeking activities that can be expected from groups such as political parties and profes­ sional associations. We expect the selected groups to involve social in­ teraction that builds trust and cooperative habits, which is the reason why we label it bridging social capital. The average score of the density of group membership in 54 European regions equals .34 with a standard deviation of .18. The highest score (.80) is obtained in the eastern part of the Netherlands (Oost-Nederland), and the lowest score (.08) in the North-Eastern part of Spain (Noroeste). All data is based on 1990 information. 5.3 Bonding social capital and family ties We measure preferences for family ties (preference parameter φ in the model) by EVS data on the relative importance of the closed social cir­

170

CHAPTER SIX

cle11. On a scale of 1-4 (very important - not at all important) respon­ dents are asked to indicate the importance in their life of family, and friends and acquaintances. By using factor analysis we re-scaled the two items in one dimension reflecting bonding social capital. Both on the in­ dividual and the regional level the chosen items converge into one di­ mension. The average value of bonding social capital in European re­ gions is –.077. The regions where people attach the highest value to the close social circle can be found in the southern part of Europe. The re­ gion with the highest score on bonding social capital is the French Medi­ terranean (.23) and the region where people attach least importance to family and friends is the German region Bremen (–.46). 5.4 Materialism To operationalize the degree of materialistic attitude towards society we use two proxies. First we use the well-known materialism-postmaterialism that Inglehart (1997, 2000) introduced. It is based on the relative importance respondents attach to the following items: a. b. c. d.

Maintaining order in the nation Giving people more say in important government decisions Fighting rising prices Protecting freedom of speech.

Of each of these four statements respondents are asked to indicate the most important and the next most important statement. The materialist/postmaterialist value is created as follows. If the respondent’s first and second choices are both materialist items (i.e., maintaining order and fighting rising prices), the score is ‘1’. If the respondent’s first and sec­ ond choices are both postmaterialist items (i.e., giving people more say and protecting free speech), the score is ‘3’. If the two choices are any mixture of materialist and postmaterialist items, the score is ‘2’. In sum, a high score on this variable reflects a postmaterialistic attitude and a low score reflects a materialistic attitude. The mean score equals 2.04 with a maximum value of 2.29 in the region Berlin (Germany). The most materialistic according to Inglehart’s materialism index are the people in the Italian region Campania (1.68). ______________ 11

We have no measures of time spent in closed networks (bonding social capital). This means that we cannot test equation (8) of the model. In other words, we look at purely stated preference instead of revealed preference with respect to bonding social capital. Instead, for bridging capital we use a measure closer to a revealed preference in­ dicator (actual network participation).

BRIDGING AND BONDING SOCIAL CAPITAL

171

In addition to the operationalization of materialism based on Inglehart, we used a second proxy. EVS contains several questions on the im­ portance people attach to various aspects of a job. Based on the question ‘which of the following aspects of a job you personally think are impor­ tant?’ respondents are asked to indicate a number of aspects12. Among these aspects some refer to materialistic values (e.g., good pay) and oth­ ers to immaterialistic values (e.g., useful job for society). We selected the following items that reflect an immaterialistic attitude towards a job: a. Pleasant people to work with b. A useful job for society c. Meeting people. Using factor analysis we re-scaled these items into one dimension and aggregated the individual scores to mean scores for each of our 54 re­ gions. The variable is scaled from immaterialistic to materialistic. We choose to label this variable job-related materialism. Hence, high scores on the variable job related materialism reflect a materialistic attitude. The highest score (most materialistic) is obtained in the French region Sud-Ouest (.56). The lowest score can be found in the eastern part of the Netherlands (–.58). Table 6.3 presents descriptive statistics of the vari­ ables defined above and used in the empirical tests. Table 6.3

Descriptive statistics

Bonding social capital Bridging social capital Materialism-Postmaterialism (Inglehart) Job-related Materialism Investment Schooling Growth 1950-1998 Initial income

Mean –.077 .34 2.04 .059 24.25 .51 .029 –.003

Std. dev. .17 .18 .13 .28 3.74 .067 .33 .25

N = 54. Note that ‘Growth 1950-1998’ and ‘Initial income’ are measured relative to the national average for reasons described in the text. The relative high standard deviation for both variables therefore does not have the usual interpretation. Growth is measured in percentage points. Initial income, as well as schooling, investment and bridging social capital are measured in logarithms (see main text for details). ______________ 12

The total list of aspects respondents are asked to choose from is: good pay, pleasant people to work with, not too much pressure, good job security, good chances for promo­ tion, a job respected by people in general, good hours, an opportunity to use initiative, a useful job for society, generous holidays, meeting people, a job in which you feel you can achieve something, a responsible job, a job that is interesting, a job that meets one’s abili­ ties.

172

CHAPTER SIX

6 Testing the model Figure 6.3 depicts our testing strategy. The boxes correspond to the theo­ retical model in Figure 6.2, but the labels now refer to our data. For ex­ ample, our measure of growth is regional economic growth 1950-1998 and one of our measures for materialism is Inglehart’s index for materialism/postmaterialism.

Importance family and friends Group membership

Inglehart’s postmaterialism

Job-related materialism Regionaleconomic growth 1950-1998

Importance family and friends

Standard economic variables: – Initial income – Schooling – Investment – Spillover – Agglomeration

Figure 6.3 Graphical representation of the empirical model

Our aim is to test the model in Figure 6.3. In particular, we are interested in the sign of the relationship between growth and bridging capital. Here we have to take into account that bridging social capital and growth are simultaneously determined. To avoid a simultaneity bias, we need to in­ strument for bridging social capital. Hence we use a two-stage least squares (2SLS) testing strategy13. In the first stage, we instrument social ______________ 13

We have checked for a possible endogeneity bias by using a Hausman test. It is common to test whether it is necessary to use an instrumental variable and estimate a

BRIDGING AND BONDING SOCIAL CAPITAL

173

capital, by regressing our measure of bridging capital on our measures of materialism, family ties and investment productivity. Doing so, we test for the signs of the arrows in the North-East part of the figure (and of equation (13)). In the second stage, we use instrumented bridging capi­ tal, together with investment and convergence measures, as regressors for growth. Doing so we test for the signs of the left-hand side of the figure (and of equation (21) with (12) substituted). Needless to say, we are most interested in finding the empirically relevant sign of the relation between growth and bridging social capital which could not be deter­ mined a priori and was accordingly denoted by a question mark in Figu­ re 6.2. We have taken a standard growth framework that is similar to the one used by Beugelsdijk and Van Schaik, and Beugelsdijk and Noorder­ haven (see chapter four and five in this volume). We analyze 54 regions in the period 1950-1998. Our basic regression analysis includes initial level of production, school enrolment rate, investment ratio, spatial autocorrelation (spillover) and a variable that captures the concentration of human capital (agglomeration). We have taken log-specifications for the first three variables. The results are summarized in Table 6.4. We estimate different mod­ els. The first is our basic model in which our dependent variable is the average regional-economic growth of per capita income between 1950 and 1998. In addition to the basic model we estimate a number of other model specifications. The basic model in column (1) shows that bridging social capital has a positive and significant effect on regional growth. Bonding social capi­ tal has the negative sign predicted by our model, but is insignificant in the second stage. However, in the first stage, bonding social capital (or better, the preference for family ties) negatively affects bridging capital, in accordance with the model. Also materialism determines bridging capital with the correct sign and significant coefficient. The results on the effects of bridging capital on growth are worth being highlighted. Note that from the model we could not sign this effect unambiguously because of two opposing forces. Empirically, we find a positive effect, which means that bridging capital is good for growth. This positive ef­ fect is statistically significant, but quite small in economic terms. A one percent standard deviation in bridging capital raises growth by only 0.17*0.18 = 0.03 percentage points. A assessment of the economic sig­ ______________

2SLS regression, i.e., whether a set of estimates obtained by least squares is consistent or not. We performed an augmented regression and concluded that estimating an OLS would not yield consistent estimates.

174

CHAPTER SIX

nificance of the result that is more consistent with our estimation proce­ dure yields a bigger number: a one standard deviation change in our three instruments (family ties and two types of materialism) raises growth through bridging capital by 0.11 percentage points. Over our 48 years sample period this amounts to the non-negligible increase of 5.4% in (last year’s) regional income. The social capital variables in the basic model perform even better than the traditional variables like schooling and investment, of which the coefficient is insignificant. While school­ ing is often a problematic variable in growth regressions (Krueger & Lindahl, 2001), investment usually is a robust variable (Levine & Renelt, 1991). Note however, that we included national rather than re­ gional investment rates. In model (2) we change our period of observation 1950-1998 into 1984-1998. In this case results of course only change in the second stage, as the dependent variable changes. For our study, the most impor­ tant change occurs with respect to the direct effect of bonding social capital on growth. In model (2) this effect is significantly negative in ac­ cordance with the model. This is an improvement relative to our basic model (1), which does not yield a significant direct relationship between growth and bonding social capital. Also remarkable is that the effect of bridging social capital becomes more than twice as large as in the basic model. At the same time initial income becomes insignificant and schooling and investment become significant. The economic interpretation is that in the more recent period, the process of catching-up is completed and regional (and national) differences play a larger role in explaining growth differentials. The overall fit of this model is worse given the Rsquared of .53 in model (1) and .44 in model (2). This is mainly caused by the poor fit of the standard economic variables, especially initial in­ come. Whereas in the longer period of 1950-1998 convergence effects can be observed, our results indicate that for a shorter period 1984-1998 this effect cannot be empirically confirmed. This result is not remarkable and fits the general thought. Other authors have shown that on the Euro­ pean regional level especially in the 80s there was no convergence, some even suggest relative divergence (e.g., Fagerberg & Verspagen, 1995; Maurseth, 2001). In the third, fourth and fifth model specifications we reduced the number of instruments or added one.

175

BRIDGING AND BONDING SOCIAL CAPITAL

Table 6.4

IV-regression

2nd Stage Dependent variable: regional economic growth (1) (2) (3) (4)

Basic

’84-’98

(5)

(6)

JobIngleTrust Regions related hart’s included excluded materia- materiaas lism lism instru­ excluded excluded ment

Initial income –.92(.16) ** –.42(.51) –.92(.16) ** –.92(.16) ** –.92(.16) ** –.84(.18)** Schooling .49(.34) 1.71(.76) ** .49(.35) .50(.35) .49(.34) .43(.35) Investment .33(.23) 1.28(.53) ** .34(.23) .33(.24) .33(.24) .35(.24) Spillover .28(.12) ** .64(.29) ** .27(.12) ** .28(.12) ** .28(.12) ** .27(.13)** Agglomeration .43(.24) * .30(.57) .42(.24) * .44(.24) * .43(.24) * .37(.24) Bonding social –.37(.23) –1.07(.51) ** –.36(.23) –.39(.22) * –.37(.22) –.35(.23) capital Bridging social .17(.08) ** .46(.18) ** .19(.08) ** .14(.09) .17(.08) ** .15(.08) capital R-squared N

.53 54

.44 54

.53 54

.53 54

.53 54

.45 52

1st Stage Dependent variable: bridging social capital Bonding social –.81(.37) ** –.80(.36) ** –.63(.38)* –1.02(.43)** –.82(.37) ** –.81(.38)** capital Job-related –.63(.27) ** –.60(.27) ** – –1.35(.24) ** –.57(.29) ** –.63(.28)** materialism Inglehart’s 2.53(.60) ** 2.6(.61) ** 3.42(.49) ** – 2.46(.61) ** 2.56(.62)** postmaterialism Trust – – – – .43(.63) – R-squared N

.58 54

.58 54

.54 54

.43 54

.59 54

.58 52

Standard errors between parentheses. * p < .1, ** p < .05. The basic equation refers to the growth period 1950-1998. We only reported the variables of interest in the first stage and excluded the other exogenous varia­ bles. In model 6 we excluded the regions that have the maximum and minimum residual in the second stage of the regression. The regions we excluded are Schleswig-Holstein (Germany) and Nord Ovest (Italy). The regression in model 6 is based on the basic re­ gression as shown in model 1.

Model (3) shows the results when the variable Job related materialism is left out. Compared with the basic model this does not yield different re­ sults. Leaving out Inglehart’s materialism index does however yield dif­

176

CHAPTER SIX

ferences. As model (4) shows, bridging social capital is not significantly positive related to growth as it is in all the other models. The overall fit of the 1st stage model goes considerably down from .58 in model (1) to .43, suggesting it is important to include Inglehart’s materialism index in the 1st stage. Adding trust as an instrument to the 1st stage regression does not yield differences with the basic model. As analyzed and discussed by Beugelsdijk and Van Schaik (see chapter five in this volume), trust is not significantly related to regional economic growth. The results in Table 6.2 suggest that trust is not indirectly related to growth either. The rela­ tion between trust and bridging social capital is not significant when we use trust as an instrument for bridging social capital. In case we add trust as an instrument and exclude the other instruments the above conclusion does not change. In our last model we tested if the reduction of observations influences our results. We have left out the regions that had the highest and lowest residual in the 2nd stage of our basic regression model (1). The regions left are Schleswig-Holstein (Germany) and Nord Ovest (Italy). The analysis for the reduced sample of 52 regions does not differ greatly of the results obtained in the basic regression on 54 regions. The main dif­ ference can be found in the fact that bridging social capital is not related to growth at the 5% significance level, but at 10% (though the reduction in significance is marginal, namely 6% versus 4%)14. 7 Conclusion and discussion We have developed a model to formalize the link between social capital, defined as participation in social networks, and economic growth. We identified two channels through which social capital and economic growth can be interrelated. First, network participation is a timeconsuming process, which crowds out working and learning time and therefore tends to be negatively correlated with growth. Second, partici­ pation in networks that span different communities may create bridging capital. Trust is generated in these networks, which protects members against rent-seeking activities. The reason is that participants that know ______________ 14

We also excluded the observations with maximum and minimum value of growth (Bayern in Germany, respectively Nord Ovest in Northern Italy) and the maximum and minimum value for initial income (Hamburg, respectively South Italy). Thirdly, we used a so-called recursive method to check of the composition of the sample influenced our re­ sults. All these checks suggest that our results are robust with respect to the potential in­ fluence of outliers.

BRIDGING AND BONDING SOCIAL CAPITAL

177

each other from the same network restrain their opportunistic behavior towards each other, to maintain reputation within the group and to avoid ostracism or lighter forms of punishment. By this second channel, the re­ lationship between growth and social capital tends to be positive. Such a positive relationship does not exist for bonding social capital and eco­ nomic growth. Bonding social capital arises from networking within own communities of close friends and family. Within the own closed circle opportunistic behavior is checked anyway, so an increase in time spent with your own close circle does not reduce opportunistic behavior in the economy. Higher levels of bonding social capital are therefore likely to go together with lower rates of economic growth, since spend­ ing more time with family and close friends comes at the cost of work­ ing and learning time. Our empirical analysis of growth in 54 European regions confirms the importance of the distinction between these two kinds of social capital. Bridging social capital is empirically good for growth, while a large importance attached to family ties is negatively re­ lated to growth. We have also stressed the fact that social capital is a choice variable that has to be explained from deeper economic and cultural variables. We think of cultural values as relatively stable over time and differing markedly across regions (cf. Baker et. al., 1981; Inglehart, 1977, 1997; Rokeach, 1973). The stability of ‘cultural’ variables over time answers the question if it is allowed to explain regional growth differentials in Europe between 1950-1998 and 1984-1998. Moreover, using a shorter period of analysis, e.g., 1991-1998 implies the use of short run growth rates, which are likely to be biased. One of the main contributions of this chapter is to provide empirical evidence for the link between differences in culture and social attitudes, on the one hand, and economic perform­ ance, on the other hand. A central variable in our analysis is materialism. For our European regional data, more importance attached to material possession is correlated with lower participation in voluntary organiza­ tions, which results through reduced bridging social capital in lower growth. Apart from generating explicit results on social values and eco­ nomic performance, our two-stage approach also allowed us to address the simultaneity problems of which other studies have been criticized. In future research, more explicit attention could be paid to the distinc­ tion between bridging and bonding social capital. Note that bonding so­ cial capital was latent in our analysis. When data is available on actual time spend with family and friends, a more explicit analysis is possible (see also Bovenberg, chapter fourteen in this volume, on the distinction between revealed and stated preferences).

178

CHAPTER SIX

Future empirical research is also needed to make the connection be­ tween the model and the empirics more precise with respect to one of the central mechanisms in our model. Measures of rent-seeking and corrup­ tion should be negatively correlated with measures of bridging social capital if our protection against rent-seeking effects is truly relevant. Un­ fortunately, this type of data on the regional level in Europe is hard to find. Also the theoretical modeling can be refined. In particular, in future work we plan to integrate into our growth framework the microeconom­ ics of reputation, opportunistic behavior and efficiency losses from cheating. We are convinced that general equilibrium modeling with micro-economic foundation can further our insights in the link between so­ cial values and economic performance and can fruitfully guide the em­ pirics of social capital and cultural values. References Baker, K., R, Dalton & K. Hildebrandt 1981. Germany Transformed. Cambridge: Cambridge University Press. Barro, R.J. 1991. Economic growth in a cross section of countries. Quarterly Journal of Eco­ nomics 106: 407–443. —— & X. Sala-i-Martin 1995. Economic Growth. New York: McGraw-Hill. Baumol, W. 1986. Productivity growth, convergence, and welfare: what the long run data show. American Economic Review 76: 1072–1085. Belk, R.W. 1984. Three scales to measures constructs related to materialism: reliability, valid­ ity and relationships to measure happiness. Pp. 291–297 in T. Kinnear (ed.), Advances in Consumer Research Vol. 11. —— 1985. Materialism: Trait aspects of living in the material world. Journal of Consumer Re­ search 12: 265–280. Beugelsdijk, S. 2002. Mapping the Landscape of Social Capital in Economics. Unpublished Manuscript. Tilburg: Tilburg University. —— & N.G. Noorderhaven 2003. Entrepreneurial attitude and regional economic growth, a cross section on 54 European regions. Chapter four in this volume. —— & A.B.T.M. van Schaik 2003. Social capital and regional economic growth. Chapter five in this volume). Bils, M. & P.J. Klenow 2000. Does schooling cause growth? American Economic Review 90: 1160–1183. Boggs, C. 2001. Social capital and political fantasy: Robert Putnam’s Bowling alone. Theory and Society 30: 281–297. Boix, C. & D.N. Posner 1998. Social capital: explaining its origins and effects on government performance. British Journal of Political Science 28: 686–693. Bovenberg, A.L. 2003. Unity produces diversity: The economics of Europe’s social capital. Chapter fourteen in this volume. Burt, R. 1992. The social structure of competition. Pp. 262–287 in N. Nohria & R. Eccles (eds.), Networks and Organizations, Structure, Form and Action. Boston: Harvard Busi­ ness School Press. Coleman, J.S. 1990. Foundations of Social Theory. Cambridge: Harvard University Press. Dekker, P., R. Koopmans & A. van den Broek 1997. Voluntary associations, social movements and individual political behaviour in Western Europe: a micro-macro puzzle. Pp. 220–239 in J.W. van Deth (ed.) Private groups and public Life: Social participation, voluntary as­ sociations and political involvement in representative democracies. London: Routledge.

BRIDGING AND BONDING SOCIAL CAPITAL

179

Fagerberg, J. & B. Verspagen 1995. Heading for Divergence? Regional Growth in Europe Re­ considered. MERIT Working Paper 2/95–014. Fine, B. 2001. Social Capital versus Social Theory: Political Economy and Social Science at the Turn of the Millennium. New York: Routledge. Francois, P. 2002. Social Capital and Economic Development. London: Routledge. Fukuyama, F. 1995. Trust: the social virtues and the creation of prosperity. New York: The Free Press. Gulati, R. 1999. Network location and learning: the influence of network resources and firm capabilities on alliance formation. Strategic Management Journal 20: 397–420. Harriss, J. & P. de Renzio 1997. ‘Missing link’ or analytically missing? The concept of social capital. Journal of International Development 9: 919–937. Inglehart, R. 1977. The Silent Revolution: Changing Values and Political Styles in Advanced Industrial Society. Princeton: Princeton University Press. —— 1997. Modernization and Postmodernization: Cultural, Economic and Political Change in 43 Societies. Princeton: Princeton University Press. —— & W.E. Baker. 2000. Modernization, cultural change and the persistence of traditional values. American Sociological Review 65: 19–51. Jackman, R.W. & R.A. Miller 1996. A renaissance of political culture? American Journal of Political Science 40: 632–659. Knack, S. & P. Keefer 1997. Does social capital have an economic pay-off? A cross country investigation. Quarterly Journal of Economics. 112: 1251–1288. Krueger, A.B. & M. Lindahl 2001. Education for growth: why and for whom? Journal of Eco­ nomic Literature 39: 1101–1136. Levine, R. & D. Renelt 1992. A sensitivity analysis of cross-country regressions. American Economic Review 82: 942–963. Lucas, R.E. 1988. On the mechanics of economic development. Journal of Monetary Econo­ mics 22: 3–42. Mankiw, N.G., D. Romer & D.Weil 1992. A contribution to the empirics of economic growth. Quarterly Journal of Economics 107: 407–431. Maurseth, P.B. 2001. Convergence, geography and technology. Structural Change and Eco­ nomic Dynamics 12: 247–276. Mincer, J. 1974. Schooling, Experience and Earnings. New York: Columbia University Press. Molle, W., B. van Holst & H. Smit 1980. Regional Disparity and Economic Development in the European Community. Westmead: Saxon House. Nahapiet, J. & S. Ghoshal 1998. Social capital, intellectual capital and the organisational advantage. Academy of Management Review 23: 242–266. Nelson, R.R. & E.S. Phelps 1966. Investment in humans, technological diffusion and economic growth. American Economic Review 56: 69–75. Paxton, P. 1999. Is social capital declining in the United States? A multiple indicator assess­ ment. American Journal of Sociology 105: 88–127. Putnam, R.D. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster. ——, R. Leonardi & R.Y. Nanetti 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press. Rebelo, S. 1991. Long-run policy analysis and long-run growth. Journal of Political Economy 99: 500–521. Richins, M.L. & S. Dawson 1992. A consumer values orientation for materialism and its meas­ urement: scale development and validation. Journal of Consumer Research 19: 303–316. Rokeach, M. 1973. The Nature of Human Values. New York: Free Press. Tarrow, S. 1996. Making social science work across space and time: a critical reflection on Robert Putnam' s making democracy work. The American Political Science Review 90: 389–398.

This page intentionally left blank

APPENDIX

A STATIC AND A DYNAMIC MODEL

182

CHAPTER SIX

A Static model Log-linearizing (2), (3), and (4) and the budget constraint in (1), we find after substituting out endogenous variables:

c� =

z� = −ε Zv + v� ,

(7’)

f� = ε Fv v� + φ� ,

(8’)

w ⎛ fw ⎞ � ⎡− v − f εFv + zεZv+ + (v + (1− v)εBz εZv+ )ζB⎤⎦ v� − ⎜ ⎟ φ+ w� , (9’) ⎣ c ⎝ c ⎠

⎞ ⎛ ⎞ ⎛ σ −σ ζB � � � c� = ⎜1+ εBzεZv+[εSf σvf +(1−εSf )σcs ]⎟ v� +⎜ vf cs εSf ⎟φ+µ+σ cs w. (10’) ⎜ ⎟ 1−ζ B σ − 1 ⎝ ⎠ ⎝ vf ⎠

Variables with a tilde are logarithmic deviations from the initial equilib­ rium; variables without a tilde refer to the initial equilibrium; εij denotes the elasticity of i with respect to j, and εij+ denotes the positively defined elasticity of i with respect to j:

ε Bz = Bz z / B > 0 ,

ε Sf = S f f / s ∈ [0,1] , ε Zv + =

v − Bzz z / > 0 ,

1 − v B z

ε Fv = 1 + σvf

ζB ε Bz ε Zv + > 1 . 1− ζB

Note that (7’)-(10’) are the log-linear equivalents of (7)-(10) in the main text; they are used to determine the partial derivatives of the functions Z, F, T and C.

BRIDGING AND BONDING SOCIAL CAPITAL

183

B Dynamic model The complete model reads: maximize µ[ln c y + δ ln co ] + ln s ,

(B.1)

subject to s = S ( f , v) and

(B.2)

cy =(1−v − f − z −l)eψl wh + (1−v)[B(z)w− D(z )eψl wh ]−

1 co , 1+r

(B.3)

where δ is the discount factor and cy and co denote consumption when young and when old, respectively. Note that we assume a Diamond-type of OLG model with ��������������������������������������� cs = 1). For simplicity, social interaction matters only for young agents. The first order conditions can be written as [note that (primed) equa­ tion numbers without prefix B correspond to (non-primed) equations in the main text]:

(1 − v) Bz ( z ) w = w ,

(2)

Sv ( f , v) w⎤ ⎡ = 1 − ⎢ D( z ) − B( z ) ⎥ , S f ( f , v) w⎦ ⎣ µS ( f , v) S f ( f , v) = , cy w

(3)

(4’)

l = [(1 − v)(1 + D( z )) − f − z] −1/ ψ ,

(11’)

co = δ(1 + r )c y .

(B.4)

Substituting (11’) and (B.4) in budget constraint (B.3), we find:

(1 + δ)c y = w[1/ ψ + (1 − v) B( z ) w / w] . Eliminating c y between this equation and (4’), we find:

µ(1 + δ) =

S f ( f , v) S ( f , v)

[1/ ψ + (1 − v) B ( z ) w / w] .

(10’)

184

CHAPTER SIX

From now on we focus on symmetry. As before, from (2) we derive z = Z(v) and from (3) we derive f = F(v,φ). Substituting these results in (11’) and log-linearizing, we find:

⎛ −v − f εFv + zεZv+ +[v + (1− v)εBz εZv+ ](1+ζ)B ⎞ ⎛ f l� = ⎜ ⎟ v� − ⎜ l ⎝ ⎠ ⎝l

⎞� 1 � . ⎟ φ+ ψ ⎠ ψl

(B.5)

Substituting z = Z (v) and f = F (v; φ) into (10’) and log-linearizing, we find:

v� =

⎛ w ⎞1 ⎤ −1 ⎡ � ⎢ ε Sf φ + µ� + ⎜ ⎟ ψ� ⎥ , εV ⎣ ⎝ (1 + δ)c ⎠ ψ ⎦

(13’)

where

εV =1+

ζB wB εBzεZv+ ⎣⎡εSf σvf + (1−εSf )⎤⎦ + [v+ (1− v)εBzεZv+ ] > 0. 1−ζB (1+δ)c

Solving (B.5) and (13’) for l we find:

⎛ ε Lv ε Sf + f / l ⎞ � ⎛ ε Lv ⎞ ⎛ εCv − εTv ⎞ 1 l� = − ⎜ ⎟φ−⎜ ⎟ µ� + ⎜ ⎟ ψ� . εV ⎝ ⎠ ⎝ εV ⎠ ⎝ εV ⎠ lψ Equations (13’) and (12’) are used to complete Table 6.2.

(12’)

CHAPTER SEVEN

SHALL THE TWAIN EVER MEET? DIFFERENCES AND CHANGES IN SOCIO-ECONOMIC JUSTICE NORMS AND BELIEFS IN EASTERN AND WESTERN EUROPE AT THE TURN OF THE MILLENNIUM WIL ARTS, JOHN GELISSEN & RUUD LUIJKX 1 Introduction Observers remarked in the early 1990s that the Velvet revolutions of 1989 and the collapse of the Soviet Union in 1991 could be compared to earthquakes that in due time will even out the watershed that has divided Eastern and Western Europe for most of the twentieth century. After these events the European landscape has changed fundamentally indeed. The most important changes were the transformation of the now postcommunist societies in Eastern Europe from planned economies and one-party authoritarian political systems towards market economies and pluralistic democratic polities. These major transitions seemed to herald in the early 1990s a reunited Europe in which ‘East’ and ‘West’ would come to resemble each other to a high degree. Many people both in the East and the West breezily insisted that Eastern Europeans wanted and would get roughly the same things—market economics, multi-party de­ mocracy, and the rule of law—that Western Europeans already pos­ sessed. At last the twain would meet! After the first decade or so after the fall of communism—at the time of this writing—we have to conclude, however, that the transformation did not work out as rapidly and as neatly as many had originally antici­ pated. Although in most of Eastern Europe the political revolutions car­ ried these countries into multi-party democracies in a matter of months, the socio-economic reorientation took years, and might even stretch through generations in some of them (Mason & Kluegel, 2000). Soon the politicians and constituents of the new Eastern European multi-party democracies came to realize that the creation of de facto plu­ ralistic democracies implied not only the implementation of constitu­ tional and legal reforms, but also the creation of democratic cultures and institutions. This was a daunting enough task, but it was relatively sim­ ple compared to the required effort to build market economies. One of the reasons that a resounding success in the socio-economic sphere re­

186

CHAPTER SEVEN

mained forthcoming in many Eastern European countries was that the in­ troduction of the new market economy was not pursued unambiguously. Most of the reformed old and formerly dissident new political and eco­ nomic elites of those countries attempted, albeit sometimes only half­ heartedly, to do away with the negative features of the old socio­ economic systems. Simultaneously, however, they tried to hold on to some of the features that they perceived as beneficial ones. As a conse­ quence privatization has stalled all over Eastern Europe. Stroppy labor unions demand concessions from foreign investors. The governments still maintain controlling or majority stakes in many companies. Giving up the reins of economic control does not come easily. That the political and economic elites in post-communist Eastern Europe made their way with difficulty and ambiguity through the transi­ tion was not only due to their own hesitations and ambivalences. It be­ came also increasingly apparent to them that the legitimization of mar­ kets does not only depend on the successful implementation of adequate institutions and correct legal procedures, but also on public opinion. In the early years of the transition public opinion in post-communist East­ ern Europe was volatile and capricious. It vented people’s hopeful feel­ ings regarding the beneficial effects of the proposed reforms in combina­ tion with anxious anticipations of the things to come. In the mid 1990s after several years of experience with the new multi-party democracy and market economy public opinion was on firmer ground. Most Eastern Europeans appeared to be not very happy with either the socio-economic or the political systems of their transforming states. What happened, they asked, to the much anticipated perks of the market economy and political pluralism? They were, however, more ambivalent about the transition toward a market economy than they were about multi-party democracy. Dashed hopes and disaffection with the market economy marked much of Eastern Europe. The reason for this increasing dissatisfaction with the socio-economic reforms was that the costs of the transition were in most Eastern European countries quite high. The increase in number of unem­ ployed, underemployed and retired people led to an alarming rise of poverty levels in these countries (Milanovic, 1998). In Western Europe in that period of time public opinion was much more stable and well bal­ anced than in Eastern Europe. The reason for this stability is a simple one. The vicissitudes of fortune were much more favorable to the public at large.

SHALL THE TWAIN EVER MEET?

187

2 Research questions In this chapter we will give further thoughts to the culture shifts, or more precisely the continuities and changes in public opinion, that occurred in Europe in that remarkable period of time. We will also look from a cross-national perspective at the cultural differences and similarities in contemporary Europe at the turn of the millennium. We will direct our attention more specifically to the vicissitudes of the popular legitimiza­ tion of the transition from a planned to a market economy in Eastern Europe and compare public opinion in Eastern Europe with the attitudes of Western Europeans. For a popular legitimization of the transition from a planned to a market economy it is not sufficient that the proper market institutions are developed and the correct legal procedures are followed. Nor is it sufficient that the citizens of post-communist states come to welcome the efficiency and tolerate the warts of the market economy. They also have to believe that, all things considered, this mar­ ket economic system is fundamentally just. They have to develop, in other words, a sense of ‘market justice’ and to abandon, at least in part, their former beliefs in ‘socialist justice’. The question of whether market justice norms and beliefs were dominant in Eastern Europe at the turn of the millennium is one of the issues we want to address in this chapter. Public opinion research has shown that in important respects market justice norms and beliefs that could legitimize capitalism were not prevalent among post-communist publics in the early transition years (cf. Kluegel, Mason & Wegener, 1995, 1999; Arts & Gijsberts, 1998). Al­ though citizens of post-communist states favored the creation of market economies in principle, they also widely endorsed many socialist norms and beliefs that prevailed in the communist era. They wanted to create societies that are freer and more democratic than the former communist systems were, but at the same time they favored more just and egalitar­ ian societal structures than can be found in some of the developed capi­ talist societies. In the early days the post-communist publics supported capitalist principles on average equally or even more strongly than Western publics, but they were also more egalitarian. They wanted to create societies were people who work hard can earn and keep good money, but where those in need are also taken care of and where a mod­ erate leveling of income differences takes place. They favored likewise a much stronger role for the state in directing the economy and ensuring the social safety net than their counterparts in the democratic capitalist states of Western Europe. Many Eastern Europeans initially subscribed to different, conflicting values, norms and beliefs at the same time. They showed in a manner of speaking signs of ideological split-consciousness.

188

CHAPTER SEVEN

Wishful thinking, i.e., unrealistically optimistic views of the market, as well as fear of the unknown seem to have played a crucial role. Being poised between hope and fear, quite a few Eastern Europeans tried to cope with the uncertainties by combining—at least in their minds—the best of two worlds. This kind of divided loyalty or split consciousness is also found in Western European countries, but seemed at least in the early stages of the transition to be more pronounced in the postcommunist states. When in the middle of the 1990s the post-communist countries of Eastern Europe progressed further into the transition popular justice norms and beliefs became more defined through experience with market economic phenomena (Kluegel, Mason & Wegener, 1999; Mason & Kluegel, 2000). The common embrace of market principles born of early transition revolutionary zeal faded, resulting in reduced average levels of endorsement in all post-communist publics. That is to say that the aver­ age level of support for capitalist principles in the East turned to a nor­ mal level, as defined by average support in the West. The divided loyalty between capitalist and socialist principles diminished, but was still more pronounced in Eastern Europe than in the West in the second stage of the transition. Is it wise to inquire into the preferences for socio-economic justice norms and beliefs in the public opinion of Eastern and Western Europe without distinguishing between Western and Eastern countries? Is East ‘East’ and West ‘West’? Looking back at the course of the transition in Eastern Europe from today’s perspective we cannot avoid the observa­ tion that at least the statement that ‘East is East’ is too rough a descrip­ tion of the post-communist condition. Not all post-communist countries have steered the same course. In some countries the transition toward a genuine pluralistic democracy and really competitive market economy has been more or less successful, but is often as far as the market econ­ omy is concerned still an uphill battle. In other countries the transition halted halfway and in still other ones the process got off to a bad start and never gained momentum. Some of them have more or less success­ fully followed a path toward a consolidated market economy, whereas others are still transitional economies and still others can be defined as statist economies that are muddling through. Despite fitful attempts to progress, Bulgaria and Romania, like Belarus, Russia, and Ukraine, are firmly in the East. By contrast, Slovenia looks at first glance little diffe­ rent from rich neighbor Austria, and seems to be essentially western. The Baltic countries are tiny; besides, Estonia is practically Nordic in all but income. They, together with the Czech Republic, Hungary, Poland,

SHALL THE TWAIN EVER MEET?

189

and Slovakia, make up the first tranche of Western Europe’s eastward enlargement. Not only are there country differences within ‘East’ and ‘West’, but there are also differences within these countries. In spite of the fact that all is not well in post-communist Eastern Europe, one also cannot avoid noticing that some people have fared fairly well from an economic point of view whereas others have experienced a downward spiral in their life chances. We may presume that the winners of the transition will have welcomed the new economic developments and embraced market justice norms, while the losers will have lamented the loss of the old system and think back longingly to socialist justice norms. We may also assume that the younger generations are far too busy with their search for jobs and opportunities to be bothered with cold-war history lessons. Young peo­ ple will therefore be the main vehicles of value change. Research find­ ings have born these assumptions out (Arts & Gijsberts, 1998). So we must not only make a distinction between the countries of Eastern Europe, but also between social categories and birth cohorts within these countries. And what applies to the East will, however to a lesser degree, apply to the West. Doing so, we fortunately do not have to start from scratch. We can take a further step forwards by extending the analyses done by the mem­ bers of the so-called International Social Justice Project (ISJP) on the 1991 and 1996 waves of their cross-national surveys (Kluegel, Mason & Wegener, 1995; Mason & Kluegel, 2000; Wegener, 2000). Data of the European Values Study (EVS) waves of 1990/1993 and 1999/2000 open up possibilities to compare market and social justice norms and beliefs in the first phase and the third phase of the transition for several Eastern European countries. They make it also possible for us to compare Eas­ tern European public opinion with that of Western Europe at two points in time. The questions we want to address in this chapter are more precisely the following. What were the differences and the similarities in socio­ economic justice norms and beliefs between and within the populations of the countries of the East and the West of Europe at the turn of the mil­ lennium? What were the continuities and changes in the popular legiti­ mization of socio-economic justice principles in the countries of Eastern and Western Europe in the past decade? Can we explain these differen­ ces and similarities, those changes and continuities and if so, how?

190

CHAPTER SEVEN

3 Theoretical framework Cross-national culture shifts are often studied by analyzing changes in people’s value patterns. Modernization theory and Culturalism are the dominant paradigms applied in this kind of research project. Both para­ digms are, however, not flawless and do contradict each other to a cer­ tain degree. According to modernization theory culture shifts are primar­ ily consequences of changing technological and economical conditions. It is assumed that people have experiences, develop attitudes, and form norms and beliefs in response to the forces that the technological and economic conditions create. The theory holds that, within broad limits, the same situational pressures, and the same framework for living, will be experienced as similar and will generate the same or a similar attitu­ dinal response by people from different countries (Arts, Hermkens & Van Wijck, 1999). If we apply this theory to our research questions the core proposition of modernization theory is the following. In so far as the modernization process, i.e., the development of a market economy, leads to a standard environment in Eastern and Western Europe with standard, institutional pressures for the populations of the different European countries and particular groups within these countries, to that degree this process should produce standard patterns of experiences, atti­ tudes, norms and beliefs. Modernization theorists therefore predict a convergence of East and West not only on the institutional level, but also as far as value patterns and interrelated systems of norms and beliefs are concerned. Most culturalists for their part would agree to the following three propositions (Wegener, 2000): 1). The population of a particular country will only support the political and economic systems of that country if its institutions and legal procedures are congruent with the shared values of that population. 2) Individuals are primarily motivated by cultural orien­ tations and only in the second place by instrumental considerations and adaptive impulses; that is, norms and values guide their behavior more than considerations of expediency. 3) Cultural orientations are a product of a shared socialization history constituting an integral part of a person’s social identity that is fairly resistant to change; it may take genera­ tions before a system of cultural values is modified noticeable. Applying this theoretical perspective to our research question one would predict that cultural change in Eastern Europe will take place by the replacement of older generations by younger ones. The theoretical assumptions of both modernization theory (Arts & Halman, 2002; Haller, 2002) and culturalism (Wegener, 2000; De Graaf, 2002) have, however, been called into question. As far as modernization

SHALL THE TWAIN EVER MEET?

191

theory is concerned critics pointed, particularly, to the empirical anoma­ lies that this theory had to content with. The pivotal assumption that technological and economic advances are the fuelling forces behind value changes appeared highly questionable. Within the EVS-context Bailey (1992), for example, argued that modernization theory, by stress­ ing technological and economic determinism, leaves no room for the impact of other nation specific situations and developments. De Moor (1994) likewise stated that modernization theory is too general to explain the dynamics of change in the various value domains and the changes in each domain in the different countries. Culturalism did not fare much better in the critique. Wegener (2000) has argued that it can hardly be denied that also other factors have an in­ fluence on shaping system support. Besides congruence between shared values on the one hand and institutions and legal procedures on the other also the economic development, the actions of the political and eco­ nomic elites, and the functioning of political and economic institutions will have a noticeable effect. Furthermore is it not obvious that, as a general condition, cultural orientations dominate the social behavior of individuals. Finally, one of the astonishing findings of studying the tran­ sition societies of Eastern Europe was that public opinion underwent drastic changes leaving behind to a considerable extent the socialist con­ victions that had dominated it for such a long time from one day to the other. Although many were also quick in rediscovering nostalgically their preferences for the previous system at a later transition stage. These value fluctuations as such are incongruent with the socialization thesis that culturalists take for granted. Does this mean that we have to face the explanatory goal of this chap­ ter empty-handed? No, this does not necessarily have to be the case. Both within the context of modernization theory and within culturalism there have been new promising developments. Ronald Inglehart (1990, 1997) for example has suggested that at least some culturalists’ amend­ ments are necessary if value researchers want to continue working with modernization theory as their paradigm. One amendment he proposed was that we must not focus all our attention on long-term developments, but that we have to take into consideration also short-term changes, such as the different phases of the business cycle, and short time events, such as wars and revolutions. They too have an impact on people’s values. The impact of the circumstances is not only temporarily, but also in cer­ tain stages of the life course a lasting one. Especially children that are of an impressionable age, i.e., are in their formative years, will later on re­ member those developments and events vividly and act according to these lasting impressions. To this observation, he added the observation

192

CHAPTER SEVEN

from the culturalist paradigm that value changes take place largely through intergenerational population replacement. Cultural shifts will therefore be much more visible if we compare younger cohorts with older ones. As a result of these deliberations he advanced both a sociali­ zation hypothesis and a scarcity hypothesis to modify and improve mod­ ernization theory. Recently Inglehart added a second amendment to his first one. To­ gether with Baker (Inglehart & Baker, 2000) he tested a more sophisti­ cated version of modernization theory by using data from the (European and World) Values Surveys data source. They found that technological and economic developments tend to push, as predicted, value patterns in a common direction. Rather than converging, however, the value pat­ terns specific to certain countries seem to move on parallel trajectories shaped by their cultural heritages. Inglehart and Baker therefore doubt that the forces of modernization will produce a homogenized world cul­ ture in the foreseeable future. They conclude that an even more sophisti­ cated version of modernization theory is necessary. One assumption of this new version should be that culture shifts are path dependent. The broad cultural heritage of a society leaves an imprint on value patterns that endures despite modernization. A second assumption should be that technological and economic developments tend to transform societies in a certain direction, but the actual processes and paths are neither fixed nor irreversible. Another example of a promising combination of the modernization and culturalist paradigms is a proposal by Szakolczai and Füstös (1998). They assume that value systems are composed of several layers, each carrying the ‘stamp’ of different axial moments, i.e., moments in which history took a decisive cultural turn. Differences among countries are therefore strongly influenced by axial considerations, like the legacy of Catholicism, or the impact of Protestantism, the influence of the Enlight­ enment, or the institutionalization of communism or social democracy. Sophisticated culturalists, moreover, realize that it would be foolish to argue that differences at the level of individual value preferences today are solely due to religious or ideological legacies. Different levels of modernization too will lead to differences in value orientations. These theoretical developments within modernization theory and cul­ turalism create the impression that a theoretical synthesis is dawning. This synthesis does not only acknowledge the importance of collective level factors, but also individual level variables. Especially demographic factors, like age, or education, gender or place of dwelling will have a decisive effect on the norms and beliefs of individuals.

SHALL THE TWAIN EVER MEET?

193

4 Hypotheses What we can safely assume if we heed the calls of modernization theory and culturalism is that people’s socio-economic justice beliefs and norms are shaped by collective as well as individual level factors. Col­ lective level factors will uniformly influence public opinion across dif­ ferent social groups within a given country. Individual level factors, however, will differentially influence public opinion between and within different social groups within a given country. We can also safely as­ sume that the collapse of communism has led to an axial crisis in Eastern Europe in the early days of the transition. As far as socio-economic jus­ tice norms and beliefs are concerned this axial crisis can be interpreted as a situation of socio-economic anomie (Arts, Hermkens & Van Wijck, 1995). If a country finds itself in a state of socio-economic anomie social and economic relations and actions are no longer regulated by a shared system of norms and beliefs, i.e., the formerly hegemonic communist ideology. Individual attitudes have been disconnected from formerly shared values, norms and beliefs, and individual disorientation is the re­ sult. At the societal level socio-economic anomie leads to the delegitimi­ zation of a socio-economic system, formerly governed by a system of socialist justice norms and beliefs. Finally, we can safely assume that over the first ten years of the transition, popular support for market jus­ tice beliefs and norms in Eastern Europe has become more ‘result ori­ ented’ (Kluegel, Mason & Wegener, 1999). People have become accus­ tomed to market economic phenomena. Only when markets and their in­ stitutions have been around there for some time Eastern Europeans have had the chance to form a grounded opinion and to pass judgment accord­ ingly. These opinions and judgments do not only depend on ‘objective’ information and ‘subjective’ experiences, but also on how they frame the new situation. Those frames are to an important extent shaped by the socio-economic, institutional and ideological context of the past. They affect people’s definition of the situation at hand and prestructure their socio-economic justice norms and beliefs (Arts & Gelissen, 2001). If we want to explain the socio-economic justice norms and beliefs that people from Eastern and Western Europe hold we have to invoke the assistance of the afore mentioned theoretical approaches: moderniza­ tion theory and culturalism. We also have to realize that our explanations must include at least two levels of analysis. One level is that of the con­ textual variables. This level is, in our case, the country level. The con­ textual characteristics which we include in our examination of European countries in 1999 are so-called global variables (e.g., post-communist and capitalist, post-totalitarian and democratic).

194

CHAPTER SEVEN

Another level of analysis is that of individual variables. Not only con­ textual factors but also the characteristics of individual citizens (such as age, gender, education, income, employment position, ideological stance, and union membership) determine the likelihood that these peo­ ple will cherish either market or socialist justice norms and beliefs. In­ cluding these individual characteristics into an explanation of socio­ economic justice norms and beliefs is important for two reasons. First, we would like to explain individual differences in socio-economic jus­ tice norms and beliefs apart from differences between countries. Second, by including individual characteristics, we control for so-called compo­ sitional effects. If individual characteristics explain, to some extent, an individual’s socio-economic justice norms and beliefs and if these indi­ vidual characteristics are unequally distributed across countries, then they also explain, to some extent the differences in socio-economic jus­ tice norms and beliefs across countries. Referring to the theoretical framework we can now formulate the fol­ lowing hypotheses. We start with three contextual factors hypotheses that will be tested on the 1999 data. The first contextual hypothesis is the modernization hypothesis. We assume that the socio-economic justice norms and beliefs of all the citizens of a specific nation state are influ­ enced in the same direction by the economic conditions of the country to which they belong irrespective of how these people themselves are af­ fected by these conditions, i.e., controlling for their individual character­ istics, but assuming that people in all countries are influenced in the same way by their individual characteristics. This hypothesis is based on the outline of modernization theory rendered in section 3 of this chapter: 1. For each European country: if a country for a considerable time has had a successful market economy generating economic growth or if a country has recently made a successful transition to a market eco­ nomy generating economic growth, citizens of that country will on average sooner cherish market justice norms and beliefs than socialist justice norms and beliefs, controlling for compositional differences. If a country is, however, still muddling through the transition to a mar­ ket economy without generating economic growth or if a country has still a planned economy, its citizens will sooner cherish socialist jus­ tice norm and beliefs than market justice norms and beliefs, control­ ling for compositional differences. The second contextual hypothesis, we will label the culturalists hypothe­ sis. Culturalist theory, as reproduced in section 3, supposes that on the map op Europe fault lines can be found. Former communist countries

SHALL THE TWAIN EVER MEET?

195

differ culturally from capitalist countries and countries with a democratic pedigree differ from countries with a totalitarian or authoritarian past. Another fault line is, for example, the division between Latin Christen­ dom and Orthodox Christianity. Furthermore, diffusion processes have an effect on the culture of a particular country. Neighboring countries are more similar to each other than countries geographically removed from each other. This hypothesis reads as follows: 2. For each Eastern European country: the closer a country’s geographic and cultural proximity to Western Europe, the higher the probability that the citizens of that country will hold market justice norms and be­ liefs and the more geographically and culturally removed a country is from Western Europe the higher the probability that its citizens will hold socialist justice norms and beliefs. The third contextual hypothesis we will call the legitimation hypothesis. Countries in which the rule of law is uphold have a higher degree of le­ gitimacy than countries in which this is not the case. This hypothesis holds, controlling for compositional differences: 3. For each Eastern European country:

– The higher the degree of privatization and the higher the aggregate (perceived) corruption in a country, the sooner the population will cherish socialist justice norms and beliefs. – The higher the degree of privatization and the lower the degree of aggregate (perceived) corruption in a country, the sooner the popula­ tion will hold market justice norms and beliefs. – The lower the degree of privatization and the higher the aggregate (perceived) corruption in a country, the sooner the population will cherish market justice norms and beliefs. – The lower the degree of privatization and the lower the degree of ag­ gregate (perceived) corruption in a country, the sooner the popula­ tion will hold socialist justice norms and beliefs. Now we proceed to the individual level hypotheses. What are the effects of individual characteristics on the socio-economic justice norms and be­ liefs of people? The first individual level hypothesis is a supplement to the modernization hypothesis. People differ as to whether they are fore­ ordained to gain or to lose by particular economic conditions. This hy­ pothesis we will dub the structural location hypothesis:

196

CHAPTER SEVEN

4. For each person: those who have a high probability to gain from a market economy and to lose from a planned economy (the young, those with higher education, those on higher income, and men) will sooner cherish market justice norms and beliefs whereas the potential losers of a market economy and winners of a planned economy (the elderly, those with only primary education, those on lower incomes, and women) will sooner cherish socialist justice norms and beliefs. The second individual level hypothesis we will call the ideological stance hypothesis: 5. For each person: those who place themselves to the left in the political spectrum will sooner cherish socialist norms and beliefs and those who place themselves to the right will sooner hold market justice norms and beliefs. The same holds for labor unionists who will sooner prefer socialist justice norms and beliefs whereas nonunionists sooner prefer market justice norms and beliefs. Finally, we formulate a transitional anomie hypothesis. This hypothesis will be derived from anomie theory as reproduced in section 3. In a so­ cietal situation of rapid social change a considerable amount of people experience a discrepancy between the values concerning the goals to be striven for and the institutionalized norms governing the means to attain these goals. This discrepancy easily leads to anomie: a societal situation where there are unclear, conflicting or unintegrated norms. In an anomic situation a society does not impress norms upon its members that bring their goals and means into harmony. Therefore people themselves have to find ways to adjust to this incompatibility. We assume that in times of transitional anomie, i.e., in times of eco­ nomic reconstruction and ideological instability, factors as to what cul­ ture people belong and how they are located in social structure determine to a lesser degree the probabilities of their choice of market or socialist justice norms and beliefs, than in times of economic and ideological sta­ bility. This hypothesis goes as follows: 6. For each European country: Under ‘normal’ societal conditions people’s socio-economic justice norms and beliefs are to a higher degree determined by their cultural belonging and structural location, than in a societal situation of socio-economic anomie. This hypothesis, in combination with two auxiliary hypotheses, gives birth to two predictions. The first auxiliary hypothesis is that Western

SHALL THE TWAIN EVER MEET?

197

Europe in both 1990/1993 and 1999/2000 can be characterized as being in a ‘normal’ societal condition, whereas Eastern Europe was in 1990/1993 in a condition of socio-economic anomie and in 1999/2000 was approaching normalcy. This leads to the first prediction: 7. Holding constant for differences between countries, in Western Euro­ pean countries people’s choices of socio-economic justice norms and beliefs will be in about the same degree explained by cultural belong­ ing and structural location in 1990/1991 and 1999/2000, whereas in Eastern Europe these choices will be explained in 1999/2000 to a higher degree by these factors, than in 1990/1993. The second auxiliary hypothesis is that because of the transformation of Eastern European economies from command to market economies a cul­ ture shift has taken place from socialist to market justice norms and be­ liefs. Because of the fact that culture shifts mainly take place by way of cohort replacement this shift will not to the same degree have occurred in each age cohort. The younger generations in Eastern Europe will have sooner and to a higher degree come to terms with the transformation to­ wards a market economy than the older generations. This leads to the second prediction: 8. Holding constant for compositional differences, in Eastern Europe in both 1990/1993 and 1999/2000 the younger cohorts will to a higher level have cherished market justice norms and beliefs, than the older ones. 5 Data and operationalization To empirically test the hypotheses 1 to 5, we will use European Values Study data from the 1999/2000 wave1. The following countries/regions are included: Austria, Belarus, Belgium, Bulgaria, Croatia, the Czech Republic, Denmark, Estonia, Finland, France, Germany (East and West are treated separately), Great Britain, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Northern Ire­ land, Poland, Portugal, Romania, Russia, Slovakia, Slovenia, and Spain. To test hypothesis 6 or rather predictions 7 and 8, we will compare findings of the EVS 1990/1993 wave with data from the 1999/2000 wave. The problem is, however, that in 1990/1993 only a limited number of countries from Central and Eastern Europe took part. For that reason ______________ 1

See for an introduction of EVS data the first chapter in this volume.

198

CHAPTER SEVEN

the comparative hypotheses are tested on data from the following coun­ tries: Austria, Belgium, Bulgaria, Czechoslovakia (separated in the Czech Republic and Slovakia), Denmark, Estonia, France, Germany (East and West separately), Great Britain, Hungary, Iceland, Ireland, It­ aly, Latvia, Lithuania, the Netherlands, Northern Ireland, Poland, Portu­ gal, Slovenia, and Spain. We note that all analyses will be restricted to people in the age range 18-64. The dependent variables are measures for socio-economic justice norms and beliefs. The questionnaires for 1990/1993 and 1999/2000 are not identical. For that reason, we are forced to use a more limited opera­ tionalization for the temporal hypothesis 6 and predictions 7 and 8 than for the other hypotheses. The operationalization of our first indicator of socio-economic justice norms and beliefs—public support for market justice—for 1999/2000 is based on the question: ‘In order to be considered ‘just’, what should a society provide? Please tell me for each statement if it is important or unimportant to you. 1 means very important; 5 means not important at all.’ Included are the following items: 1. ‘Eliminating big inequalities in income between citizens.’ 2. ‘Guaranteeing that basic needs are met for all, in terms of food, hou­ sing, cloths, education, health.’ 3. ‘Recognizing people on their merits.’ We will use the first principal component of these items. A comparison of its country-specific factor loadings shows that the differences between these loadings across countries are well within limits to allow crossnational comparison of this construct. Items indicating support for so­ cialist justice (equality and need) are reversed. Because the final con­ struct has a highly skewed distribution, we apply a logarithmic transfor­ mation to it, namely (ln(just99+2.07667)). The operationalization for our second indicator of socio-economic justice norms and beliefs—public support for government intervention— for 1999/2000 is the question: ‘Now I’d like you to tell me your views on various issues. How would you place your views on this scale?’ Scales range from 1 to 10. The following items are included: 1. ‘Individuals should take more responsibility for providing for them­ selves’ vs. ‘The state should take more responsibility to ensure that everybody is provided for.’

SHALL THE TWAIN EVER MEET?

199

2. ‘Competition is good. It stimulates people to work hard and develop new ideas’ vs. ‘Competition is harmful. It brings out the worst in people.’ 3. ‘The state should give more freedom to firms’ vs. ‘The state should control firms more effectively.’ We will use the first principal component of these items, in which the first mentioned options indicate market justice and the second socialist justice. Also in this case we checked whether there existed fundamental differences in the country-specific factor loadings across countries, but the results gave no reason to expect that comparisons of this construct would be seriously biased. Because of the lack of information we cannot make a comparison of 1990/1993 and 1999/2000 for support for market justice. For support for government intervention, we only have the first and the second indicator available for both waves. For that reason, we will analyze both variables (1-10 scales) separately. The independent variables are both individual and collective level de­ terminants. Individual level determinants, i.e., social categories and birth cohorts

– Sex: (0) female; (1) male. – Age in years. – Education in categories for the analysis of the 1999/2000 data: (1) inadequate; (2) completed compulsory/basic vocational; (3) interme­ diate vocational/general education; (4) 2nd maturity level; (5) 3rd cer­ tificate. – Age leaving school: (1) at 12 years or younger; (2) 13 years;......; (9) 20 years; (10) at 21 years or older. – Household income measured in deciles. – Union membership: (0) not a member; (1) a member. – Political left/right self placement: (1) left; …; (10) right; this variable is not available for the Baltic countries in 1992/1993. – Employment variable in categories, with dummies for those who are employed full-time or part-time, the retired, students, unemployed persons, and persons who are otherwise not gainfully employed. Here, the reference category consists of persons who are gainfully employed. – In the longitudinal analysis, we distinguish among the following birth cohorts: 1937-1945; 1946-1954; 1955-1963; 1964-1972. Here, the oldest birth cohort is the reference category.

200

CHAPTER SEVEN

Collective level determinants First, we make a distinction within Eastern Europe of the geographically and culturally close to the West ‘Intermediate’ countries and the more remote ‘Post-totalitarian’ countries. This classification is based on Mellens (1999):

– Intermediate cluster: Eastern Germany, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, and Slovenia. – Post-totalitarian cluster: Belarus, Bulgaria, Rumania, Russia, and Ukraine Because of the limited number of countries for the 1990/19931999/2000 comparison, the distinction between Intermediate and Posttotalitarian Eastern European countries cannot be made. Finally, we include several global contextual variables. First of all, we use GDP per capita in 1998, as well as the growth in GDP in the pe­ riod 1990-1998 as a predictor for the overall level of support for market justice and state intervention across countries. These data come from Maddison (2001) and they are used in the 1999/2000 analysis. We derive further ‘objective’ information concerning collective level determinants from the Freedom House reports Nations in Transit 2001. This report measures progress and setbacks in political and economic re­ form in 27 countries of Central and Eastern Europe and the New Inde­ pendent States of the former Soviet Union. This volume covers events from July 1, 1999, through October 31, 2000. These are the years that the third wave of the EVS survey has been fielded. The 2001 report is an updated edition of surveys published in 2000, 1998, 1997, and 1995. The report rates countries on a comparative basis in three broad thematic categories: democratization, the rule of law, and economic liberalization. Democratization scores encompass the average of ratings for four di­ mensions of civic and political life: political process, civil society, inde­ pendent media, and governance and public administration. Rule of law scores are an average derived from ratings for constitutional, legislative, and judicial framework and for corruption. Economic liberalization scores represent average ratings for the areas of privatization, macroeco­ nomic policy, and microeconomic policy. The ratings are based on a 1 to 7 scale, with 1 representing the most favorable level and 7 the most re­ pressive, or state-dominated, level of political and economic practice. Based on their scores, countries are divided into the following classifica­ tions of polities: consolidated democracies, transitional governments, and consolidated autocracies. Using a similar typology for economic

SHALL THE TWAIN EVER MEET?

201

policy, the countries are divided into consolidated market economies, transitional economies, and consolidated statist economies. 6 Models To test our hypotheses, we apply both multilevel analysis and OLS re­ gression. For the analyses on the 1999 data, we estimate several socalled random intercept models. The first level is the individual level and the second level is the country level. In these models, countries only dif­ fer with respect to the mean value of the dependent variables. Further­ more, it is assumed that the effects of the independent variables are ho­ mogeneous across groups. We analyze the impact of individual and con­ textual variables in several steps. Following the estimation of an ‘empty’ baseline model, we add individual-level characteristics to the model in order to control for so-called composition effects. Then, global contex­ tual variables, which are expected to affect the overall level of suppor­ tiveness for market justice and government intervention, are added to the models containing the compositional variables. The correlation coeffi­ cients presented are not standardized, because we want to compare be­ tween models and have no hypotheses about the differential impacts of independent variables within each model. To test our cross-temporal predictions, we apply OLS regression to several subsets a pooled dataset of the 1990 and 1999 EVS data. These subsets are defined by region (East vs. West) and wave (1990 vs. 1999). In these models, we include dummy variables for the countries, using Sweden as the country of reference when analyzing Western Europe and Bulgaria when analyzing Eastern Europe. In this way, we are able to as­ sess the direct relationships between the individual-level variables and the dependent variables, controlling for differences, which might exist between the countries within a particular region. 7 Findings We begin the discussion of the results by first examining the extent to which compositional and contextual characteristics are directly related to the first dependent variable under study, namely, the extent to which people endorse the principle of market justice. The findings are pre­ sented in Table 7.1. Note that these models are initially based on the samples of both the Eastern and Western European countries.

202

CHAPTER SEVEN

Table 7.1 Random intercept regression of public support for market justice: Eastern and Western European countries in 1999/2000 Eastern and Western European countries Model 1A Model 1B Model 1C Individual-level characteristics Gender (ref. = women) Age Union membership (ref. = no) Political orientation Household income Educational attainment (ref.= completed compulsory/basic vocational) Inadequate Intermediate vocational/general education 2nd maturity level 3rd certificate Employment status (ref. = working) Retired Student Unemployed Other Contextual-level characteristics Eastern European country (ref. = West) Region (ref. = West) Intermediate Eastern Post-totalitarian Eastern GDP Growth GDP 1998 Level of corruption Level of Privatization Corruption X Privatization Intercept Variance within countries Variance between countries Chi-square Df Intraclass correlation Reduction of variance within countries Reduction of variance between countries

.028 *** –.003 *** –.041 *** .028 *** .037 ***

.028 *** –.003 *** –.041 *** .028 *** .037 ***

.072 ** .046 ** .045 ** .073 **

.071 ** .046 ** .046 ** .074

–.031 * –.035 –.025 * –.029 **

–.031 * –.035 –.025 * –.029 ** –.121 **

.634 *** .196 .019 1388.9 27 .088

.596 *** .189 .020 1529.4 27 .096 .036 –.053

.656 *** .188 .017 1285.2 26 .083 .041 .105

SHALL THE TWAIN EVER MEET?

203

Table 7.1 (continued) Eastern and Eastern European countries only Western European countries Model 1D Model 1E Model 1F Model 1G Model 1H Model 1I

Model 1J

.028 *** –003 *** –.041 *** .028 *** .037 ***

.028 *** –.003 *** –.041 *** .028 *** .037 ***

.028 *** –.003 *** –.041 *** .028 *** .037 ***

.041 *** –.003 *** –.065 *** .019 *** .038 ***

.041 *** –.003 *** –.064 *** .019 *** .038 ***

.041 *** –.003 *** –.064 *** .019 *** .038 ***

.072 ** .046 ** .046 ** .074 **

.071 ** .046 ** .046 ** .073 *

.071 ** .045 ** .046 ** .073 *

.209 *** .052 ** .069 *** .148 ***

.209 *** .053 ** .069 *** .148 ***

.209 *** .053 ** .069 *** .148 ***

–.031 * –.035 –.025 * –.029 **

–.031 * –.035 –.025 * –.029 **

–.031 * –.035 –.025 * –.029 **

–.113 ** –.139 **

–.153 –.205 ** –.165

.656 **

.680 ***

–.026 –.241 ** –.020 –.033

–.026 –.241 ** –.020 –.034

–.026 –.241 ** –.020 –.033

.038 –.045

–.039 –.187 .030 .506 ***

–.302 ** .019 **†

.596 ***

.578 ***

.507 ***

.507 ***

.200 .191 .191 .188 .188 .188 .191 .011 .010 .011 .018 .018 .013 .011 444.3 336.9 319.1 1285.4 1240.7 1026.3 424.3 25 24 25 12 12 10 9 .052 .051 .054 .087 .087 .066 .053 .047 .047 .041 .041 .041 .047 .073 .005 .053 .053 .305 .033 * p < .1; ** p < .05; *** p< .01 (two-tailed test). Note: † coefficient multiplied by 1000.

204

CHAPTER SEVEN

Model 1A is a random effects analysis of variance model in which nei­ ther compositional nor contextual variables figure. It provides the basic partition of variability in the data between the individual and contextual level. Inspection of the variance components of this model shows that there is much more variability within countries than between countries. For this model, the intraclass correlation coefficient, which gives the fraction of the total variability, which is due to the contextual level, amounts to .088. However, the Chi-square test for the contextual vari­ ance components is highly significant, which is evidence that there ex­ ists significant variation among countries in their endorsement of market justice. When we turn to the analyses of the direct associations between pu­ blic support for market justice and individual-level ideological and struc­ tural characteristics (Model 1B), we corroborate many of our expecta­ tions with respect to the effects of these variables. As hypothesized, men favor market justice principles more strongly than women. Moreover, as a person’s age increases, he or she is less likely to endorse market jus­ tice; whether this is an effect of ageing or cohort replacement cannot be decided with the data at hand. Union membership is negatively related to supportiveness of market justice, whereas those who place themselves to the right of the political spectrum are, as expected more supportive of market justice. One’s household income is also positively related to sup­ portiveness of market justice principles; thus, those on higher household incomes are also more likely to support market justice. Inspection of the effects of educational attainment shows that compared to those with only basic educational attainment, persons in the higher educational catego­ ries are more likely to endorse market justice principles; this finding also holds unexpectedly for those individuals without any schooling. Finally, the effects of the employment status variables show that compared to employed persons, the retired, the unemployed and persons in the resid­ ual category are less likely to endorse market justice principles. The pro­ portion variance explained at the individual level by these variables yields .036. In Model 1C we test whether a difference in public endorsement of market justice exists between the Western European countries and the Eastern European countries, controlling for compositional differences between the countries. Note that adding the contextual characteristics does not substantially alter the effects of the compositional variables; therefore, in the following we limit the discussion to the effects of the contextual variables. The evidence based on Model 1C indicates that the publics of the Eastern European countries significantly less strongly support the principles of market justice. In Model 1D we try to disentan­

SHALL THE TWAIN EVER MEET?

205

gle the effect of the Eastern European cluster by distinguishing between an Intermediate and a Post-totalitarian region. Here, the results indicate that the publics of both Eastern European clusters significantly less en­ dorse the principles of market justice, with the publics of the Posttotalitarian cluster being the least supportive of market justice. When controlling for the level of economic growth, these differences become more pronounced (Model 1E), but the pattern found in Model 1D holds in this model as well. Finally, in Model 1F we examine the effects of the level of economic development and the level of economic growth simul­ taneously, holding constant for compositional differences. The level of economic growth seems to have an anomalous effect: it apparently de­ creases the level of support for market justice2. On the other hand, in countries that had a higher average Gross Domestic Product per capita in 1998, the publics appear to more strongly endorse the principle of mar­ ket justice than the publics in the countries with a lower level of eco­ nomic development. To test hypothesis 3, we present Models 1G to 1J. Note that these models are based on the samples of the Eastern European countries only. From inspection of the variance components estimated on the empty Model 1G, we infer that about 5 per cent of the variance in public en­ dorsement of market justice is due to differences between Eastern Euro­ pean countries. In Model 1H we control for compositional differences between these countries. All the effects are similar with respect to the di­ rection of the associations found in Model 1B, but differences exist in the strength of the direct relationships. Most notably, we find that for the Eastern European countries, only students are less supportive of market justice than the employed, whereas no differences exist between the other transfer classes and employed persons. Finally, in Models 1I and 1J we test hypothesis 3. The findings indicate that our hypotheses with respect to the impact of the level of corruption and privatization on the degree of endorsement of market justice must be rejected: holding con­ stant for compositional differences, neither the additive effects of the level of corruption and privatization, nor their linear interaction is sig­ nificantly related to the level of public support for market justice. In Table 7.2, we consider the random intercept regression models for the second indicator for market or socialist justice: public support for government intervention.

______________ 2

This could be an artefact caused by ceiling and bottom effects or by the differential influences of growth and decline rates.

206

CHAPTER SEVEN

Table 7.2 Random intercept regression of public support for government intervention: Eastern and Western European countries in 1999/2000 Eastern and Western European Countries Model 1A Model 1B Model 1C Individual-level characteristics Gender (ref. = women) Age Union membership (ref. = no) Political orientation Household income Educational attainment (ref.= completed compulsory/basic vocational) Inadequate Intermediate vocational/general education 2nd maturity level 3rd certificate Employment status (ref. = working) Retired Student Unemployed Other

–.132 *** .003 * .112 *** –.084 *** –.121 ***

–.132 *** .003 * .112 *** –.084 *** –.121 ***

.081 –.084 *** –.103 ** –.193 ***

.082 –.085 ** –.105 ** –.193 **

.083 ** –.049 .159 ** .043

.082 ** –.048 .159 ** .044

Contextual-level characteristics Eastern European country (ref. = West) Region (ref. = West) Intermediate Eastern Post-totalitarian Eastern GDP Growth GDP 1998 Level of corruption Level of Privatization Corruption X Privatization Intercept Variance within countries Variance between countries Chi-square Df Intraclass correlation Reduction of variance within countries Reduction of variance between countries

.221 **

–.021

.082

–.031

.913 .069 1145.1 28 .070

.844 .073 1318.8 28 .080 .076 –.058

.844 .063 1095.9 27 .069 .076 .087

SHALL THE TWAIN EVER MEET?

Table 7.2

207

(continued)

Eastern and Eastern European countries only Western European countries Model 1D Model 1E Model 1F Model 1G Model 1H Model 1I

Model 1J

–.132 *** .003 * .112 *** –.084 *** –.121 ***

–.132 *** .003 * .112 *** –.084 *** –.121 ***

–.132 *** .003 * .113 *** –.084 *** –.121 ***

–.135 *** .005 ** .097 ** –.070 *** –.127 ***

–.135 *** .005 ** .099 ** –.070 *** –.127 ***

–.135 *** .005 ** .099 ** –.070 *** –.127 ***

.083 –.084 ** –.104 ** –.193 **

.083 –.084 ** –.104 ** –.193 **

.082 –.084 ** –.104 ** –.193 **

.027 *** –.171 *** –.193 *** –.410 ***

.027 *** –.168 *** –.192 *** –.408 ***

.028 *** –.167 *** –.190 *** –.410 ***

.082 ** –.048 .158 ** .044

.082 ** –.048 .158 ** .044

.082 ** –.047 .159 ** .044

.064 –.266 .064 .045

.064 –.266 .064 .045

.064 –.265 .064 .045

.284 ** .095

.286 ** .099 .009

.002 –.091

–.249 –.525 .093 .317 **

–.032

–.033

.244 –.022 **†

.083

.089

.320 **

.312 **

.997 .914 .914 .844 .844 .844 .914 .069 .064 .059 .062 .061 .063 .065 448.6 461.6 467 418.6 1097.2 1097.5 1099.5 13 13 11 10 26 25 26 .065 .065 .061 .064 .067 .069 .072 .083 .083 .083 .076 .076 .076 .072 .140 .101 .116 .087 .058 * p < .1; ** p < .05; *** p< .01 (two-tailed test). Note: † coefficient multiplied by 1000.

208

CHAPTER SEVEN

As is shown by the baseline Model 2A, here too we find that the withincountry differences are far larger than the between-country differences. This is exemplified by the estimates of the variance components, which yield an intraclass correlation coefficient of .07. In Model 2B we control for compositional differences between the countries. We see, first of all, that men, on average, are less supportive of government intervention than women. As a person’s age increases, he or she appears to be more supportive of government intervention. Again, whether this finding concerns an effect of age or cohort replacement cannot be resolved with the cross-sectional data analyzed here. As we expected, union members more strongly endorse government interven­ tion than non-members, whereas persons who place themselves to the right of the political spectrum are less likely to support government in­ tervention. Moreover, we find that as the household income increases, people are, on average, less likely to consent to government intervention. The effects of the educational attainment variables suggest some support for the self-interest hypothesis: controlling for other compositional diffe­ rences, it appears that as one’s level of schooling increases, one is, on average, more reluctant to accepting government intervention. Finally, we find some evidence for differences between the transfer classes and employed persons. The findings indicate that retired and unemployed persons are more in favor of government intervention. Together, these compositional differences account for a 7.6 per cent reduction in the va­ riance at the individual level. In Model 2C, we test the hypothesis that no difference exists between the countries of Western and Eastern Europe, controlling for composi­ tional differences. The evidence indicates that this hypothesis has to be rejected: the people of the Eastern European countries appear to be more in favor of government intervention than the people of the Western European countries. Model 2D assesses whether this finding holds for both the Intermediate and the Post-totalitarian countries of the East in comparison to the Western European countries. Here, we find that in comparison with the publics of the West, citizens of the Intermediate Eastern European countries are more in favor of government interven­ tion, whereas the opinions of the citizens of the Post-totalitarian coun­ tries, on average, do not deviate significantly from the opinions of peo­ ple living in Western Europe. This finding also holds when we addition­ ally control for the economic success, which countries have had in the period 1990-1999 (Model 2E). Finally, Model 2F shows that a country’s level of economic development is directly related to the public’s endor­ sement of government intervention, whereas its growth level is not.

SHALL THE TWAIN EVER MEET?

209

Hypothesis 3 relates only to the group of Eastern European countries, and the Models 2G to 2J are based on the samples of these countries. In­ spection of the baseline Model 2G yields an intraclass correlation coeffi­ cient of .065, which is close to the intraclass correlation coefficient esti­ mated on the samples of the Eastern and Western European countries to­ gether. Comparing the effects of the compositional variables of Model 2H to Model 2B, we see that the direction of the direct relationships largely remain similar, but the strength of the relationships has changed. Most notably, differences between the lowest level and the higher levels of educational attainment are more pronounced. In addition, whereas we found significant differences between the transfer classes and the work­ ing population in Model 2B, these differences appear not to be present in the sub-sample of the Eastern European countries. Finally, the results of models 2I and 2J yield no support for hypothesis 3: holding constant for compositional differences between the countries, neither the level of cor­ ruption nor the level of privatization appears to be directly related to the public’s endorsement of government intervention; the interaction effect between these two contextual characteristics also fails to reach statistical significance. In Table 7.3, we present the findings of separate OLS regression analyses for each wave of the EVS and for the Eastern and Western European regions separately. These analyses enable us to test predictions 7 and 8. According to these predictions, it is expected that holding con­ stant for differences between countries, in Western European countries people’s choices of socio-economic justice norms and beliefs will be in about the same degree explained by cultural belonging and structural lo­ cation in 1990/1993 and 1999/2000, whereas in Eastern Europe these choices will be explained in 1999/2000 to a higher degree by these fac­ tors, than in 1990/1993 (Prediction 7). In addition, we anticipate that, holding constant for differences between countries and other composi­ tional differences, in Eastern Europe in both 1990/1993 and 1999/2000 the younger cohorts will to a higher level have cherished market justice norms and beliefs, than the older ones (prediction 8). The evidence indicates that prediction 7 must, for the greater part, be rejected. First, a comparison of Model 3A to Model 3B shows that the included individual-level characteristics explain—for the Western region—more of the variance in the level of public support for state inter­ vention in 1999/2000 than in 1990/1993. In addition, comparisons of the explained variances of Model 3E to 3F and 3G to 3H show that the in­ cluded individual-level variables actually explain less variance in the level of support for competition in 1999/2000 than in 1990/1993, and this holds for both the Eastern and Western region. The only positive

210

CHAPTER SEVEN

evidence for prediction 7 is the finding that for the Eastern region, the included individual-level characteristics explain more variance in the level of public support for state intervention in 1999 than in 1990, hol­ ding constant for other differences between the countries. Table 7.3

Public support for state intervention and competition

Western cluster (ref. = Sweden) France United Kingdom West Germany East Germany Austria Italy Spain Portugal Netherlands Belgium Denmark Northern Ireland Ireland

Eastern cluster (ref. = Bulgaria) Estonia Latvia Lithuania Poland Czech Republic Slovakia Hungary Slovakia

Public support for Public support for state intervention competition Model 3A Model 3B Model 3E Model 3F 1990 1999 1990 1999 –.883 *** .070 –.946 *** –1.494 *** –1.943 *** .371 *** .699 *** –.213 * –.501 *** –.351 *** –.775 *** .381 * –.260 **

–1.606 *** –.923 *** –1.815 *** –.301 ** –1.662 *** –.174 * –.075 –1.199 *** –.861 *** –.479 *** –1.307 *** –1.141 *** –1.071 ***

–1.350 *** –1.131 *** –.566 *** .002 –.215 –1.580 *** –1.427 *** –1.416 *** –1.541 *** –1.436 *** –1.260 *** –1.294 *** –.973 ***

–1.324 *** –.715 *** –.054 .169 .358 *** –.682 *** –.727 *** –.995 *** –1.179 *** –1.431 *** –.452 *** –.366 *** –.281 ***

Public support for Public support for state intervention competition Model 3C Model 3D Model 3G Model 3H 1990 1999 1990 1990

.833 *** 2.032 *** 1.270 *** .625 *** –.763 *** .665 *** 1.343 *** 1.123 ***

1.475 *** 2.024 *** .829 *** .947 *** .117 1.832 *** 1.305 *** 1.695 ***

.966 *** 1.166 *** .809 *** .072 1.808 *** 1.047 *** .367 *** .561 ***

–.402 *** .794 *** .152 .180 * 1.001 *** 1.005 *** .383 *** .533 ***

SHALL THE TWAIN EVER MEET?

Table 7.3

211

(continued)

Individual-level characteristics Gender (ref. = women) Birth cohort (ref. = co hort 1937-1945) Cohort 1946-1954 Cohort 1955-1963 Cohort 1964-1972 Union membership (ref. = no) Household income Educational attainment in years Employment status (ref. = working) retired student unemployed other Constant Observations R-squared

Individual-level characteristics Gender (ref. = women) Birth cohort (ref.=co1946-1954 hort 1937-1945) Cohort Cohort 1955-1963 Cohort 1964-1972 Union membership (ref. = no) Household income Educational attainment in years Employment status (ref. = working) retired student unemployed other Constant Observations R-squared * p < 0.1, ** p < .05,

*** p < .01.

Public support for state intervention Model 3A Model 3B Model 3C West 1990 West 1999 East 1990 –.448 *** –.138 *** –.388 *** **

–.048 .054 * .112 *** .323 *** –.210 *** –.051

.158 ** .179 .051 .062 *** –.299 *** –.059

.348 ** –.034 .599 *** –.013 5.691 *** 17053 .06

Model 3D East 1999 –.128 ***

.002 .089 .159 ** –.080 *** –.180 *** –.093

.111 .129 * .011 .092 –.303 *** –.052 ***

.227 .256 *** .689 *** .204

.212 .145 .625 *** .099

.205 ** .289 .672 *** .193 **

5.895 *** 14267 .09

5.568 *** 17053 .06

***

Public support for competition Model 3E Model 3F Model 3G West 1990 West 1999 East 1990 *** .342 *** .301 *** .325 **

4.919 14267 .09

Model 3H East 1999 .257 *** **

.064 .056 –.003 ** –.085 *** .175 *** .030

–.143 *** –.186 –.027 ** –.133 *** .193 ** .021

.021 .019 –.047 *** .173 *** .156 *** .051

–.162 *** –.221 –.079 –.072 *** .202 ** .019

–.208 .061 –.342 *** .080

–.145 * .178 –.389 *** .011

–.132 –.079 *** –.447 –.037

–.070 .258 –.330 *** –.094

6.847 *** 16891 .09

6.304 *** 14067 .06

5.719 16891 .08

***

5.803 *** 14067 .04

212

CHAPTER SEVEN

Finally, our findings suggest that prediction 8, which states that in Eas­ tern Europe in both 1990/1993 and 1999/2000 the younger cohorts will— to a higher level—have cherished market justice norms and be­ liefs than the older ones, must be rejected as well. Inspection of the ef­ fects of the dummy variables for the various age groups in Models 3C, 3D, 3G and 3H shows a slight tendency among the members of the more recent cohorts of actually being more in favor of state intervention than the members of the oldest birth cohort. In addition, as is shown by Model 3G, members of the birth cohorts 1946-1954 and 1955-1963 ap­ pear to be less in favor of competition than the members of the oldest cohort. We note that the overall effect of age group is rather limited: a simultaneous test of whether the age group variables, as a whole, are significant must be rejected for Model 3C, 3D and 3G at the 1 per cent significance level. Only in Model 3H is the variable age group statisti­ cally significant (p <.01). 8 Conclusions and discussion This chapter is an attempt to gain an insight into differences and simila­ rities, changes and continuities in the socio-economic justice norms and beliefs held by the populations of Eastern and Western European coun­ tries and social categories within these populations. We started from the premise, that Eastern Europe during the transition has become more and more similar to Western Europe in this respect. To test this assumption we used data from the EVS waves of 1990/1993 and 1999/2000. As a theoretical framework we adopted a synthesis of modernization theory and culturalism. From this framework we derived several hypotheses and predictions. Taking the results of our 1999/2000 wave analyses together and look­ ing at the main collective level influences, we can conclude that at the turn of the millennium, in the third phase of the transition, Eastern Euro­ peans still more strongly supported socialist justice principles and gov­ ernment intervention, than Western Europeans did. The pivotal question of this chapter was whether the twain will eventually, in a later phase of the transition, meet as far as their support for a market economy is con­ cerned. The answer is an ambiguous one. Modernization theory predicts convergence as a product of the further modernization of both Eastern and Western European countries. Because of the law of diminishing re­ turns modernization would slow down in countries that are already highly modernized and would be faster in countries that start from a lower level of modernization. The moment they arrive at the same level op modernization their populations will cherish the same kind of values

SHALL THE TWAIN EVER MEET?

213

to more or less the same degree. In line with this prediction is the finding that the higher the average GDP per capita of a country, the stronger the support for market justice principles and the weaker the support for gov­ ernment intervention. Economic growth, however, did not have the pre­ dicted consequences. The tacit assumption we cherished that nothing succeeds like success did not hold empirically. We argued, however, in footnote 2 that the finding could be an artefact. Therefore we do not want to make too much fuss about this falsification. Eastern European countries that are culturally and geographically closer to the West supported market justice principles to a higher degree than countries further removed, but in case of the supportiveness of gov­ ernment intervention the reverse is true. Post-totalitarian Eastern Euro­ pean countries are more reluctant regarding government intervention, than Intermediate Eastern European countries. This could be an indica­ tion that because of the state of economic affairs in the Post-totalitarian countries public opinion shows a considerable degree of anxiety. They support market justice principles to a lower degree than Intermediate countries, but at the same time also support government intervention to a lower degree. It looks as if they have the impression that during the tran­ sition they have jumped out of the frying pan into the fire. Still, the twain could meet in due time. Eastern European countries may very well eventually come to hold more or less the same kind of market justice norms and beliefs as Western European countries do, not only as a consequence of further modernization, but also as a result of closer economic and political cooperation and increased communication. For these latter processes would lead after some time to the diffusion of market justice ideas from the Western to the Eastern part of the conti­ nent. To our surprise neither the additive nor the multiplicative levels of privatization and corruption were related to the level of public support for market justice principles and government intervention. In spite of some hard thinking our sociological imagination does not offer plausible conjectures why this should be the case. We are at a loss for explana­ tions. To cut a long story about the contextual hypotheses short we can con­ clude that the modernization and culturalist hypotheses fared fairly well, whereas the legitimation hypothesis failed. How did the individual level hypotheses fare? Both the structural lo­ cation and the ideological stance hypotheses were corroborated. Our ex­ pectations came true that the actual or potential winners of a market economy (the young, those with higher education, those on higher in­ come and men) and those who place themselves ideologically to the

214

CHAPTER SEVEN

right are more in favor of market justice principles and less supportive of government intervention, than the actual or potential losers (the old, those with lower education, those on lower income, and women), labor unionists, and those who place themselves ideologically to the left. If we now turn to the comparison of the findings of the 1990/1993 and the 1999/2000 waves the conclusion must be that the anomie hy­ pothesis and the predictions derived from it did not work out as we ex­ pected. If our data are to believed support for socialist justice principles and government intervention are not as closely connected as we ex­ pected and they have fluctuated in both Eastern and Western Europe in the last decade in a way that was not predicted by us3. The culturalist as­ sumption that in Eastern Europe the younger generation would take the lead in the transition to market justice has also been proven wrong. If we assume that our results are robust rather than in some way artefactual, the obvious question is why the anomie hypothesis and its derivatives failed so dismally. One plausible answer could be that public opinion is much more vola­ tile than both modernization theorists and culturalists suppose. We as­ sumed that a linear development has taken place from the early to the late 1990s. There are good reasons, however, to presume that culture shifts also take cyclical and incidental forms. This means that to really put our anomie hypothesis to the empirical test we would need panel data or, even better, life event histories. Only then would we have the possibility to disentangle cohort, life course and period effects. A con­ tiguous conjecture could then be that culture shifts are to a considerable degree sooner products of contingency, than of determinacy. If we may finish on a methodological note, one might argue that our theoretical arguments and empirical analyses about socio-economic jus­ tice norms and beliefs in Eastern and Western Europe at the end of the second millennium fail to consider the complexity of the issues at hand. Modernization theory assumes that culture shifts are caused by techno­ logical and economic developments, which are more or less autono­ mous. Culturalism assumes that cultural developments are more or less autonomous and cause technological and economic shifts. If one wants to combine both approaches one has to address the question of which is the prevailing direction of causality or even whether there are causal loops due to feedback processes. If we answer the latter question in the ______________ 3

Van den Broek and De Moor (1994), analyzing the findings of the 1990/1993 wave for several Eastern European countries, mention some limitations of these data due to possible complications involved in interviewing people in these countries: until shortly before the survey, almost everyone was living a double life, systematically saying one thing in public and another in private.

SHALL THE TWAIN EVER MEET?

215

affirmative a host of theoretical and empirical estimation problems are created. One could mention, for example, the problem of cultural and structural lags. The central idea behind that problem is that various ele­ ments in an economy, a polity and a culture can change at different rates, and so integration may be far from perfect. Solving these theoretical and estimation problems is, however, as far as the state-of-the-art in EVS is concerned music of the future. We therefore decided, for the time being, to keep within bounds and sing instead a relatively simple song. References Arts, W. & J. Gelissen 2001. Welfare States, Solidarity and Justice Principles: Does the Type Really Matter? Acta Sociologica 44: 283–299. —— & M. Gijsberts 1998. After the Velvet Revolutions: Altered Life-Chances, Fragile Le­ gitimacy, and Split-Consciousness in Post-Communist Eastern Europe. Social Justice Re­ search 11: 143–172. —— & L. Halman 2002. Risk and Trust. Value Change in the Second Age of Modernity. Pp. 321–348 in P. Chmielewski, T. Krauze & W. Wesolowski (eds.), Kultura, Osobowosc, Polityka, Culture, Personality, Politics. Essays in honour of Aleksandra Jasinska-Kania. Warszawa: Wydawnictwo Naukowe ‘Scholar’. ——, P. Hermkens & P. van Wijck 1995. Anomie, Distributive Injustice and Dissatisfaction with Material Well-Being in Eastern Europe: A Comparative Study. International Jour­ nal of Comparative Sociology 36: 1–16. ——, P. Hermkens & P. van Wijck 1999. Modernisation Theory, Income Evaluation, and the Transition in Eastern Europe. International Journal of Comparative Sociology 40: 61–78. Bailey, J. 1992. Social Europe: Unity and Diversity – An Introduction. Pp. 1–14 in Bailey J. (ed.), Social Europe. London: Longman. De Graaf, N.D. 2002. De verklaringskracht van cultuur. Mens & Maatschappij 77: 137–152. De Moor, R. 1994. Epilogue. Pp. 229–232 in P. Ester, L. Halman & R. de Moor (eds.), The In­ dividualizing Society. Value Change in Europe and North America. Tilburg: Tilburg Uni­ versity Press. Karatnycky, A., A. Motyl & A. Schnetzer 2001. Nations in Transit 2001: Civil Society, Democ­ racy, and Markets in East Central Europe and the Newly Independent States. Washing­ ton: Freedom House. Haller, M. 2002. Theory and Method in the Comparative Study of Values. Critique and Alternative to Inglehart. European Sociological Review 18: 139–158. Inglehart, R. 1990. Culture Shift in Advanced Industrial Society. Princeton: Princeton Univer­ sity Press. —— 1997. Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies. Princeton: Princeton University Press. —— & W.E. Baker 2000. Modernization, Cultural Change, and the Persistence of Traditional Values. American Sociological Review 65: 19–51. Kluegel, J.R., D.S. Mason & B. Wegener (eds.) 1995. Social Justice and Political Change. Public Opinion in Capitalist and Post-communist States. New York: Walter de Gruyter. ——, D.S. Mason & B. Wegener 1999. The Legitimation of Capitalism in the Post-Communist Transition. Public Opinion about Market Justice, 1991-1996. European Sociological Re­ view 15: 251–283. Maddison, A. 2001. The World Economy: A Millenial Perspective. Paris: Development Centre of the Organisation for Economic Cooperation and Development. Mason, D.S. & J.R. Kluegel (eds.) 2000. Marketing Democracy. Changing Opinion about Ine­ quality and Politics in East Central Europe. Lanham: Rowman & Littlefield.

216

CHAPTER SEVEN

Mellens, M. 1999. Uniformity and Diversity defined. In J. de Beer. & L. van Wissen (eds.), Europe: One Continent, Different Worlds: Population Scenarios for the 21st Century. Dordrecht: Kluwer Academic Publishers. Milanovic, B. 1998. Income, Inequality, and Poverty during the Transition from Planned to Market Economy. Washington: The World Bank. Szakolczai, A. & L. Füstös 1998. Value systems in Axial Moments. A Comparative Analysis of 24 European Countries. European Sociological Review 14: 211–229. Van den Broek, A. & R. de Moor 1994. Eastern Europe after 1989. Pp. 197–228 in P. Ester, L. Halman & R. de Moor (eds.), The Individualizing Society. Value Change in Europe and North America. Tilburg: Tilburg University Press. Wegener, B. 2000. Political Culture and Post-Communist Transition – A Social Justice Ap­ proach: Introduction. Social Justice Research 13: 75–82.

CHAPTER EIGHT

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT PAUL DEKKER, PETER ESTER & HENK VINKEN 1 More civil society, better citizens? The analysis of the rise and decline of civil society has become a boom­ ing academic enterprise among social scientists. Reflecting upon the so­ cial, cultural, historical, economic, and political dynamics affecting the erosion of community in Western societies has turned into a major topic of (post)modern accounts of the Res Publica. It even seems that in an era of widespread disillusion with the highly fragmented disciplinary status of the social sciences, particularly of sociology, the preoccupation with the collapse of community develops into what appears to be a growing unifying theme (Misztal, 1996). The number of publications by social scientists on the loss of community, often phrased as the decline of so­ cial capital, is augmenting (e.g., Norris, 1999; Skocpol & Fiorina, 1999; Van Deth et al., 1999; Putnam, 2000, 2002; Dekker & Uslaner, 2001; Edwards et al., 2001). The debate on the waxing and waning of commu­ nity, the decline of civic life, the weakening of social bonds, the inflation of social capital, on what makes a good citizen, on what the good society stands for, is—once again—a core theme within the social sciences1. Contributions vary from highly normative approaches, embodied in the morally pronounced writings by (new) communitarians (e.g., Etzioni, 1993, 1996, 2001), to more matter-of-fact quantitative approaches that study empirical trends in citizen involvement (e.g., Norris, 1999a; Put­ nam, 2002). We witness a significant growth of studies on trends in citi­ zens’ civic virtues, political participation, volunteering, and involvement in informal social networks, which recently accumulated in Robert Putnam’s both much applauded and criticized book Bowling alone. The col­ lapse and revival of American community (2000). Putnam’s main mes­ sage is that in the last quarter-century Americans have become increas­ ingly disconnected from their families, friends, neighbors, communities, social institutions, and public life; in short: American communities are ______________ 1

The concern of the loss of community was of course a major concern among many of the founding fathers of the social sciences including Marx, Weber, Durkheim, and Simmel.

218

CHAPTER EIGHT

confronted with a serious and painful loss of social capital. Using a wide variety of, though rather conventional, indicators, Putnam shows that Americans are partaking less and less in the political game, are less in­ volved in religious and secular social activities, their civic participation is declining, they attend fewer informal social gatherings, and their so­ cial connections have substantially thinned. Americans trust their fellow Americans less, feel that honesty and morality are weakening, and in­ creasingly mistrust institutions and political authorities. Many American social theorists have expressed deep concerns with the loss of civil society, i.e. the private commitment with the public cause through collective participation in a wide array of voluntary asso­ ciations. A commitment that, as Alexis de Tocqueville already observed in the mid-1830s, has always been a major feature of American society. Today’s theorists put somewhat different emphases on accounting for this loss of commitment ranging from the prevailing dominance of the market calculus in the moral realm (Wolfe, 1989), the clash of opposing global forces (Barber, 1995), the severe malfunctioning of basic social institutions (Bellah et al., 1995), the absence of shared core values (Etzi­ oni, 2001), loose primary connections (Wuthnow, 1998), the decoupling of human behavior and norms of sociability (Fukuyama, 1995), the weakening of institutions (such as the family and the neighborhood) me­ diating between the private life of the individual and the megastructures of public life (Berger & Neuhaus, 1996), and the erosion of the moral fabric of family, community, and the nation (Sandel, 1996)2. Though these observers of the contemporary era point to different cultural, political, and institutional factors that have changed the rela­ tionships between the citizen, the state, the market, and the societal mid­ field, they all share the notion that a flourishing civil society and a vital democracy presupposes high levels of social trust, cohesion, and social participation. A vibrant democratic civil society is only believed to pros­ per if based on high levels of social trust accompanied by support for democracy and political involvement of its citizenry. Mutual social trust among citizens is assumed to establish strong community bonds, to posi­ tively affect the functioning of societal institutions, to generate a healthy social and cultural climate, to build social capital, and to generate tangi­ ble economic and governing revenues. Inglehart (1990, 1997) found that interpersonal trust and democratic orientations are strongly associated with both economic development and democratic stability. Trust fosters moral bonds, enables cooperation and pro-social behavior, and decreases transaction costs. High trust societies appear to be economically more ______________ 2

Cf. Eberly (2000) for an overview of classic civil society essays.

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

219

prosperous, politically more effective, and institutionally more stable than low trust societies as they are rooted in cultures of reciprocity, moral obligation, shared ethical values, and duty towards community (Fukuyama, 1995). Social trust reinforces spontaneous sociability, vol­ untary participation, and the advancement of common goals. Trust is a prerequisite of social order, facilitates social integration and solidarity, secures cooperation, diminishes social dilemmas, stabilizes social inter­ actions, is the touchstone of social capital, and cultivates face-to-face re­ lationships (Misztal, 1996). Social trust is believed to be strongly associ­ ated with civic engagement: ‘People who trust their fellow citizens vol­ unteer more often, contribute more to charity, participate more often in politics and community organizations, serve more readily on juries, give blood more frequently, comply more fully with their tax obligations, are more tolerant of minority views, and display many other forms of civic virtue (…) People who trust others are all-round good citizens, and those more engaged in community life are both more trusting and more trust­ worthy’ (Putnam, 2000: 136–137). Thus, mutual social trust among citi­ zens not only strengthens pro-social behavior at the micro-level but also generates widespread feelings of support for democratic rule and institu­ tions, and a willingness to engage in political participation and voluntary associations (Inglehart, 1990). This basic assumption of positive rela­ tionships between social trust, support for democratic institutions, and political engagement is a core notion of the present debate on the merits of the civil society. Formulated in this way the present debate is a reprise of the famous civic culture study by Almond and Verba (1963) that showed that social trust and social participation are at the nucleus of a democratic culture and may generate political surplus effects. ‘Member­ ship in some association, even if the individual does not consider the memberships politically relevant and even if it does not involve his ac­ tive participation, does lead to a more competent citizenry. Pluralism, even if not explicitly political pluralism, may indeed be one of the most important foundations of political democracy’ (Almond & Verba, 1989 (1963): 265). The political spillover impact of participation in voluntary social organizations may stem from two effects. On the one hand a ‘so­ cialization effect’ may result as citizens acquire political interest and po­ litical skills through their participation in voluntary organizations; on the other hand a ‘mobilization effect’ may occur as voluntary organizations increasingly have to enter the political arena and consequently need to recruit politically competent members (Verba et al., 1995)3. Decennia af­ ter the classic 1960s Almond & Verba study, Putnam (2000) points at ______________ 3

Cf. Verba & Nie (1972), Berry et al. (1993), Putnam (1993) for the effects of volun­ teering at the macro-level.

220

CHAPTER EIGHT

‘virtuous circles’ of trust and involvement which are directly related to the quality of democracy. Although this basic notion of the inter­ connectedness of trust, democratic support, and political participation is at the very heart of the civil society debate, its empirical robustness is scarcely tested from a comparative point of view (Dekker & Van den Broek, 1998)4. Is it indeed true that participation in civil society organi­ zations is associated with high levels of social trust, democratic support, and political engagement? Are patterns cross-nationally stable or do they hold for some national cultures but not for others? Are observable macro-patterns also to be found at the individual level? The assumed positive relationships between civil society and political democracy are by definition located at the macro level. Using individual indicators of civic engagement, social trust, and democratic support, however, offers the possibility to scrutinize underlying mechanisms and explanatory fac­ tors. Do citizens in countries characterized by high levels of social par­ ticipation indeed show more trust and democratic support or is this only true for the participants themselves? Is it a matter of context or composi­ tion? Finally: what about generational differences? Various civil society theorists argue that the decline in civic engagement is particularly ob­ servable among younger generations. Says Putnam (2000: 255): ‘It is as though the postwar generations were exposed to some anticivic X-ray that permanently and increasingly rendered them less likely to connect with the community.’ But is the assumed civic decay a cross-national typical feature, or is it a case of American exceptionalism? These are pressing questions and this chapter attempts to provide some empirical answers based on EVS data. For our comparative purposes we restrict the analyses to 22 countries: the EU member-states (with separately West and East Germany, and Great Britain and Northern Ireland) and five EU candidates. The structure of this chapter is as follows. We first explore civil soci­ ety in the twenty countries we want to investigate in this chapter, and in­ troduce the indicators for participation in civil society, social trust and democratic involvement that will be used for our analyses. We then look at the relationships between these indicators and their cultural back­ grounds from a macro-perspective. How strong are relationships be­ tween national levels of participation, trust and democratic involvement, and how important is participation as a cause of trust and involvement ______________ 4

Dekker and Van den Broek (1998) cross-nationally tested a number of hypotheses on the effects of modernization and individualization on involvement in voluntary associa­ tions both at the macro- and micro-level. They found, among other things, that the posi­ tive political effects of engaging in voluntary associations are by no means restricted to associations of a political nature.

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

221

compared to cultural factors? Next, we analyze the interconnectedness of participation, trust and involvement at the individual level. Finally, we bring in the generational perspective by looking at intercohort differen­ ces in civic engagement and its cross-national stability. Is it indeed true that younger generations do take the lead in ‘killing civic engagement’? 2 Measuring civil society, trust and democratic involvement

2.1 Civil society There are many definitions of civil society in existence (see e.g., Selig­ man, 1995; Walzer, 1995; Ehrenberg, 1999; Deakin, 2001; Chambers & Kymlicka, 2002) but in this chapter we simply choose for the non-state, non-market and non-private societal domain where voluntary association between citizens is dominant (cf. Dekker & Van den Broek, 1998, War­ ren, 2001). The problem is how that can be operationalized at the level of societal associations and networks as well as at the level of individual involvement. At the macro level we may use voluntary associations, but do we include all of them? Must some be excluded because they are not really so voluntary or should we enlarge the selection to include organi­ zations and networks that are no formal voluntary associations, but ful­ fill similar roles for the citizens and the polity? What are the criteria and how could they ever be applied in cross-national research? Then there is the individual level. Membership is sometimes nothing more than an administrative handling, but active membership is difficult to measure and only including those persons that do volunteer work is too stringent a selection criterion. Clear cut criteria for deciding which associations to include that are now in that gray area somewhere between civil society and the other realms are lacking. Because of the importance of voluntariness in our conception of civil society, associations which one ‘grows up in’, or which one can only withdraw oneself from at considerable costs, are not part of civil society. As voluntariness implies a certain plurality on the supply side, institutions that hold a monopoly in a given domain, like state churches or monopolistic trade unions, fail to qualify as well (Dahrendorf, 1991). Inner-directed associations that have no impact on society at all might be excluded too, but in our view associations do not have to be directed to public discourse as such, and neither do they need to be focused on a common good or to originate in altruistic motives. In the EVS study, membership and participation in volunteer associa­ tions are measured by asking respondents about their affiliation with and their participation in a great number of types of volunteer associations.

222

CHAPTER EIGHT

Strictly speaking, membership is an inappropriate term, as many of these associations do not know formal membership. Therefore, the expression ‘belonging to’ was used in the EVS-questionnaire. For linguistic conven­ ience, however, to belong to a voluntary association is referred to here as membership. In those cases in which persons indicated doing unpaid work in a voluntary association that they had not mentioned as an asso­ ciation they felt they belonged to, their volunteering is interpreted as suf­ ficient proof of membership of that association. In Table A8.1 in the ap­ pendix we present an overview for all countries of membership and/or volunteer work in all the reference categories. Figure 8.1 allows for an impression of the differences between countries when the differentiation between membership and volunteer work is made (cf. Dekker & Van den Broek, 1998, for EVS 1989/1990 figures). 90

gb

80

70 fr gr pt

it

60 es

Volunteers as % of members

hu

ie si

ee

be

senl

fi

pl

50

cz

lu ni

at dk

eg

40

wg

30 10

20

30

40

50

60

70

80

90

100

Members as % of population

Note: See Table A8.1 for the meaning of the abbreviations. Members = belonging to at least one of the organizations listed a1-a14 in Table A8.1 with the exception of a2 (reli­ gion). Volunteers = doing voluntary work for at least one of these organizations (all vol­ unteers are supposed to be members).

Figure 8.1 Civil society patterns

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

223

It appears that membership of civil society organizations is particularly widespread in Scandinavia and the Netherlands, but the percentage of members that are active as volunteers is modest. Opposite of this pattern are the Southern European countries with smaller percentages of the population belonging to organizations but larger percentages of the members volunteering (however, because of the substantially lower membership rates, volunteers are a much smaller percentage of the popu­ lation than in Northern Europe). Compared to the ‘broad civil societies’ of the North and the ‘elitist civil societies’ of the South, most other European countries, including the new democracies of Central and Eas­ tern Europe, combine modest membership rates with modest activity of the members. Opposite of these countries were in 1989/1990 (Dekker & Van den Broek, 1998) and again at the end of the 1990’s (Hodginson, 2003) the ‘active civil societies’ of North-America with many members and much volunteering of the members. Compared to 1989/1990 and findings in other research, the level of volunteering in West Germany is implausibly low, and that of Great Britain is implausibly high. One of the reasons for the peculiar outlier position of Great Britain is probably the strong underreporting of normal membership in sport clubs5. In this chapter we will not continue along the path of differentiating between the types of involvement in civil society. Instead we want to fo­ cus on the trust and political correlates of one similar kind of associa­ tional involvement in the countries in the study. For several associational categories and countries we have no idea what kind of organizations are behind the labels. Belonging to an organization in the area of ‘health’, can mean that someone financially supports the Red Cross, is a client of a service-delivering organization, or is an active member of a self-help group. Belonging to an organization for ‘nature and environment’ can mean a regular gift for WNF or Greenpeace, a membership card that al­ lows for a walk in a nature reserve, activity in a local advocacy group, and much more. This diversity is as such a problem, and it is a huge problem for comparative cross-national research because the proportions of the different kinds of organizations vary notably between countries6. In the tradition of the studies of civic culture and civic community men­ ______________ 5

See Table A8.1. The unlikely low involvement in sports in Great Britain will cause more surprising findings, but as long as it remains unclear what the causes for eventual underreporting are, we feel it is better to keep it with a question mark in the analysis than just drop it because we do not like the results. 6 See SCP (2000: 152-164) for an overview of frequencies of various voluntary or­ ganizations in European countries. There are different national traditions of organizing leisure and collective action. A comparison, for instance, of people involved in environ­ mental action in a Northern and in a Southern European country could in fact be a com­ parison of mainly donators with mainly demonstrators. Such a comparison is fully legiti­ mate, but it is not what we want to do here.

224

CHAPTER EIGHT

tioned before, we prefer to focus on involvement in organizations that are not intrinsically political, and that probably still have a high rate of associational ‘face-to-face’ interaction (cf. Almond & Verba, 1963; Put­ nam, 2000). The present civil society debate is not about the political relevance of political parties but about the social and political effects of membership of a sports club, singing choirs et cetera., and almost all theories describe mechanisms causing these effects stemming from faceto-face contacts between members (see, however, Selle & Strømsnes, 2001). With these criteria in mind we came to our choice for ‘leisure participation’ as indicated in the last row of Table A8.1: belonging to and/or volunteering for ‘education, arts and music and cultural activities’ and/or ‘sports or recreation’. Involvement in leisure groups, so it can be argued, is not intrinsically political and presupposes high levels of ‘faceto-face’ communication and thus is a fine example of non-political asso­ ciational involvement in civic culture and civic community. 2.2 Trust and democratic involvement Table A8.1 shows the indicators for trust and democratic involvement as well. Trust is measured by the illustrious ‘generalized social trust ques­ tion’. One may doubt whether this is a good question to measure the kind of trust that is generated in social networks and that might indicate ‘social capital’ (Edwards et al., 2001; Dekker, 2003), but as a measure­ ment of civic culture and individual confidence and efficacy it is relevant for social and political participation, and has been used in numerous studies before. The measurement of democratic involvement is a more difficult topic: there is no agreed-upon simple indicator and the available EVS data do not include exactly the kind of variables one is looking for. To measure political participation, the EVS-questionnaire only contains items that are biased towards political action7. Three to four indicators will be used in this chapter, one based on a question about the frequency of discuss­ ing politics with friends, one based on questions about joining boycotts and lawful demonstrations to measure support for political action, one measuring support for the idea that democracy is the best system, and more or less as a backup, not available in all countries, a self-assessment of political interest. The original questions and indicators are listed in Table A8.1. ______________ 7

The questions including the answers ‘have done’ and ‘might do’ about petitions, demonstrations, boycotts and occupations come from the ‘political action’-research. We will use two of this series. A question about vote intention in case of elections turns out to be unusable for an electoral participation indicator.

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

225

As a starter for the next section, Figure 8.2 shows the relationship of leisure participation with trust and three main political indicators at the level of national aggregates. 70

100 dk

se

se nl

60

90

fi

50

si

eg ni

40

70

es

nl

itgreg fr

80

dk fi belu cz ie wg

gb

ie at wg

it gb

30

at

be

60 pt es

lu

ni

cz ee

hu

frsi gr

pl

20

50 ee pl

pt

10

40

0

hu

30 0

10

20

30

40

50

60

70

80

0

10

2

20

30

40

50

60

70

80

2

Social trust (R = .51)

Political action (R = .56)

95

80

eg

cz

dk

70 wg

85

wg

gr

ee

dk

nl

se

60

fr

at

lu be

pl

75

at fi

gr

se

50

si

nl

gb

lu

itni

it ni fr

cz

pt es

40

be

65

iefi eg

ie 30

55

si

hu

hu

pl

es pt

20

ee

gb

45

10 0

10

20

30

40

50

2

60

Discuss politics (R = .26)

70

80

0

10

20

30

40

50

60

70

80

2

Democracy best (R = .38)

Figure 8.2 Relationships between leisure participation (horizontal axis) and the alleged benefits of participation

226

CHAPTER EIGHT

The four scatterplots include least-squares regression curves that best fit the data points, eventually with a squared term. The fits are better for po­ litical action and social trust than for discussing politics and the support of the idea that democracy is the best system. With the relative small number of countries we analyze, the results are quite sensitive for out­ liers. Analyses without the Netherlands or Sweden show very similar re­ sults; without both countries, we see a substantial drop of explained variance for social trust (R2 = .31) and a very different curve (almost lin­ ear) for the opinion that democracy is the best system. These results con­ firm the sensibility of small number analyses, but on the other hand it must be stressed that our 22 cases are not a small sample of a large uni­ verse but cover most of Western and Central Europe. It is not a weak­ ness of the analysis that results for Europe including the Netherlands and Sweden are different from Europe excluding these countries. They are a significant part of Western Europe and not just accidental statistical out­ liers. In this chapter we do not analyze a sample of European countries to discover global patterns, but to get a better understanding of the im­ portance of associational involvement for the European societies. In this perspective we will first examine in the next section whether there are cultural factors that better explain the national differences in the alleged benefits of participation in civil society. 3 Nations As argued, systematic empirical assessments of the association of macro-level diversity of European societies with the distinct levels of trust and democratic commitment in these societies are scarce and largely underdeveloped. In this section we aim to systematically assess the impact of different sets of cultural diversity characteristics in con­ junction with and relative to various indicators for economic, political and religious features that sharply demarcate the various nations within this continent of pluralism, especially when looking beyond the former ‘iron curtain’ and including the newly proposed EU-candidate countries into these analyses. First of all, a wide range of cultural indicators from different social science disciplines are included, ranging from the four most important cultural dimensions discerned by Hofstede (1980; 2001), the moderniza­ tion and postmodernization indices developed by Inglehart (1990; 1997), and the value universals outlined by Schwartz (1994)8. It can be expec­ ______________ 8

Country scores on Hofstede’s dimensions of power distance, uncertainty avoidance, individualism, and masculinity are included (see esp. Hofstede, 2001: 500-502). From In­ glehart (1997) we have included country scores for postmaterialism, the survival/well-

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

227

ted that many of these value-based cultural dimensions, such as power distance, postmaterialism, and hierarchy beliefs, are strongly related to trust and democratic engagement. However, an assessment of the rela­ tive impact of each of these dimensions has not yet been performed. Our analysis is the first to do so. Other pluralism features with an established impact on trust and democratic engagement are economic, political and religious. Economic indicators, such as the level of economic develop­ ment and economic growth, are for instance known to relate to levels of trust (e.g., Fukuyama, 1995; see also Beugelsdijk & Smulders, chapter six in this volume). Literature shows that political diversity, especially concerning political rights and civil liberties or concerning the recovery from communist rule, have a serious impact on the trust and democratic involvement9. Religious demarcations, particularly those between coun­ tries with a Catholic or a Protestant history, also seem highly relevant for various aspects of civil society, among which trust and democratic in­ volvement10. We have included a large number of different indicators for economic, political, and religious European diversity in our analyses11. With a series of multiple regression analyses the effect of the renowned indicators for the cultural, economic, political and religious diversity of European societies was assessed12. We systematically present the main ______________

being index and the traditional/secular-rational authority index, based on analyses of the 3rd wave EVS-dataset. From Schwartz (1994: 112-115) we have included country scores of all of his seven value universal indices: conservatism, hierarchy, mastery, affective autonomy, intellectual autonomy, egalitarian commitment, and harmony. 9 See the analyses of Inglehart (1997) for a strong case on this relationship.

10

See e.g., Putnam (2000: 65-79) for an overview.

11

For the economic domain we have included GNP per capita of the year 2000, in­ come inequality (a Gini index referring to various years) and economic growth in the 1990-1999 period (average annual % of growth of GDP) all from 2001 World Develop­ ment Indicators database collected by The World Bank (www.worldbank.org/data/databytopic/databytopic.html). Political diversity indicators include scores for the perceived corruption levels in 1988-1992 period and in the year 1999 as well as for the change in these levels between both years derived of the Internet Centre for Corruption Research (www.gwdg.de/~uwvw/icr.html). Furthermore we included freedom or level of democ­ racy indicators based on scores for political rights and civil liberties collected by Freedom House for the 1989-1990 and the 1999-2000 period as well as scores for the change in freedom between the two periods (www.freedomhouse.org/research/freeworld/2001/tables.html). An extra variable indicating which country belonged to the former pre-1990 communist world was included in the dataset (cf. Inglehart & Baker, 2000). A set of vari­ ables tapping into religious diversity was included. We discerned historically Roman Catholic from historically Protestant countries in the dataset (cf. Inglehart & Baker, 2000). Based on the 3rd-wave EVS dataset scores were included for church attendance (% of attendance of at least once a week; cf. Verweij, 1998), and for the proportions of peo­ ple belonging to a Catholic and to a Protestant denomination and people who indicate not to be a member of any denomination at present. 12 The regression analyses with the dependent variables trust, discussing politics with friends, took place in two stages. In a first stage all variables of each domain (culture, eco­ nomics, politics, religious history) were included in separate analyses per domain. The to­

228

CHAPTER EIGHT

outcomes of these analyses in this section. Findings are presented in Ta­ ble 8.1 and 8.2. Low appreciation of uncertainty avoidance and high emphasis on postmodern or well-being values yield higher levels of trust. Cultures with a high tolerance for new and ambiguous situations and with a weak orientation towards rules that might restrict exploration and experiment (Hofstede, 2001), as well as cultures in which maximizing well-being and the pursuit for quality of life instead of maximizing economic growth and efforts to provide food, clothing, and shelter (Inglehart, 1997; Inglehart & Baker, 2000), are cultures in which people are more likely to trust other people. In the economic domain the level of eco­ nomic development (expressed in GNP per capita in 2000) has a similar effect. The perceived level of corruption in the year 1999 is in the politi­ cal domain the only variable contributing to the level of trust: in coun­ tries with high subjective levels of corruption trust is low. In the reli­ gious domain we find that in countries that are historically predomi­ nantly Protestant, people express more trust. In a final analysis the influ­ ential macro-level factors for trust are both cultural and religious: a high importance of well-being values and a Protestant tradition contribute positively and almost equally strong to the levels of trust in a society. Other indicators for cultural, economic, political or religious diversity do not seem to matter much. Nor does the fact that some societies have high levels of involvement in leisure (culture and sports) associations and others have not. Trust, in short, is fundamentally an issue of culture and religious tradition. Discussing politics with friends is hardly related to macro-level diver­ sity features of European societies. In the cultural domain, Schwartz’ value universal of harmony is negatively related to informal political de­ bate. These more transcendent harmony values, such as ‘a world of beauty’, ‘protecting the environment’ and ‘unity with nature’, according to Schwartz (1994: 105–106) stand in clearest ‘opposition to enhance­ ment value types that promote actively changing the world through self­ ______________

lerance for multicollinearity was carefully checked in these analyses. Within each domain the different sets of variables (e.g., Hofstede’s indicators, Inglehart’s, and Schwartz’ in the cultural domain) were entered in the analyses in blocks (method = stepwise; pin < .05; pout < .010) to help determine which of the indicators within one set affect the dependent variables. In the cultural domain analyses with the Schwartz’ indicators was done sepa­ rately from the ones with Hofstede’s and Inglehart’s due to a larger number of missing data on the Schwartz’ indicators. The results of these analyses are reported in Table 8.1. In a second stage of the analyses all variables from the separate domain analyses were en­ tered in one analysis (method = stepwise; pin < .05; pout < .010), again distinguishing the Hofstede and Inglehart indicators and Schwartz variables in two separate analyses for the above mentioned reason. At this stage the civil engagement indicator of belonging to lei­ sure organizations was included. These analyses yielded the results reported in Table 8.2.

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

229

assertion and exploitation of people and resources’ (as is found in Schwartz’ opposite mastery and hierarchy values). This might indicate that in countries in which harmony values prevail, informal political de­ bate might well be seen as an act against the strive for transcendence and more as an expression of self-assertion and of prioritizing the vulgar immanent issue of the division of power to master the world rather than transcending individual or group interests to serve preservation and fit­ ting into the social and material environment. In these cultures informal political debate is thus less likely to occur. Addressing the effects on po­ litical discussion among friends of all the indicators of diversity at once, only the Schwartz’ harmony index remains part of the explanation; all other indicators lose their explanatory power in the analyses. At the level of nations, therefore, variance in informal political debates, is a predomi­ nantly cultural issue. At first sight, explaining political action at the macro-level is a more promising exercise. Culturally, masculinity levels and proportions of postmaterialists seem to matter most. The more masculine cultures are, the more unequally gender roles are divided, the more at the macro-level this coincides with high emphases on recognition, ambitions, and con­ frontations rather than on empathy, caring, and compromise, and the less people in these cultures are likely to involve in (unconventional) politi­ cal action. The higher the proportion of postmaterialists, stressing efforts to improve quality of life and according to Inglehart (1997: 78) only one part of the more broader cultural shift away from survival values to­ wards well-being values, the higher the level of political action. Of the economic indicators, GNP per capita also contributes positively. Among the political indicators negative effects on political action are related to corruption levels in the European transition years of 1988-1992 and the specific political history of being an ex-communist country or not. High perceived levels of corruption more than ten years ago and a communist past play down levels of political action at the turn of the millennium. Of the set of religious variables it is again a Protestant history that contrib­ utes positively. The overall analyses, including all indicators with a proven effect from the different domains in one summarizing regression analysis, show that only the ex-communist political history and the level of membership of leisure associations stand out. The political history of established communism negatively influences political action, high level involvement in leisure associations does so positively. Political action at the macro-level is not cultural, economic or religious. It is related to the great political-ideological divide of the 20th century as well as to an ac­ tive civil society, even when located at the level of leisure.

230

CHAPTER EIGHT

Table 8.1 Domains Culture

Summary multiple regressions analyses per domain Social trust Variables Effects Hofstede’s Un- Negative certainty Avoid­ ance

Discuss politics Variables Effects

Inglehart’s Sur- Positive vival Well being Schwartz’ Harmony Economy Politics

Religion

Negative

GNP / Capita Positive 2000 Corruption 1999 Negative

Historically protestant

Positive

Our core theme of the evaluation of democracy as a preferable political system has several bases in the domains of culture, economics, politics, and civil society, as the separate domain-specific analyses show. The proportion of postmaterialist and the stress on mastery values (part of the value syndrome opposite to the harmony values mentioned above) both positively influence beliefs in democracy. These beliefs seem to relate to an individual Macher’s mentality and those who value the postmodern issues of well-being that lie beyond economic survival and societal sta­ bility. Economic development (in terms of GNP per capita) contributes positively, a communist history negatively to these beliefs. The prefer­ ence for democracy according to the final overall analyses, however, loses all its cultural, economic, religious and civil society macro-level ties and is solely based on the difference between ex-communist and non-ex-communist societies with the former displaying significant lower beliefs in democracy than the latter. From the cultural domain Hofstede’s power distance dimension and Inglehart’s traditional/secular-rational authority index influence political interest, the first one negatively and the second one positively. Power distance taps into the acceptance and dependence of and respect for hier­ archies, social inequalities, status figures and authorities. The traditional/secular-rational authority index substantively aligns with the Hofste­

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

231

Table 8.1 (continued) Political action Democracy best Variables Effects Variables Hofstede’s Negative Masculinity

Inglehart’s % Positive Inglehart’s % Postmaterialists Postmaterialists

Effects

Political interest Variables Hofstede’s Power Distance

Effects Negative

Positive Inglehart’s Traditional- Positive Rational Authority

Schwartz’ Mastery Positive GNP / Capita 2000 Corruption 1988-1992

Positive GNP / Capita 2000 Positive

Ex-communist

Negative Ex-communist

Historically protestant

Positive

Negative Negative Freedom 1989-1990

Negative

Historically catholic

Negative

% non-church mem- Positive bers Note: Multiple regression analyses method = stepwise (pin < .05, pout > .10), listwise de­ letion of missing cases; separate analyses for the dimension culture with 1. Hofstede’s & Inglehart’s dimensions (max N = 19 missing data from East Germany, Great Britain and Northern Ireland) and 2. these dimensions and Schwartz’ dimensions (max N = 14, miss­ ing data from Austria, Belgium, the Czech Republic, Great Britain, Ireland, Luxemburg, Northern Ireland and Sweden); for the dimension economy GNP data and economic growth data are missing from East Germany, income inequality data from East Germany and Northern Ireland (max N = 20); for the dimension politics corruption data are missing from East Germany and Northern Ireland (max N = 20); there are no missing data for the dimension religion (max N = 22).

de dimension of power distance. Key words in the traditional authority pole of the index are obedience, respect for authority, respect for parents and a high valuation of the role of institutions such as religion and the family (e.g., Inglehart, 1997: 82). At the secular-rational pole of the in­ dex individual responsibility, personal achievement motivation, thrift and determination and the authority of political institutions are highly valued. High power distance cultures match with those situated at the traditional authority pole. The more one accentuates hierarchy and ac­ cepts power elites the less politics will be regarded as an interesting arena and vice versa the more one underlines personal responsibility re­ gardless of traditional institutionalized authorities the more one is inter­ ested in politics, the separate analyses for the cultural domain show. The level of freedom in the 1989-1990 period affects political interest too, be it negatively. The higher this level of freedom in this period of European

232

CHAPTER EIGHT

Table 8.2

Final multiple regression analyses Social trust

Culture Hofstede’s Power Distance Hofstede’s Uncertainty Avoidance Hofstede’s Masculinity Inglehart’s Materialism-Postmaterialism

Discuss politics

β

B . Exit . .

Inglehart’s Survival-Wellbeing 66.4 Inglehart’s Traditional-Rational Authority .

p .

β

B

p

.

.

.

.

. .

. .

. . .

. . .

. . .

.49

.002

. .

. .

. .

Schwartz’ Mastery

.

.

.

.

Schwartz’ Harmony

.

–16.9

–.64

.013

GNP / Capita 2000

Exit

.

.

.

Politics

Corruption 1988-1992 Corruption 1999

.

Exit

.

.

. .

. .

. .

.

.

. .

. .

. .

. .

. .

Economy

Freedom 1989-1990 Ex-communist Religion Historically catholic

.

.

.

.

.

.

Historically protestant % non-church members

18.6 .

.51 .

.002 .

. .

. .

. .

Leisure participation

Exit .192

89.0

Constant R2 (Adjusted)

Exit

7.2 .81

.000

.005 .36

.013

upheaval, the less people are politically interested. Looking closer, espe­ cially countries that were in a poor situation in terms of freedom in the early 1990s (e.g., Czech Republic, East Germany, Poland) the more inte­ rested their population ten years later are compared to countries where the populace experienced hardly any change in terms of the political and civil liberties. From the religious domain the historical background of Catholicism is a negative factor for political interest. Positive is the num­ ber of non-church members: the higher the proportion of people saying they do not belong to any church, the higher the level of political interest in a society. The final analyses of political interest show that the civil so­ ciety factor is the one and only determinant that explains cross-national variation in political interest. Culture, political and religious history ef­ fects disappear in favor of the influence of civil society in the leisure do­

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

Table 8.2

(continued)

Political action β

B

Democracy best p

β

B

Political interest p

β

B

.

.

.

.

.

.

Exit

. Exit Exit

.

.

. . Exit

. .

. .

. . .

. .

. .

. .

. .

. .

. .

. Exit

.

.

.

Exit

.

.

.

.

.

.

Exit

Exit

p

. .

. .

.

.

.

.

.

.

.

.

.

Exit .

.

.

. .

. .

. .

. .

. .

. .

. –14.6

. –.43

. .028

. –13.2

. –.82

. .001

Exit .

.

.

.

.

.

.

.

.

. .

. .

. .

. .

. .

. .

.47

.53

.009

Exit

.000

90.9

56.6

233

Exit .

. .61

.000

.

Exit 25.8

.77

.006 .001

.61 .001 .64 .001 .54 .006 Note: Exit = removed from multiple regression analyses with method = stepwise (pin < .05, pout > .10); Listwise deletion of missing cases; GNP / Capita 2000 deleted from analyses for political action due to low collinearity tolerance < .200 (results similar).

main. A strong involvement in this civil society is the best guarantee for high levels of political interest. The macro-level analyses, in summary, show that trust and informal political debate have fundamentally cultural roots. Countries with con­ trasting cultural profiles vary also in the levels of trust between people and the incidence of political debate among people. The more emphasis in these cultures is put on postmodern well-being values, the more peo­ ple trust one another. Protestant dominance in a country’s history has the same positive effect on trust. Moreover, stress on transcendent harmony values tones down political discussions in the private realm. Political ac­

234

CHAPTER EIGHT

tion and political interest are more than in culture, economics or religion rooted in civil society. Strong involvement in leisure, culture and sports associations boosts unconventional political action (boycotting and demonstrating) and political interest. Political action is hampered if a country was part of the communist block in the 20th century. Views on democracy as a political system, finally, have a similar basis in political history. In countries that were part of the communist world, doubts on democracy as an ultimate political system are highest. 4 Individuals Shifting from national aggregates to individual data, we first look in this section at the impact of some standard socio-demographic variables on differences in leisure participation, and then on the importance of leisure participation and these variables to explain differences in the alleged so­ cietal benefits of participation. We conducted separate logistic regression analyses for all 22 (sub)nations. The results are shown in Table 8.3 as (adjusted) odds ratios; coefficients greater than 1 indicate positive effects and coefficients smaller than 1 negative effects. The first block of Table 8.3 shows that in most countries leisure par­ ticipation is more popular among men and, to a lesser degree, among the youngest part of the population. This reflects the higher popularity of club sports among younger men. The exception of Great Britain, where women tend to participate more than men, is probably a consequence of the very small number of British respondents that reported involvement in sport clubs. With a minor exception of a non-significant coefficient in Portugal, the higher educated are clearly over-represented among the participants in all countries. With a few exceptions, church attendance is not significantly related to leisure participation. The strong effect of edu­ cation across the board is somewhat surprising as our participation measurement includes the ‘unsophisticated’ area of ‘sports and recrea­ tion’ (beside ‘education, arts and music and cultural activities’, a proba­ bly more high-browed area, that however attracts less people in most countries; see Table A8.1). Thus, it appears at least implausible that in­ volvement in voluntary associations is a vital compensation for a lack of individual resources for social trust and democratic political involve­ ment. In the rest of the table we will see what positive effects of partici­ pation remain after adjustment for the effects of education (and the other three socio-demographic variables). The first impression of the coefficients printed in italic is that of overall positive effects of leisure participation on the chance that an in­ dividual has social trust, is involved in political discussions and political

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

235

action, and supports democracy as the best political system. There are no significant negative effects of participation. A second finding is that in general leisure participation is more im­ portant for political action than for the other two political traits and also somewhat more important than for social trust. It is also a finding in other research that political participation is a stronger correlate of social participation than social trust or other attitudes (cf. Dekker, 2002: 24– 27). This finding can give rise to the speculation that the organizational impact of social participation is more important today than its socializing role: there are many other places nowadays where people can learn to trust other people, discuss politics or develop a positive attitude toward democracy, but the organizational infrastructure of associations and their institutional links with (local) politics are often still important conditions for actual political involvement. However, one can think of more trivial explanations for a relatively strong participation-participation link. With the data at hand we cannot say more about the mechanisms behind the statistical correlates. A third finding is that adjustment for the effects of the socio­ demographic variables leads to somewhat smaller effects of leisure par­ ticipation, but not substantially smaller. In a few cases small significant effects turn into statistical insignificance, but in most cases participation effects survive adjustment for the other factors. If leisure participation makes a difference, it does so above the socio-demographic traits and not as a consequence of them. However, if one compares the adjusted effects of participation and the other variables, it becomes clear that across the board education is statis­ tically more important than leisure participation. Again, this is not a pe­ culiar finding for our leisure participation indicator of civic involvement. Also for indicators of social participation that include interest organiza­ tions an other areas, education seems to be at least as significant as a de­ terminant of political attitudes and participation and social trust (Dekker, 2002: 29–30). Differences in Table 8.3 between the countries as regards the effects of leisure participation have not been systematically analyzed. However, there are evidently no simple patterns of strong and weak effects follow­ ing a wind direction or an order of the countries according any of the cultural dimensions of the last paragraph. In the next section we will ad­ dress the final issue of this chapter, i.e. to examine how crucial genera­ tions are in explaining differences of participation and the alleged benign effects of participation.

236 Table 8.3

CHAPTER EIGHT

Backgrounds of leisure participation and the benefits of leisure participation in 20 countries: Odds ratios AT

BE

CZ EG

Leisure participation Female Age 45+ Higher education Church attendance Social trust Participation unadjusted Participation adjusted Female Age 45+ Higher education Church attendance Discuss politics Participation unadjusted Participation adjusted Female Age 45+ Higher education Church attendance Political action Participation unadjusted Participation adjusted Female Age 45+ Higher education Church attendance Democracy best Participation unadjusted Participation adjusted Female Age 45+ Higher education Church attendance

DE WG

DK

EE

ES

FI

.5 .8 2.0 1.2

.6 .7 2.4 1.2

.5 .5 2.1 .9

.5 .6 1.8 .9

.7 .6 2.2 1.3

.7 .8 2.2 .9

.8 .4 2.0 1.1

.8 .7 2.8 1.1

.7 .6 2.5 1.2

2.0 1.7 .8 1.2 2.0 1.3

2.0 1.7 .8 1.1 2.0 1.1

1.2 1.1 1.1 1.4 1.7 1.2

1.5 1.3 .8 1.0 2.3 1.6

1.0 .9 .8 .9 1.5 1.7

2.4 2.1 1.0 .7 1.8 1.1

1.7 1.6 .9 1.3 1.8 1.0

1.2 1.1 1.0 .8 1.3 1.1

1.5 1.4 1.1 1.2 1.4 1.3

2.0 1.7 .6 1.4 2.3 1.2

2.1 1.7 .6 1.3 2.6 1.7

1.2 1.2 .7 1.6 1.6 .9

3.1 2.7 .7 1.3 3.7 1.8

2.6 2.1 .3 1.0 2.1 1.6

1.7 1.5 .7 .9 1.9 1.4

1.3 1.3 .7 1.5 1.6 .7

1.7 1.3 .5 1.0 2.1 1.2

1.3 1.2 .7 2.0 1.7 .7

2.4 1.9 .6 .5 2.1 .7

2.5 1.7 .6 .5 3.2 1.0

1.7 1.3 .7 .7 2.1 .8

3.7 2.9 .7 .6 2.1 .9

2.1 1.6 .6 .5 2.1 .9

1.9 1.5 .8 .3 1.5 .8

2.1 1.8 .6 .9 2.1 .8

3.3 2.5 .6 .6 2.3 .6

2.0 1.7 .6 .4 1.2 .5

1.3 1.1 .7 1.2 1.6 .9

1.2 1.0 .8 1.8 2.5 1.1

1.3 1.2 .8 1.2 1.7 1.2

1.3 1.2 .9 .9 1.4 1.6

1.2 1.3 1.2 2.1 1.3 1.3

1.3 1.2 .7 2.1 2.5 1.2

1.9 1.7 .5 1.0 1.7 .9

1.5 1.4 .9 .9 1.2 1.0

1.2 1.1 .9 1.1 1.6 .9

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

Table 8.3 FR

GR

237

(continued) HU

IE

IT

LU

NL

PL

PT

SE

SI

UK GB

NI

.7 .8 2.5 1.1

.6 .8 4.2 1.1

.8 .8 5.6 1.1

.4 .7 2.8 1.7

.6 .7 2.7 1.1

.8 .7 1.9 1.5

1.2 1.0 3.2 1.0

.3 .7 3.7 .5

.2 1.0 1.2 .7

.8 .7 1.9 .9

.5 .5 3.6 .9

2.2 1.3 3.1 2.8

.5 .8 3.1 1.5

1.5 1.3 1.1 1.0 2.0 1.4

2.2 2.4 1.0 2.2 1.2 .4

2.3 1.9 1.0 1.0 1.9 1.2

2.0 1.7 .6 1.2 1.5 1.5

1.9 1.6 .9 1.3 2.6 1.2

1.5 1.4 1.0 1.3 2.1 1.3

2.1 1.7 .9 .9 2.2 1.0

1.6 1.8 1.3 1.4 1.5 1.8

2.7 2.5 .6 .8 1.2 1.8

1.7 1.6 1.0 .8 1.7 1.0

2.1 1.6 1.0 1.2 2.8 1.0

2.4 1.9 .9 1.4 2.1 1.5

1.3 1.2 .8 1.2 1.1 1.5

1.6 1.5 .6 1.9 2.0 1.2

1.7 1.4 .4 1.4 1.5 .9

2.4 1.9 .6 1.4 2.2 .8

2.0 1.7 .6 1.8 2.1 1.1

2.5 1.9 .5 1.3 2.9 .8

1.9 1.7 .9 1.7 2.3 1.7

1.9 1.6 .7 1.0 2.2 1.2

2.3 1.5 .5 1.0 3.2 1.5

1.8 1.3 .4 .8 1.3 1.2

1.9 1.7 .9 1.3 2.6 1.0

1.6 1.2 .6 1.1 2.0 1.0

3.0 2.8 .4 1.8 1.9 1.8

2.1 1.9 .6 1.3 1.4 1.4

3.7 2.8 .5 .6 2.5 .8

2.9 1.9 .6 .5 2.1 1.4

3.7 2.6 .7 .6 2.4 1.1

2.2 1.7 .5 .7 2.1 .7

3.5 2.4 .5 .5 2.4 .7

2.0 1.8 .8 .6 2.0 .7

2.0 1.7 .5 .4 2.0 .7

3.3 2.3 .5 .7 1.6 .8

2.8 2.2 .6 .6 1.6 .7

2.4 2.1 1.0 .4 1.7 .9

2.4 1.5 .6 .6 2.5 .6

1.2 1.2 .4 .8 2.2 1.0

2.2 2.0 .6 .7 1.1 .7

1.3 1.9 2.2 1.2 1.1 1.6 1.2 1.6 .8 1.2 1.1 .9 .9 1.2 2.1 1.8 1.1 1.0 1.5 1.0 1.3 .7 1.1 .9 .8 .8 .7 1.3 1.1 .7 .8 .9 .7 .9 .7 .8 .9 .7 .5 1.3 1.0 1.4 1.5 1.7 1.7 1.8 1.4 1.4 1.2 1.9 2.2 1.9 1.4 .9 2.8 1.3 2.3 1.9 2.3 2.1 1.0 1.7 2.1 1.5 1.3 1.2 1.3 1.0 1.4 1.1 1.2 1.1 .8 .9 .8 .9 1.4 1.1 Note: Higher education: in each country about half of the oldest respondents when com­ pleting education; church attendance: attends religious services at least once a month; sig­ nificant odds ratios (p < .05 two-tailed) are put in italic.

238

CHAPTER EIGHT

5 Participation in civil society and generational change The current debate on what is taking place among generations in terms of their civic engagement is strikingly conflicting. The generational per­ spective dominates the discussion on what the past, present, and future state of civic engagement is. In almost every observation and analysis in the field the rise, decline, and—according to a few number of analysts— the revival of civic engagement is closely connected to the issue of gene­ rations. Robert Putnam (2000: 283) suggests that 50% of the decline in most forms of civic engagement and social capital in the US is due to generational change, making the generational argument the single most important one in his analysis13. The process of the gradual replacement of the civic prewar generation of ‘self-made citizens’ by the postwar generations of baby boomers and Gen X’ers (and the like), is what ‘killed civic engagement’. One of the problems with this diagnosis is that it stops with pinpointing the mechanism of change: generational re­ placement. What it fails to indicate are the basic commonly shared fea­ tures of these different generations that make them more or less engaged in civil society14. An answer to the straightforward question why youn­ ger, postwar generations, including the sixties generation, are less in­ volved in their community or have less social capital than older genera­ tions is simply missing. Or ad hoc explanations are offered such as ex­ cessive TV watching. Similar problems have American popular studies that au contraire predict the rise of a ‘new great generation’ of optimis­ tic, ‘more civic-spirited’ young people manifesting a ‘wide array of posi­ tive social habits (...) and good conduct’ (Howe & Strauss, 2000: 4, 13)15. The energy with which an inevitable turn for the better is commu­ nicated on almost every page of these studies is proportionally reversed to the number of sound empirical arguments that substantiate the sudden rise of this new great generation. Both Putnam and these upbeat studies make a strong generational statement—Putnam for alarming decline, the others for powerful revival—but both do not elucidate their most basic assumptions. Are younger generations of the postwar era less civicly in­ volved or not, and how to account for this? Are younger generations out­ side the US equally less (or more) embedded in community life, civicly ______________ 13

It has to be stressed, though, that Putnam is well aware of the possibility that the younger generation today ‘is no less engaged than their predecessors, but engaged in new ways’ (Putnam, 2000: 26) but he is not very convincing in elaborating this very likely al­ ternative interpretation of intergenerational civic participation in postwar American so­ ciety. 14 See for partly overlapping arguments against the Putnam’s conception and use of generation as an explanatory factor the recent study by Scott McLean (2002). 15 See also Tapscott (1999) for similar statements.

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

239

engaged, oriented towards social life, or are we witnessing a typical American phenomenon? In order to answer these fundamental questions we will analyze whether younger generations in Western societies are indeed less participating in advancing the good society compared to older generations. We will not include all the EVS-countries in our generational assess­ ment but restrict the analysis to seven countries (in alphabetic order): Belgium, Denmark, France, Great Britain, Italy, the Netherlands, and West Germany16. This smaller selection of countries allows us to more closely examine positions of generations over time, in our case between the years 1981, the first wave of EVS, and 1999. Three generations are discerned labeled the ‘War Generation’ (born before 1940), the ‘Baby Boom Generation’ (born between 1940 and 1960) and the ‘Baby Bust Generation’ (born after 1960). Generation studies across Europe argue that cohorts whose members have, in their formative years, experienced the Second World War in the 1940s and its aftermath in the 1950s (War Generation), the cultural and political upheaval of the 1960s and 1970s (Baby Boom Generation), and the severe economic crisis and pervasive political shifts of the 1980s and 1990s (Baby Bust Generation) form dis­ tinctive generations (see, e.g., Arber & Attias-Donfut, 2000; Diepstraten et al., 1999a,b; Edmunds & Turner, 2002; Kohli & Szydlik, 2000; Van den Broek, 1996; 2001)17. Members of each of these three generations are assumed to share a strong sense of a common history and a similarly common destiny. Following generation theory, members of each of the three generations should also share basic behaviors, values, attitudes and opinions, in our case behaviors, values, attitudes and opinions regarding civic engagement, trust and political involvement. These generations, furthermore, overlap with the ones mentioned by Putnam (2000) and de­ picted above: the prewar generation, the baby boomers, and Gen X’ers. We have examined the importance of leisure participation, in this case culture participation, for the level of social trusts and extent of political involvement of the three generations18. Table 8.4 sketches the first de­ tails. ______________ 16

Only West Germany participated in the EVS-wave of 1981.

Of course, more detailed country-specific categorizations of generations are possi­ ble: e.g., the Weimar, Third Reich, or post-Berlin Wall generation in Germany, or the ‘generation of 1914’ in several European countries involved in the First World War (Wohl, 1979). Van den Broek (2001), however, convincingly shows that many countryspecific generation classifications overlap with the more general pattern of a (pre)war generation, sixties or protest generation, and post-sixties generation. 18 Culture participation (being a member and/or volunteer for organizations in educa­ tion, arts, music or cultural activities) is part of the concept of Leisure participation in the previous sections of this chapter. The other part is sports or recreation organizations. This part is not measured in the 1981-wave of EVS. 17

240

CHAPTER EIGHT

Table 8.4 Proportions members/volunteers for arts/culture activities organizations by generations in 7 countries 1981-1999 BE 81 War Generation 9.4 Baby Boom Generation 13.5 Baby Bust Generation 8.7 Total 10.8 N 124

90 11.8 20.7 19.8 17.1 477

99 16.8 19.8 21.1 19.5 371

WG 81 8.7 3.9 6.9 6.8 89

90 12.1 13.1 10.3 12.0 252

99 8.8 9.0 7.6 8.4 87

DK 81 5.1 5.8 10.5 5.8 69

90 10.5 14.3 12.5 12.5 129

99 13.3 17.6 16.7 16.3 165

With only a few exceptions members of the War Generation, born before 1940, participate least in cultural organizations in most years of the European Values Study. The exceptions are found in 1981. More than twenty years ago in Belgium, France, Italy, and, the Netherlands, mem­ bers of the Baby Bust Generation, born in 1960 or later, participated least. In West Germany, however, the Baby Busters are clearly least in­ volved in culture organizations from the 1990s onwards. The general trend is upward: in most countries and for most genera­ tions participation as a member and/or volunteer in culture organizations increases. France, and, again, West Germany are the exceptions. After a rise between 1981 and 1990 participation rates in 1999 dropped among all three generations in France and West Germany, albeit to a level (slightly) higher than in 1981. Generally speaking, the sharpest rise in participation rates is found among the Baby Bust Generation, especially in Belgium and Italy. Differences with the rise among the Baby Boom­ ers, born between 1940 and 1960, are usually small. In Germany and Denmark rising participation among the Baby Boomers is highest. This last finding contributes to the idea that the West German young generation in particular is the overall exception. The participation rate in organized culture among German Baby Busters is lowest and hardly any higher in 1999 than it was in 1981. This underlines the analysis of post­ war Germany by Schülze (2000) who detects a decollectivization of classic Feierabendvergnügen activities in favor of activities that boost excitement and action, have the allure of eccentricity and Genuss and are found in the domains of informal leisure and consumption. Perhaps West German Baby Busters are the prototypical young generation parting from classicly organized culture and indulging in new forms of enter­ tainment that can be individually enjoyed and satisfy the pursuit for the joyeux experience outside formal settings. This analysis underlines the importance for civil society studies of taking more modern individual­ ized types of participation into account (cf. Vinken & Ester, 2002; see also below).

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

Table 8.4

241

(continued)

FR GB 81 90 99 81 6.3 8.0 6.0 6.1 5.8 9.7 8.8 7.4 5.3 10.4 8.6 8.2 6.0 9.3 8.0 6.8 72 93 129 79 Note: War Generation born < Generation born > 1959.

IT NL 90 99 81 90 99 81 90 99 11.2 7.0 3.1 3.6 8.5 11.1 25.6 42.2 10.3 10.2 5.1 4.7 8.8 16.8 41.1 50.0 9.2 11.6 1.7 6.7 12.5 10.4 42.4 45.2 10.4 9.9 3.7 4.8 10.4 13.4 36.3 46.2 153 96 50 97 207 163 369 462 1940; Baby Boom Generation born 1940-1959; Baby Bust

Taking account of these country differences we have assessed the bene­ fits for social trust and political involvement of being member of younger generations and participating in culture organization. We have done so in a few distinct steps. Table 8.5 reports on these steps and their results. Table 8.5

Social trust and political involvement by culture participation and generations in 7 countries 1981 and 1999: Odds ratios

Culture participation Wave 1999 Culture participation in 1999 Culture participation Wave 1999 Culture participation in 1999 Baby Boom/Bust Generation Baby Boom/Bust Generation in 1999 Baby Boom/Bust Generation participating in 1999 Culture participation Wave 1999 Culture participation in 1999 Baby Boom Generation Baby Bust Generation Baby Boom Generation in 1999 Baby Bust Generation in 1999 Baby Boom Generation participating in 1999

Social trust 1.7 1.2 1.3 1.7 1.1 1.3 1.3 1.0 1.0 1.7 1.1 1.3 1.3 1.3 1.0 1.0 1.1

Discuss politics 2.0 2.1 1.1 2.1 1.8 .9 3.7 .8 1.3 2.1 1.8 .9 3.4 5.4 .7 .7 1.6

Political action 2.4 1.3 1.1 2.4 1.3 1.1 1.5 .8 1.0 2.4 1.3 1.2 1.6 1.1 .9 1.1 1.3

Baby Bust Generation participating in 1999 .9 1.1 .9 Note: Baby Boom Generation born 1940-1959; Baby Bust Generation born > 1959; significant odds ratios (p < .05 two-tailed) are put in italic; controlled for country.

Culture participation is important for social trust, as we have mentioned before and is shown again here. Those who participate in organized cul­ tural activities, i.e. activities for which one should be a member or can volunteer for, trust other people more. There is also an effect of the year

242

CHAPTER EIGHT

1999, a proxy of a period effect in other words. In the year 1999 people have more social trust than in 1981, and especially people who partici­ pate in culture organizations in 1999. When these higher levels of trust are primarily generational, than the effect of time and of participation in 1999 should diminish if not disappear in favor of generation member­ ship. This is exactly what happened after including generational mem­ bership in the analysis. The period effect has dropped to insignificance. Culture participation as such is still crucial for trust, but doing this in 1999 no longer is. Important for having trust besides participating in cul­ ture organizations in general, is being a member of a younger generation (Baby Boom and/or Baby Bust Generations)19. Independent of culture participation, younger generations grown adult (long) after the Second World War trust other people more than those who have been shaped by experiences of the Second World War. Culture participation and time (1999) relate to informal political de­ bate. Discussing politics with friends is done more by people who par­ ticipate in formal cultural activities and who live at the turn of millen­ nium. The latter effect of time weakens, but stays significant when in­ cluding generations in the analyses. Postwar generations, and especially the Baby Bust Generation, are by far more inclined to discuss politics within informal social settings. Postwar generations (taken as a whole and separately), however, are less willing to discuss politics in the 1999­ period. In reference to the significant time or period effect, one may conclude that political discussion seem a general non-generational fea­ ture of the end of the 20th century. In the early 1980s, however, this type of discussions were much more generational with more debate among younger generations. Still, this conclusion can be specified even further. Baby Boomers participating in cultural organizations in 1999 are more involved in political discussions. This participation effect is not found among the youngest, Baby Bust, generation. Participation in organized culture does not seem to rise the importance of political debate for the youngest generation at the turn of the century. Their higher involvement in political discussions is furthermore both time-less and time-bound. Time-less, because as a generation they discuss politics more than the oldest generation independent of time (or participation on formal culture ______________ 19

See Diepstraten et al. (1999a, b; compare Van den Broek, 1996, 2001) for argu­ ments to divide generations in a (pre)war and postwar typology, in our case the War ver­ sus Baby Boom and Baby Bust generations. In our analyses we first verify the importance of this crude two-type classification and in a second step we discern the Baby Boom from the Baby Bust generation (with the War Generation as the reference category in both cases). We also include two-way and three-way interaction terms of generations with time and culture participation. Country positions are also included in the analyses, but not re­ ported here. The results in other words are controlled for country differences.

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

243

organizations). It is also time-bound, because this generation-specific in­ volvement in political discussions is, however, most remarkable at the start of the 1980s. People involved in culture participation and in the year 1999 are also more willing to take political action. The period effect does not decline and does stay significant after taking generational differences into ac­ count. Apparently the late 1990s are the period of vivid political activity. Most politically active are the younger, especially the Baby Boom Generation, regardless of time or participatory involvement. The Baby Bust generation is in terms of political action not any different from the War Generation, either in the early 1980s or late 1990s or when participating in culture organizations or not participating in these organizations. This might of course be due to the classic concept of political action used in our analyses, i.e. joining boycotts and attending demonstrations, activities, as Norris (2002: 1–2) puts it, that have moved from margin to mainstream. Alternative avenues to political activism, e.g., via consumer behavior (buying afghan bags, patchouli oil or the tiedyed T-shirt are a few examples Norris mentions in reference to the sixties generation) and leisure behavior (e.g., via activities on the Internet), play a marginal role in civil society research (see also Vinken & Ester, 2002). Political action the way we measured it is found among the participating segment of society, is explicit at the turn of the century and is typical for the sixties generation of Baby Boomers. 6 Discussion Civil society theorists share the notion that a vibrant civil society and a flourishing democracy assume high levels of social trust, cohesion, and social engagement of its citizenry. A vital democratic civil society will only thrive if rooted in widespread social trust accompanied by basic support for the democratic system and a socially and politically involved populace. Social trust, democratic support and citizen engagement foster social capital, strong communities, moral bonds, cooperation and prosocial behavior, political effectiveness and stability, and voluntary in­ volvement in advancing the common good. Trust, democratic support, and civic engagement are assumed to be strongly interconnected and fundamental pillars of a healthy civil society that strongly advocates voluntary involvement. Civic engagement in voluntary associations is presupposed to have important social spillover effects: voluntariness creates a more competent citizenry (Almond & Verba, 1989). Social trust is believed to be strongly associated with civic engagement and, vice versa, civic engagement creates social trust. Though this fundamen­

244

CHAPTER EIGHT

tal assumption of the interconnectedness of trust, democratic support, and civic engagement as voluntary participation is at the very heart of the civil society debate, its empirical robustness is scarcely tested from a systematic cross-national point of view (Dekker & Van den Broek, 1998). Is it indeed a cross-nationally observable phenomenon that par­ ticipation in voluntary organizations yields higher social trust, democ­ ratic support, and political involvement? Are observable patterns stable across different societies both at the macro- and micro-level? Do such patterns differ along generational lines as many civil society adherents believe to be the case? Are younger generations withdrawing from ad­ vancing the good society? These are important questions and this chapter attempted to provide some answers using EVS data from 22 countries. Core variables are: involvement in voluntary associations, social trust, and democratic involvement (political interest, discussing politics, sup­ port for democracy, and political action). At the macro-level we observed that trust, democratic support, and political involvement have fundamental cultural roots. Countries with diverging cultural profiles also diverge in the levels of social trust of its citizens and their willingness to engage in politics. The more these cul­ tural profiles emphasize postmodern well-being values, the more its citi­ zens express basic feelings of social trust and the less they will ignore politics. Interestingly, high levels of social trust are particularly to be ob­ served in predominantly Protestant societies. In turn, political action and political interest are clearly rooted in what makes a civil society. Strong civic involvement in leisure, cultural and sports associations was found to reinforce unconventional political action and political interest. En­ gagement in political action and supporting democracy are less popular in countries that were once part of the communist rule. At the individual level we found positive statistical effects of leisure participation on social trust and the political involvement indicators in most countries. Across the board the effects were strongest for political action—suggesting that participation-to-participation mechanisms might be more important than the socializing impact of associational life—and education appears to be more important than leisure participation to get people to trust their fellow citizens and get them involved in democratic politics. This finding, also reported in other research, should temper high expectations of the social and political blessings of associational life. Policymakers who care about the loss of social trust and political invol­ vement may better keep focused on their tiresome efforts to improve educational changes than simply join the call for community and the re­ vival of associational life. Our findings confirm that associational life can play a positive role for the development of Almond and Verba’s

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

245

‘competent citizenry’, but it is not a compensation for a lack of indivi­ dual resources. At the generational level we found that members of the youngest gen­ erations, born after the Second World War, are most involved in leisure (i.e. culture) organizations. The sharpest rise in involvement is generally found among the Baby Bust Generation born after 1960. Culture partici­ pation is important for social trust and political involvement. Cutting across time and generations, participation in formally organized cultural activities generates more social trust, more political debate in informal circles and political activism. Generational membership is important for trust and political action. As far as informal political discussion goes there are no single, straightforward generation effects. The two youngest generations, the Baby Boom (born between 1940-1960) and Baby Bust generations (born after 1960) are more trustful of other people. The Baby Boom generation furthermore is more involved in classic activism, i.e. joining boycotts and demonstrations, activism that has moved from margin to mainstream. Both younger generations as such do not take up distinct positions in discussing politics with friends. At least in the late 1990s. In the early 1980s the younger generations, especially the very young—i.e. those socialized in the 1970s—were more actively discuss­ ing politics in their social circles. In the late 1990s only the Baby Boom generation participating in formal cultural organizations is. Besides these generation effects there is a period effect. In the late 1990s political action seems most vibrant. In reply to Putnam, we may conclude, that young generations outside the US, are not less but more involved in participatory life compared to older generations, more in­ stead of less inclined to trust other people, and even more and not less politically involved, mostly regardless their levels of involvement in civil society. The two findings, Baby Boomers more frequently indulging in classic types of political activism and formally organized Baby Boomers more greatly involved in political discussions within their real-life social cir­ cle, in our view uncover a major shortcoming in political survey re­ search. As hypothesized a number of times before in this chapter, mem­ bers of young generations who are socialized or have had their major formative experiences in the late 1980s and 1990s, might well have cho­ sen types of political engagement and platforms of this engagement— e.g., by using Internet-based strategies—that are not tapped with the classic political science indicators (see Vinken & Ester, 2002). In line with the risen number of critical views on the focus of mainstream po­ litical surveys we aim to stress that it is very likely that each generation develops its own types of political activism and utilizes its own plat­

246

CHAPTER EIGHT

forms of political discussion. There is an urgent need for a profound re­ orientation in political science toward these contemporary types and platforms of political engagement. The indicators of political engage­ ment measured in the European Values Study are of the classic type. It therefore highly recommended to take a critical view of the political in­ dicators of EVS in order to be conceptually innovative and to be sensi­ tive to the signs of time. References Almond, G.A. & S. Verba (1963)1989. The civic culture. Newbury Park: Sage.

Arber, S. & C. Attias-Donfutt (eds.) 2000. The myth of generational conflict. London:

Routledge. Barber, B.R. 1995. Jihad vs. McWorld. New York: Random House. Barnes, S., M. Kaase et al. 1979. Political action. Beverly Hills: Sage. Bellah, R.N., R. Madesen, W.M. Sullivan, A. Swidler & S.M. Tipton 1992. The good society. New York: Vintage Books. Berger, P.L. & R.J. Neuhaus (1977) 1996. To empower people. Washington: American Enter­ prise Institute. Berry, J.M., K.E. Portney & K. Thomson 1993. The rebirth of urban democracy. Washington: The Brookings Institution. Beugelsdijk, S. & S. Smulders 2003. Bridging and bonding social capital: which type is good for economic growth? Chapter six in this volume. Chambers, S. & W. Kymlicka (eds.) 2002. Alternative conceptions of civil society. Princeton: Princeton University Press. Deakin, N. 2001. In search of civil society. Basingstoke: Palgrave. Dekker, P. 2002. De oplossing van de civil society. Den Haag: Social & Cultural Planning Of­ fice. —— 2003. Social capital of individuals. In P. Selle & S. Prakash (eds.), Investigating social capital. London: Sage (in press). —— & E. Uslaner (eds.) 2001. Social capital and participation in everyday life. London: Routledge. —— & A. Van den Broek 1998. Civil society in comparative perspective. Voluntas 8: 11–38. Diepstraten, I., P. Ester & H. Vinken 1999a. Mijn Generatie. Tilburg: Syntax Publishers. —— 1999b. Talkin’ ‘bout my generation. Netherlands’ Journal of Social Sciences 35: 91–109 Eberly, D.E. (ed.) 2000. The essential civil society reader. Lanham: Rowman & Klittlefield. Edmunds, J. & B.S. Turner 2002. Generations, culture and society. Buckingham: Open Univer­ sity Press. Edwards, B., M.W. Foley & M. Diani (eds.) 2001. Beyond Tocqueville. Hanover: Tufts Univer­ sity Press. Ehrenberg, J. 1999. Civil society. New York: New York University Press. Etzioni, A. 1993. The parenting deficit. London: Demos. —— 1996. The new golden rule. New York: Basic Books. —— 2001. The monochrome society. Princeton: Princeton University Press. Fukuyama, F. 1995. Trust. London: Penguin Books. Hodginson, V. 2003. Volunteering in global perspective. In P. Dekker & L. Halman (eds.), The values of volunteering. New York: Plenum (in press). Hofstede, G. 1980. Culture’s consequences. Newbury Park: Sage Publications. —— 2001. Culture’s consequences (second edition). Thousand Oaks: Sage Publications. Inglehart, R. 1990. Culture shift in advanced industrial society. Princeton: Princeton University Press. —— 1997. Modernization and postmodernization: cultural, economic, and political changes in 43 societies. Princeton: Princeton University Press.

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

247

—— & W.E. Baker 2000. Modernization, cultural change and the persistence of traditional values. American Sociological Review 65: 19–51. Kohli, M. & M. Szydlik (eds.) 2000. Generationen in Familie und Gesellschaft. Opladen: Les­ ke & Budrich. McLean, S.L. 2002. Patriotism, generational change, and the politics of sacrifice. Pp. 147–166 in S.L. McLean, D.A. Schultz & M.B. Steger (eds.), Social capital. Critical perspectives on community and ‘Bowling Alone’. New York & London: New York University Press. Misztal, B.A. 1996. Trust in modern societies. Cambridge: Polity Press. Norris, P. 1999 (ed.). Critical citizens. Oxford: Oxford University Press. —— 2002. Democratic phoenix. New York: Cambridge University Press. Nye, J.S., P.D. Zelikow & D.C. King (eds.) 1997. Why people don’t trust government. Cam­ bridge: Harvard University Press. Putnam, R.D., R. Leonardi & R.Y. Nanetti 1993. Making democracy work: civic traditions in modern Italy. Princeton: Princeton University Press. —— 2000. Bowling alone. New York: Simon & Schuster. —— (ed.) 2002. Democracies in flux. Oxford: Oxford University Press. Sandel, M.J. 1996. Democracy’s discontent. Harvard: Harvard University Press. Schwartz, S.H. 1994. Beyond individualism/collectivism. Pp. 85–122 in U. Kim et al. (eds.), Individualism and collectivism. Thousand Oaks: Sage Publications. SCP 2000. The Netherlands in a European perspective. Den Haag: Social & Cultural Planning Office. Seligman, A.B. 1995. The idea of civil society. Princeton: Princeton University Press. Selle, P. & K. Strømsnes 2001. Membership and democracy. Pp. 134–147 in P. Dekker & E. Uslaner (eds.), Social capital and participation in everyday life. London: Routledge. Schulze, G. 2000. Die Erlebnisgesellschaft. Frankfurt/Main: Campus Verlag. Skocpol, T. 1999. Advocates without members. Pp. 461–509 in T. Skocpol & M.P. Fiorino (eds.), Civic engagement in American democracy. Washington: Brookings Institution Press. —— & M.P. Fiorina (eds.) 1999. Civic engagement in American democracy. Washington: Brookings Institution Press. Van den Broek, A. 1996. Politics and generations. Tilburg: Tilburg University Press. —— 2001. De verraderlijke charme van het begrip generatie. Tijdschrift voor Sociologie 22: 329–360. Van Deth, J.W., M. Maraffi, K. Newton & P. Whiteley (eds.) 1999. Social capital and Euro­ pean democracy. London: Routledge. Verba, S. & N.H. Nie 1972. Participation in America. New York: Harper & Row. Verba, S., K.L. Schlozman & H.E. Brady 1995. Voice and equality. Cambridge: Harvard Uni­ versity Press. Vinken, H. & P. Ester 2002. The culture of engagement. Paper presented at the ISA XV World Congress of Sociology. Brisbane, Australia. Walzer, M. (ed.) 1995. Toward a global civil society. Providence: Berghahn. Warren, M.E. 2001. Democracy and association. Princeton: Princeton University Press. Wolfe, A. 1989. Whose keeper? Berkeley: University of California Press. Wuthnow, R.D. 1998. Loose connections. Cambridge: Harvard University Press.

APPENDIX

INDICATORS (%) IN 20 COUNTRIES

250

CHAPTER EIGHT

Table A8.1 Indicators (%) in 20 countries AT BE CZ ‘Please look carefully at the following list of voluntary organizations and activities and say: a) which, if any, do you belong to?; and b) which, if any, are you currently doing unpaid voluntary work for?’ yes to a) and or b): a1 social welfare services for elderly, handicapped or de­ prived people a2 religious and church organizations a3 education, arts and music and cultural activities a4 trade unions a5 political parties or groups a6 local community action on issues like poverty, em­ ployment, housing, racial equality a7 third world development or human rights a8 conservation, the environment, ecology, animal rights a9 professional associations a10 youth work (e.g., scouts, guides, youth clubs et cetera) a11 sports or recreation a12 women’s groups a13 peace movement a14 voluntary organizations concerned with health a15 other groups ‘Generally speaking, would you say that most people can be trusted or that you can’t be too careful in dealing with people?’ b1 most people can be trusted b2 cannot be too careful b3 don’t know Frequency of discussing political matters ‘when you get together with your friends’ c1 frequently c2 occasionally c3 never c4 don’t know Position toward ‘different forms of political action that people can take’ Joining in boycotts d1 have done d2 might do d3 would never do d4 don’t know Attending lawful demonstrations e1 have done e2 might do e3 would never do e4 don’t know

6 12

7

DE EG WG

3

4

26 12 7 12 14 14 19 12 6 8 20 16 11 8 7 12 7 4 3 3 3 3 9 7 4 23 4 1 9 9

5

4

1

1

10 1 0 1 11 8 2 3 9 6 3 5 8 8 1 2 24 24 18 31 9 3 4 4 2 1 0 0 6 7 2 3 12 10 5 4

32 29 23 41 31 63 69 75 55 64 5 2 2 4 4

20 15 21 27 22 54 50 66 61 61 26 36 13 12 17 0 0 0 0 0

9 11 8 5 11 33 31 28 37 39 54 53 56 51 43 3 4 8 7 7 16 39 26 44 21 35 30 40 31 41 47 30 29 20 32 1 2 5 4 5

This page intentionally left blank

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

251

Table A8.1 (continued) DK EE ES

FI FR GR HU IE

IT LU NL PL PT SE SI

UK GB NI

6

4

4

12

6

14

3

7

7

19

22

4

2

21

7

18

4

12 16 54 7 6 4 13 11 7 33 2 1 4 14

8 9 5 2 3 0 2 4 3 10 3 0 1 5

6 8 4 2 3 3 3 3 3 8 2 2 3 4

47 15 34 7 3 7 5 6 8 25 5 1 10 12

4 8 4 2 2 2 2 3 2 17 0 1 3 8

15 14 7 6 7 5 9 9 4 11 3 7 7 9

14 4 7 2 2 1 3 4 2 5 0 1 2 3

16 10 10 4 6 3 3 8 7 28 5 2 4 6

11 10 6 4 3 3 4 7 5 12 1 2 5 3

12 20 14 7 7 14 13 7 11 27 7 3 10 5

35 46 23 9 7 25 44 19 7 50 4 3 10 10

6 3 10 1 2 0 1 4 2 3 1 1 2 3

6 3 2 1 1 1 0 1 1 9 0 1 2 3

72 26 64 11 9 15 11 14 7 37 3 1 7 25

8 10 17 3 11 1 4 7 5 17 2 1 3 10

10 10 8 2 4 6 9 8 18 6 3 4 12 .

24 8 6 3 3 2 1 4 6 15 4 2 4 7

64 32 3

22 74 5

37 58 5

57 42 2

22 76 2

18 77 5

22 77 1

35 63 2

32 66 2

25 71 3

59 40 1

18 79 3

10 88 2

64 32 4

21 77 2

29 68 3

39 58 3

25 55 20 0

14 66 18 2

7 46 47 1

8 66 26 0

12 53 35 0

16 57 27 0

10 43 46 0

12 47 40 0

13 55 32 0

18 52 30 0

16 64 20 0

20 55 23 1

12 39 49 0

19 60 20 0

8 64 28 0

9 39 51 1

15 52 33 0

24 36 37 3 29 38 31 2

3 24 63 10 10 28 53 8

5 21 59 15 24 30 36 9

14 51 30 5 14 42 40 4

12 40 41 7 39 33 26 2

3 17 78 2 39 36 25 0

3 18 76 3 4 29 65 2

9 34 53 4 20 44 33 2

10 41 44 5 33 38 25 4

8 38 49 5 28 41 28 3

23 39 38 0 34 35 31

4 24 71 1 10 31 58 0

4 32 60 4 14 37 44 4

32 54 12 2 35 52 12 1

8 51 36 5 9 55 31 5

17 13 43 25 37 58 3 4 13 38 46 3

21 29 45 4

252

CHAPTER EIGHT

Table A8.1 (continued) AT BE CZ Opinion about ‘Democracy may have problems but it’s better than any other form of government’ f1 agree strongly f2 agree f3 disagree / disagree strongly f4 don’t know Interest in politics g1 very interested g2 somewhat interested g3 not very interested g4 not at all interested g5 don’t know g6 not asked Leisure participation = a3 and/or a11 Social trust = b1 Discuss politics = c1 + c2 Political action = d1 or d2 and e1 or e2 Democracy best = f1 Political interest = g1 + g2

DE EG WG

59 56 39 32 66 32 51 51 52 30 8 7 7 6 3 4 3 3 10 2 21 46 23 10 0

10 20 21 18 29 50 45 41 31 21 25 29 30 9 8 11 0 0 0 0

32 32 74 60 59 67

36 29 64 72 56 40

31 23 87 71 39 70

35 31 83 69 66 59

22 41 88 78 32 67

CIVIL SOCIETY, SOCIAL TRUST AND DEMOCRATIC INVOLVEMENT

253

Table A8.1 (continued) DK EE ES

FI FR GR HU IE

IT LU NL PL PT SE SI

UK GB NI

70 26 1 3

38 48 6 8

35 51 9 4

58 30 6 6

62 33 4 1

35 50 7 8

42 49 6 4

58 28 4 10

47 48 4 2

21 57 9 13

38 49 7 7

24 61 9 6

46 30 21 3

41 42 6 11

5 22 31 42 0

5 23 49 22 0

9 27 31 33 0

12 28 40 20 1

9 34 32 25 0

7 26 42 26 0

12 36 31 21 0

16 52 25 7 0

6 35 33 25 0

6 23 32 39 0

5 37 33 24 0

8 30 27 35 1

9 31 36 24 0

23 21 72 73 24 42

15 29 48 67 46 37

20 39 66 55 41 40

17 60 8 16

17 43 30 9 0 100 41 64 80 75 70 60

16 22 81 46 17 .

23 48 17 12

100 14 37 52 56 38 27

34 57 74 72 35 28

22 22 65 76 58 36

20 18 73 78 62 40

8 22 54 36 23 .

50 43 6 2

100 32 35 60 69 35 43

19 32 68 78 42 32

38 25 70 73 58 48

70 59 80 79 47 68

6 18 76 44 21 42

11 10 51 55 38 29

52 64 80 94 50 .

Note: AT = Austria, BE = Belgium, CZ = the Czech Republic,

DE = Germany: East (EG) and West (WG), DK= Denmark, EE = Estonia, ES = Spain,

FI = Finland, FR = France, GR = Greece, HU = Hungary, IE = Ireland, IT = Italy,

LU = Luxembourg, NL = the Netherlands, PL = Poland, PT = Portugal, SE = Sweden,

SI = Slovenia, UK = United Kingdom: Great Britain (GB) and Northern Ireland (NI).

This page intentionally left blank

CHAPTER NINE

MISSING LINKS? CONSENSUS DEMOCRACY, VOLUNTARY ASSOCIATION AND EUROPEAN VALUES WIM VAN DE DONK, FRANK HENDRIKS & RUUD LUIJKX If the members of a nation—or a class or a group—do have a conception of common purpose, democracy will enable them to express it and to work for it. If they are trying to give themselves some such guide, democracy will provide them with the freedoms to seek it. But where positive goals are lacking, democracy itself will not supply them. People do ask for a purpose in life, seeking to find something greater than themselves with which to identify – a cause, a movement, a historical or moral reality. In this quest, the central legitimising principle of modernity is at best neutral. (Samuel Beer, in Beer et al., 1973: 4–5)

1 Introduction In this chapter, we will use EVS data to explore the link between values and institutions. More specifically, we will try to relate these data and two theories, both highly influential in contemporary political science and public administration: Putnam’s Social Capital Theory and Lijphart’s Consensus Democracy Theory. Both authors (among many others, though, see Dekker et al., 2003) explicitly claim a connection between ‘effective governance’ and particular characteristics of the social and po­ litical institutions of a society. Putnam’s Social Capital Theory underscores the importance of in­ formal civic institutions that further and support the development of so­ cial capital defined in terms of trust, involvement, solidarity and coope­ rativeness (Putnam, 1993, 2000). Lijphart’s Consensus Democracy The­ ory emphasizes the importance of formal political institutions that sup­ port consensus building through power sharing and dispersal (Lijphart, 1999). Political and social systems that are characterized by high levels of social capital and a ‘consensual’ political structure seem to be ‘linked’ with higher levels of social as well as economical performance. These findings seem to be highly relevant for the current debates about the future institutional shape of the European Union. Both in the context of the discussions about the European Commission’s White pa­ per on ‘European Governance’ as in the discussions that take place in the European Convention, much attention is paid to the institutional design of the European Union. Such a discussion on the future design of the

256

CHAPTER NINE

European institutions is especially important now that the Union will be enlarged with about 10 new member-states. Without a thorough review of Europe’s political institutions, it is unlikely that it will ever meet its ambitious economic and social policy goals. But, as we will try to sug­ gest in this chapter, a mere institutional approach might not be appropri­ ate, especially when it tends to overestimate the structural dimensions (the ‘hardware’) of Europe’s future social and political institutions. In this context, the link between social capital and consensus demo­ cracy on the one hand and ‘good and effective governance’ on the other hand is highly interesting, but it has been studied extensively already, most notably by Putnam and Lijphart themselves. In the next section (section 2), we will shortly summarize their findings on the relation be­ tween ‘good governance’ and consensus democracy and social capital. Much less is known, however, about the kinds of ‘backgrounds’ these performing political (and social) systems seem to presuppose in terms of the main value orientations that characterize their socio-political envi­ ronments (their ‘software’, so to say). In this chapter, we will present the results of a first, tentative exploration into these orientations. We will try to do so by exploring two possible ‘missing links’ that re­ flect the importance of these ‘background’ dimensions. The first link we look into is the link between consensus democracy as defined by Lijp­ hart and social capital as defined by Putnam (Missing Link I). Do higher levels of consensus democracy go along with higher levels of trustengendering (informal) social networks and institutions? The second link regards the relation between these two on the one hand and value patterns—following from EVS data—on the other hand (Missing Link II). What do the data of EVS teach us about the cultural backgrounds and value orientations that go along with consensualism and social capital? We call these ‘missing links’, because as far as we know these parti­ cular links (between a specific institutional pattern and a specific value orientation) have not been specified yet. Influential studies of (political) culture do, however, suggest that such links are worth exploring. In a more general sense, ideas about a relationship between democra­ cy and political culture were discussed in Almond and Verba’s seminal work on ‘Civic Culture’ (Almond &Verba, 1973). In a comparable way, the relationship between culture (modernization and post-modernization) and democracy gets attention in recent work of Inglehart (1999, 2000). Studies that focus on the interrelation between social capital and consen­ sus democracy on the one hand, and value orientations (and changes therein) on the other hand are, however, scarce1. Therefore, in this chap­ ______________ 1

There are, however, some very interesting suggestions about the relationship be­ tween characteristics of national political cultures (that can be characterized by low and

257

MISSING LINKS?

ter, most attention will be devoted to the exploration of Missing Link I (in section 3) and Missing Link II (in section 4).

Social capital (Civic institutions)

Cultural orientations (Following from EVS)

II

I

Good governance (Performance of democracy)

Consensus democracy (Governmental institutions) Note: This chapter focuses on Missing Link I and Missing Link II, combining central no­ tions of Lijphart (1999), Putnam (1993), and EVS (2000).

Figure 9.1 Schematic depiction of the underlying logic

Figure 9.1 presents a graphic abstraction of the complex, multi-causal reality that we address in this chapter. Almond and Verba have very rightfully characterized the phenomena political culture and political structure as variables in a ‘complex, multidirectional system of causal­ ity’. The lines/arrows in this figure should therefore not be interpreted as simple monocausal relations. We are interested in the interrelation be­ tween consensus democracy as defined by Lijphart and social capital as defined by Putnam, and the interrelation of the two with basic cultural orientations. But before we come to that, let us explore (in section 2) shortly what Lijphart en Putnam have to say about the contribution of ‘consensualism’ and ‘social capital’ to the economic and social perform­ ance of political systems, and—more importantly in this contribution— how they have conceptualized these promising properties of political life and social infrastructure. In the following sections, we will see whether EVS data will enable us to say something about the interrelations (‘links’) we have described above. Can we see a relation between: a. the two characteristics (social capital-breeding informal institutions and consensus-orientedness of ______________

high power distances: low or high PDI) and institutional characteristics of political sys­ tems in Hofstede (2001: 83, 110-113). There seems to be a high similarity between the countries that Lijphart characterizes as consensus democracies and the countries that are characterized by a high PDI by Hofstede (e.g., France). See more specifically: Lijphart, 1999: 288.

258

CHAPTER NINE

formal political institutions) and b. these institutional characteristics and specific value orientations in countries that fit the same profile? 2 Patterns of democracy and social capital: Lijphart, Putnam and the European Values Study Lijphart and Consensus Democracy – Comparing patterns and perfor­ mances of democracy in thirty-six democratic systems, Lijphart (1999) concludes that consensus democracies clearly perform better than their counterparts: non-consensual or majoritarian democracies. He success­ fully challenged the popular assumption that consensus democracies per­ form better in terms of representation, and that majoritarian democracies function better in terms of performance. Lijphart shows that majoritarian democracies do not outperform the consensus democracies when it comes to keeping public order and managing the national economy. In fact, the consensus democracies even have a slightly better record (Lijp­ hart, 1999: 274). The consensus democracies, furthermore, clearly beat the majoritarian democracies with regard to the quality of democracy and democratic representation as well as with regard to what Lijphart calls the ‘kinder and gentler’ qualities of politics and policymaking2. Put differently: consensus democracies perform better on the ‘softer’, more ‘feminine’ side of politics and policy-making, while not performing worse (as has always been assumed) on the ‘harder’, more ‘masculine’ side of politics and policymaking (Lijphart, 1999: 294). In Lijphart’s work, the level of consensualism is measured along two dimensions. The first dimension (L1) is the executives-parties dimen­ sion. It measures the extent to which powers and responsibilities are shared in a given democracy. The degree of consensualism on this di­ mension is determined by five variables. A democracy is more consen­ sual to the extent that it is characterized by: 1. executive power-sharing in broad cabinets/coalitions (versus singleparty majority cabinets); 2. executive-legislative balance of power (versus a dominant execu­ tive); 3. a multiparty system (versus a two-party system); ______________ 2

‘Consensus democracies demonstrate these kinder and gentler qualities in the follow­ ing ways: they are more likely to be welfare states, they have a better record with regard to the protection of the environment, they put fewer people in prison and are less likely to use the death penalty; and the consensus democracies in the developed world are more generous with their economic assistance to the developing nations’ (Lijphart, 1999: 275­ 276).

MISSING LINKS?

259

4. a system of proportional representation (versus a winner-takes-all electoral system); 5. interest group corporatism (versus interest group pluralism). The second dimension (L2) is the federal-unitary dimension. It measures the extent to which powers and responsibilities are divided in a given democracy. The degree of consensualism on this dimension is, again, de­ termined by five variables. A democracy is more consensual to the ex­ tent that it is characterized by: 1. federal and decentralized government (versus unitary and centralized government); 2. strong bicameralism (versus dominance by a single representative chamber); 3. constitutional rigidity (versus constitutional freedom and flexibility); 4. judicial review by a constitutional court (versus a final say of the leg­ islature in constitutional matters); 5. central bank independence (versus dependent central banks). Lijphart’s variables are mainly formal-political or structural-organizational in character. The institutional structure of politics matters: this, in short, is Lijphart’s claim. Consensualism and federalism are keys to a performing political system. In a similar way, Putnam’s work on social capital and associative life, suggests the importance of a particular shape of the more informal insti­ tutions of a democratic society. Putnam and Social Capital – Putnam’s study of Italian regions con­ firms the importance of the institutional structure of politics (in his case: the institutionalization of regional government), but adds an important element too: social capital, which is more on the ‘soft’ or cultural side of politics and social life (and economics too: see Fine, 2001). Re-introduced by Bourdieu (1980) and Coleman (1990)3, this much debated notion of social capital more or less explicitly describes the importance of fami­ lies, social bonds, communities, networks and associations, as well as shared habits that enable individuals to act collectively. Successful col­ lective action requires (generalized) mutual trust and ‘soft’ regulations that exceed the logic of mere instrumental reciprocity (see Putnam, 1993; 2000; Coleman, 1990; Warren, 1999; Fukuyama, 2000; Newton, ______________ 3

Fukuyama (2000) recalls how the notion of Social Capital was first used by Lyda Judson Hanifan in 1916 to describe rural school community centers and was also used in Jane Jacob’s classic work The Death and Life of Great American Cities. See also Putnam (2000: 19).

260

CHAPTER NINE

2001; Fine, 2001). ‘Though these observers of the contemporary era point to different cultural, political, and institutional factors that have changed the relationships between the citizen, the state, the market and the societal midfield, they all share the notion that a flourishing civil so­ ciety and a vital democracy presupposes high levels of social trust, cohe­ sion, and social participation. A vibrant democratic civil society is only believed to prosper if based on high levels of social trust accompanied by support for democracy and political involvement of its citizenry.’ (Dekker et al., 2003). More specifically, the kind of trust that contributes to social capital is trust that can be generalized to people who are strang­ ers, as compared to trust that is particular, limited to one’s family or group (e.g., Warren, 1999, or Putnam, who makes a difference between ‘bonding’ and ‘bridging’ forms of social capital). A society that is characterized by generalized reciprocity is, accor­ ding to Putnam, more efficient than a distrustful society. ‘Frequent inter­ action among a diverse set of people tends to produce a norm of general­ ized reciprocity.’ (Putnam, 2000: 21). The degree of diversity seems to be especially important for the type of social capital that is produced: ‘When economic and political dealing is embedded in dense networks of social interaction, incentives for opportunism and malfeasance are re­ duced. This is why the diamond trade, with its extreme possibilities for fraud, is concentrated within close-knit ethnic enclaves’ (Putnam, 2000: 21). Such a close-knit production of social capital might be favorable for those inside these networks (or ‘communities’), but, as Putnam seems to be well aware of, the external effects of these kinds of ‘warm and cud­ dly’ social capital are by no means always positive (he mentions sectari­ anism, ethnocentrism and corruption). ‘Of all the dimensions along which forms of social capital vary, perhaps the most important is the dis­ tinction between bridging (or inclusive) and bonding (or exclusive).’ (Putnam, 2000: 22). Although Putnam makes clear that bonding and bridging forms of social capital are not ‘either-or’ categories, but ‘more or less’ dimensions’ along which we can compare different forms of so­ cial capital, it is clear that the variables that are used to measure the amount of bridging social capital will differ from those that are used to measure the bonding type. The concept of (notably bridging) social capital is closely related to the recurrent debate on the importance of a strong ‘civil society’ for ‘ma­ king democracies (and economies) work’. Especially Robert Putnam has fuelled this debate in claiming that declining stocks of social capital are related to declining levels of civic organization and association. ‘Stocks of social capital, such as trust, norms, and networks, tend to be selfreinforcing and cumulative. Virtuous circles result in social equilibria

MISSING LINKS?

261

with high levels of cooperation, trust, reciprocity, civic engagement and collective well-being (…). Defection, distrust, shirking, exploitation, iso­ lation, disorder, and stagnation intensify one another in a suffocation mi­ asma of vicious circles. This argument suggests that there may be at least two broad equilibria toward all societies that face problems of col­ lective action (that is all societies) tend to evolve and which, once at­ tained, tend to be self-reinforcing’ (Putnam, 1993: 177). Putnam’s Making Democracy Work (1993) analyses the impact of the regional governments that were set up in Italy in 1970, in fact a major reform of the formal-political structure. The new regional governments appeared to produce remarkable effects in the whole of Italy. But these effects turned out to be considerably more positive in the North than in the South of the country. According to Putnam, the institutional reforms in the North were combined with more favorable social and cultural cir­ cumstances than in the South. This difference could be satisfactory ex­ plained by varying levels of ‘social capital’: trust, involvement, soli­ darity and cooperativeness, and expressed in various forms of voluntary association. Over centuries, northern Italian regions have been able to invest more in the stock of social capital than southern Italian regions. Social capital is crucial in making democracy, and democratic institu­ tions such as regional governments, really work. This goes not only for Italy, Putnam claims, but also for the United States, traditionally known as a ‘civil society’ but now endangered by the decline of social capital il­ lustrated by the metaphor of Bowling Alone (2000). In Putnam’s re­ search, the rise and decline of social capital is highly determined by the magnitude of voluntary association. The more voluntary association, the more social capital, the better the chances of ‘making democracy work’. Lijphart, Putnam and EVS – To define Putnam-groups, based on the European Values Study 1999/2000, we follow Knack and Keefer’s defi­ nition (1997: 1274; see also the chapter of Beugelsdijk and Van Schaik in this volume) for the density of groups. For each country, the average number of groups is computed. Relevant variables in EVS 1999/2000 are those that measure whether people are a member of: religious or church organizations; education, arts, music or cultural activities; youth work, e.g., scouts, guides, youth clubs et cetera. This average number of group membership ranges from .12 in Portugal to 1.04 in Sweden. For our explorative endeavor, we selected those countries from the EVS data for which we had information about the two Lijphart dimen­ sions. Countries that have a high score on Lijphart’s first dimension (Executives-parties dimension, L1) can be characterized by an institutional structure that favors the sharing of power, for instance because of high degrees of interest group corporatism, high degrees of proportional re­

262

CHAPTER NINE

presentation and less dominant executives. The institutional structure of these democracies is more ‘deliberative’ and less ‘majoritarian’. Coun­ tries that have a high score on his second dimension (Federal-unitary dimension, L2) tend to have a political infrastructure in which decisionmaking powers are much more concentrated in a unitary and centrally organized political center. For Lijphart’s executives-parties dimension Great Britain has the lo­ west score (–1.39) and Finland the highest (1.66); for the federal-unitary dimension, Great Britain scores lowest (–1.19) and Germany highest (2.53). In Table 9.1, we present for each country the score on the Putnam-group variable and on both Lijphart dimensions. Table 9.1 Scores on the Putnam-group variables and on Lijphart’s ExecutivesParties and Federal-Unitary dimension for eighteen European countries Country at* Austria be Belgium hr Croatia dk Denmark fi Finland fr France de Germany gb Great Britain gr Greece is Iceland ie Ireland it Italy lu Luxembourg mt Malta nl Netherlands pt Portugal es Spain se Sweden

Putnam Groups .42 .39 .23 .35 .68 .15 .23 .21 .33 .93 .37 .24 .34 .22 .86 .12 .16 1.04

Lijphart’s ExecutivesParties dimension .26 1.42 – .38 1.45 1.66 – .93 .23 – 1.39 – .74 .66 .12 1.16 .29 – .90 1.16 .36 – .59 1.04

Lijphart’s FederalUnitary dimension 1.08 .21 – .44 – .38 – .83 – .17 2.53 – 1.19 – .75 – 1.03 – .42 – .11 – .89 – .39 .35 – .7 .42 – .79

* The at refers, as be and so on does, to the way in which the countries are recognized in the graphs that go with this chapter.

In section 3, this dataset will be connected to the dataset that Lijphart has made public together with ‘Patterns of Democracy’. In section 4, the aforementioned dataset will be used to see whether, and if so, how far and in which way Putnam’s associational groups and Lijphart’s consen­

MISSING LINKS?

263

sus democracies correspond to value patterns suggested by the European 4 Values Study (EVS ). 3 Missing Link I: Consensus democracy and voluntary association The first link that we wannt to explore connects the formal and informal institutional characteristics of democratic societies. Can we indeed find a link between Lijphart’s consensus democracies and Putnam’s voluntary associations and groups? Does consensualism in formal-political terms correlate with ‘associationalism’ in civic-cultural terms? Such a link between consensualism and associationalism is suggested —at a theoretical level—by many so-called neo-Tocquevillian inquiries into contemporary democracy. Authors such as Barber (1994) have drawn attention to ‘strong’ or ‘deep’ forms of democracy, characterized by a mutual reinforcement of strictly political and more informal and so­ cietal repertoires of civic participation. These forms of democracy ex­ clude, sui generis, an all too strong concentration and/or centralization of power. Moreover, they favor consensualism rather than majoritarianism. They suggest the importance of a ‘political opportunity structure’ char­ acterized by shared power, horizontal and vertical subsidiarity, pluralism and (corporatist forms of) deliberation. Moreover, in political systems that are characterized by higher levels of linguistic, ethnic or religious pluralism, many public services tend to be channeled through private, (quasi-)autonomous, non-governmental and non-profit associations. These associations and their umbrella organizations (as intermediary bodies) tend to be highly involved in the public domain. The ‘political opportunity structure’ of these regimes favors a culture of shared power 5 and responsibilities (and indeed: a larger voluntary nonprofit sector) . In short: there are good reasons to believe that there is something like a Wahlverwandtschaft, an elective affinity, between consensualism and associationalism. One could expect a consensual democratic system (characterized by inter alia a more dispersed allocation of power and proportional representation) to provoke and sustain an ‘active society’ and a more communitarian societal ‘infrastructure’. The correlations in Table 9.2 and the scores in Figure 9.2 suggest that such a link actually exists – at least partially. ______________ 4

See the introductory chapter of this volume for additional notes on data. A larger non-profit sector or voluntary sector is associated with higher levels of so­ cial capital, the ability to express a greater variety of value patterns and also with higher levels of innovation and policy learning (thus overcoming some of the negative sides of a dominant state bureaucracy – the bureaucratic constraint). See Douglas (1987). 5

264

CHAPTER NINE

Table 9.2 Correlations between Scores on the Putnam-group variable and Lijphart’s Executives-Parties and Federal-Unitary dimensions for eighteen European countries Putnam Groups Lijphart’s Executives -Parties dimension Lijphart’s Federal­ Unitary dimension

Lijphart’s ExecutivesParties dimension

.56 – .20

.07

Lijphart’s first dimension, the executives-parties dimension, correlates considerably (.56) with associations and groups as defined by Putnam. Figure 9.2 juxtaposes countries such as France and Great Britain (sco­ ring low on associationalism and low on consensualism in terms of power-sharing) to countries such as the Netherlands, Sweden, Finland and Iceland (scoring high on associationalism and high on consensualism according to Lijphart’s first dimension). This finding supports the idea that consensual, power-sharing institutions in Northern European countries present a fair breeding-ground for voluntary associations, which are also dependent upon consent in taking up joint responsibili­ ties. The correlation between Lijphart’s second dimension, the federalunitary dimension, and associations and groups as defined by Putnam is non-significantly negative, that is, virtually non-existent. This is, how­ ever, understandable when taking into account the specific items that produce consensualism on Lijphart’s second dimension: central bank in­ dependence, judicial review, constitutional rigidity, bicameralism, feder­ alization, decentralization. It is not likely that these items correlate with Putnam’s associations and groups. As far as Lijphart’s first dimension and the ‘Putnam-groups’ are con­ cerned, however, the ‘correlation’ we expected on a theoretical level seems to mirrored in the data about the 18 EVS countries we were able to use for our first endeavor. The institutional profiles of the formal and the more informal infrastructures seem to be related to each other: Put­ nam and Lijphart are ‘bedfellows’. Although the data-sets we were able to use and the explorative character of this chapter prevent us from being too conclusive, the analysis clearly supports the importance of doing more extensive and systematic theoretical and empirical research into the ‘link’ between the characteristics of formal and informal institutions of democratic societies.

265

fi dk be nl

1

it

se is

pt 0

de lu at ie hr

es -1

frmt

gr

gb r = .56

-2

Li jphart’s Executives-Parties dimension

2

MISSING LINKS?

0

.2 .4 .6 .8 Density of P utnam Groups

1

Data source: European Values Study 1999/2000

Figure 9.2 Relation between Putnam Groups and Lijphart’s Executives-Parties dimension

4 Missing Link II: Consensus democracies, voluntary associations and cultural orientations Let us now turn to our second question. Are there reasons to believe that consensualism and associationalism—conceptualized as institutional characteristics of social and political systems—are related to deeper, un­ derlying cultural patterns that reflect particular value orientations? In terms of the explorations in this chapter: do the EVS countries that have high scores on Putnam-groups, and on Lijphart’s dimensions of consen­ sualism also have high scores on particular cultural and value orienta­ tions? Both Lijphart and Putnam seem to suggest, though implicitly and, again, at a theoretical level, that such a link might exist. Lijphart (1999: 307), for instance, writes that it is plausible to assume that both consen­ sus democracy and kinder, gentler policies stem from ‘an underlying consensual and communitarian culture’. Putnam too, points at such cul­ tural dimensions in his ‘Agenda for Social Capitalists’ that follows his extensive analysis of all kinds of ‘civic disengagement’ in the United States. Some of the elements of his agenda explicitly address initiatives and ideas that foster a more communitarian attitude and forms of ‘deep

266

CHAPTER NINE

engagement’. Throughout his book, he is more or less explicitly discuss­ ing the link between the ‘civic values’ of American citizens and the in­ stitutions that are—at the same time—(re-)producing them. (Putnam, 2000: 402-414). Those who care about the future of civic institutions should, according to Putnam, care about the civic culture and the values that support it. Data derived from EVS (for the eighteen countries that were also in Lijphart’s survey) seem to support the notion of some specific underly­ ing cultural (or value) orientations carrying both associationalism and consensualism. Below these cultural orientations are discussed under the headings of generalized trust, conditional tolerance and political in­ volvement. We have established these cultural patterns by inductively examining correlations with the Putnam-groups as well as the dimen­ sions of consensualism distinguished by Lijphart. The cultural patterns just mentioned appeared to display noteworthy connections with the Putnam-groups and with Lijphart’s first dimension of consensualism. Again we stress the explorative character of our contribution: we had to rely on secondary analysis, we only used one data set (EVS) and were not able to construct variables on the basis of a systematic and theoreti­ cal exploration of the nature of this second ‘missing link’. However, as we will see below, this handicapped strategy still enables us to conclude that there are good reasons to assume that such a link exists, and de­ serves further research. Moreover Lijphart’s second dimension is the odd man out. The federal dimension is not connected to any of the patterns mentioned, it seems to have a logic of its own, and will therefore be dis­ cussed separately. Generalized Trust – Consensual, power-sharing systems, and systems with strong associationalism, are more likely to flourish in high-trust systems than in low-trust systems. Trust, in this case, is measured by the percentage of people in a country that answer that ‘most people can be trusted’ on a question where they can chose between that answer and the answer: ‘you can’t be too careful.’ In other words, we are talking about the kind of ‘generalized trust’ we discussed above. The correlation of generalized trust with the density of Putnam groups is .73 (see left graph in Figure 9.3). Putnam-groups stand better chances of flourishing when generalized trust is available: for making voluntary ac­ tion worthwhile at least a minimum of generalized trust should be avail­ able. The correlation of ‘trust’ with Lijphart’s first dimension is .64 (see right graph in Figure 9.3). Power-sharing institutions seem to be sup­ ported, at least to some extent, by a culture that transfers the message that most people can be trusted. A minimum of interpersonal trust should

267

MISSING LINKS?

.8

.8

be available for people to get together to see if some sort of consensus could be reached.

se

General ized Tru st .4

es

is

.2

ie at it de be gb lu gr fr mhrt

.2

se dk

pt

0

0

.2 .4 .6 .8 Densi ty o fPut nam Gro ups

es gb gr fr t hr m

fi

is ie date

it

be

lu pt

r = .73 0

nl

.6

fi

nl

General ized Tru st .4

.6

dk

1

r = .64 -2 -1 2 1 0 Li jp hart’s Execu ti ves -Part ies dimens ion

Data source: European Values Study 1999/2000

Figure 9.3 Relation of Generalized Trust with Putnam Groups and Lijphart’s Executives-Parties dimension

Qualified Tolerance – Consensualism (in the sense of Lijphart’s first dimension) and associationalism (in line with Putnam’s conceptualiza­ tion) are associated with a cultural pattern that values freedom and toler­ ance, albeit under certain conditions. Putnam-groups correlate .62 (see left graph in Figure 9.4) and Lijphart’s power-sharing dimension (L1) correlates .46 (see right graph in Figure 9.4) with a bundle of ‘liberties’ that individuals may take without curtailing the choices, or the possibilities, of others. The Y-axis in the graphs in Figure 9.4 (‘qualified tolerance’) is the first dimension of a principal component analysis of questions where people are asked whether the mentioned ‘liberties’ are (never) justified; the most typical items are: homosexuality; divorce; abortion; euthanasia; suicide; having casual sex.

268

1

1

CHAPTER NINE

se

dk

se nl dk .5

fi

0

lu esgb gr at de be

pt hr

gb

fi be it hr

ie

pt

-1

-1

ie

lu at de

gres

-.5

it

is fr 0

Qualified Tolerance

is fr

-.5

Qualified Tolerance

.5

nl

mt 0

.2 .4 .6 .8 Densi ty o f P utnam Gro ups

r = .62

1

mt

r = .46

0 1 2 -2 -1 Lij phart’s Executi ves-P art ies dimensio n

Data source: European Values Study 1999/2000

Figure 9.4 Relation of Qualified Tolerance with Putnam Groups and Lijphart’s Executives-Parties dimension

These findings are, by the way, similar with Putnam’s own findings (see Putnam, 2000: 356 and footnote 10: 496). Putnam writes that ‘social capital and tolerance go together’. They go together in a ‘liberal com­ munitarian’ culture, a culture of building bridges while accepting differ­ ences (‘Salem without witches’). ‘Social Joiners and civic activists are as a rule more tolerant of dissent and unconventional behavior than so­ cial isolates are (…)’; ‘(…) citizens of high-social-capital states are far more tolerant than citizens of low-social-capital states. Far from being incompatible, liberty and fraternity are mutually supportive, and this re­ mains true when we control for other factors like education, income, ur­ banism, and so on. The most tolerant communities in America are pre­ cisely the places with the greatest civic involvement. Conversely, com­ munities whose residents bowl alone are the least tolerant places in America’ (Putnam, 2000: 356). The liberties mentioned reflect an ‘open’ kind of tolerance towards fellow citizens, granting space that does not entrench on the space of others. Such liberties should, however, be distinguished from the ‘liber­ ties’ that do entrench on the choices, and the possibilities, of others. These ‘liberties’ (some would prefer to call them vices) with negative side-effects for others are bundled in a second component (‘vices 1’: that

MISSING LINKS?

269

reflect some ‘egocentric’ and ‘un-concerning’ attitudes towards others): throwing away litter in public places; drunken-driving; joy-riding; spee­ ding; and smoking in public places; and a third component (‘vices 2’: that reflects, just as much, anti-social attitudes): cheating on tax; paying cash for services to avoid taxes; lying in your own interest; claiming state benefits that you are not entitled to; accepting bribes in the course of duty. Both ‘vices 1’ and ‘vices 2’ correlate non-significantly nega­ tively with Putnam-groups as well as Lijphart’s first dimension. Political commitment/involvement – According to Putnam, civic insti­ tutions are institutions that are built upon people who are inclined to be involved (‘people who join are people who trust’). Throughout his book, he discusses the importance of all kinds and repertoires of involvement in the public (political) domain. Public apathy could be viewed as a threat to democracy: an erosion of civil society, which would rob demo­ cracy of its social base (e.g., Dahrendorf in Pharr & Putnam, 2000). In terms of EVS, political involvement is associated with a positive attitude towards all kinds of participation, which can be expressed in a series of more or less ‘political’ activities: signing a petition; joining in boycotts; attending lawful demonstrations; joining unofficial strikes; occupying buildings or factories. The degrees in which people value the importance of these forms of political involvement correlates .52 with Putnam groups and only .20 with Lijphart’s first dimension (see graphs in Figure 9.5). The odd man out – Lijphart’s second dimension, the federal/unitary dimension, is the odd man out. This dimension does not display signifi­ cant correlations with the (sets of) variables discussed above. Lijphart’s second dimension does however display correlations with two other (sets of) variables that are worth mentioning. First, Lijphart’s federal/unitary dimension is clearly associated with aversion to right-wing political extremism6 (.69; see left graph in Figure 9.6) as well as left-wing political activism (.62; see right graph in Figure 9.6). This is explainable, taking into account that federalized, power di­ viding, systems have often been developed in countries where centrifu­ gal tendencies should be feared.

______________ 6

Such countries often have a history of political violence by political minorities, pro­ ducing an aversion to political extremism among members of the majority, as well as a need for federalized, power-spreading, institutions to organize a tenable level of unity in pluriformity.

270

.6

.6

CHAPTER NINE

se

nl

gb lube it ie

0

fr

.3

is fi

ptes

is nl dk

gb

lu

it

be fi

ie de pt at

es

at

r = .52 .2 .4 .6 .8 Densi ty o f P utnam Gro ups

-.6

mt

-.6

mt

0

hr

-.3

de

gr

0

gr

Political Commitment

hr

dk

-.3

Political Commitment

.3

fr

se

1

r = .20

0 1 2 -2 -1 Lij phart’s Executi ves-P art ies dimensio n

Data source: European Values Study 1999/2000

at lu se

be mt

fr

hr ie it g bi s g r

.8 Avers io n of Left Wi ng Pol it ical Ext remi sm .2 .4 .6

de nl

es

fi dk pt

r = .69

0

0

Avers io n of R ight Wi ng Pol it ical Ext remi sm .2 .4 .6 .8

Figure 9.5 Relation of Political Commitment with Putnam Groups and Lijphart’s Executives-Parties dimension

-1 .5 -1 -.5 0 .5 1 1.5 2 2.5 Li jp hart’s Federal -Uni t ary di mensi on

nl

at

de

lu mt hr ie

be it fr

gb is sger

es

fipt dk r = .62 -1 .5 -1 -.5 0 .5 1 1.5 2 2.5 Li jp hart’s Federal -Uni tary di mensi on

Data source: European Values Study 1999 /2000

Figure 9.6 Relation of Aversion of Right and Left Wing Political Extremism with Lijphart’s Federal-Unitary dimension

271

MISSING LINKS?

1

.4

Second, Lijphart’s federal/unitary dimension is clearly correlated (.73; see left graph in Figure 9.7 with the tendency of people to say ‘I belong first of all to my region’ (instead of to larger entities like their country or Europe). People living in systems scoring higher on Lijphart’s second dimension are less likely to say that they are ‘very proud to be a citizen’ of their country than people living in states scoring lower on Lijphart’s federal/ unitary dimension7.

es

pt fr fi it lugr se hr nl is

.1

gb

r = .73

.9 .8

gb

dk fr it gr luse hr

at es

nl be

.7

ie

be

Proud to be a citizen of one’s own state

.2

dk

de

fi

de .6

.3

at

0

Belong first of all to my Region

ie is pt mt

-1.5 -1 -.5 0 .5 1 1.5 2 2.5 Lij phart’s Federal-Unitary dimension

r = -.69 -1.5 -1 -.5 0 .5 1 1.5 2 2.5 Li j phart’s Federal-Unitary dimension

Data source: European Values Study 1999/2000

Figure 9.7 Relation of Belonging first of all to my own region and Proud to be a citizen of one’s own state with Lijphart’s Federal-Unitary dimension

This variation in appreciation (in the sense of belonging to one’s region, one’s country and the European public domain) is highly relevant to the discussions about the upcoming expansion of the European Union. But more importantly, the value patterns that express attitudes like trust, tolerance and involvement in the public sphere, seem to be espe­ cially important. Our preliminary analysis suggests that those who are designing Europe’s future institutions cannot restrict themselves to a mere technical and administrative analysis. Those who care about the social and economical performances of a future European Union must also see to European values, since some values are more than others do, ______________ 7

There is a negative correlation of .69 between Lijphart’s second dimension and the item ‘proud to be a citizen of (one’s own national state)’ – see right graph Figure 9.7. Fe­ deralized systems, giving ample room to regional identity and decision making, seem to be a logical answer to cultural patterns that prefer the regional to the national.

272

CHAPTER NINE

supporting the legitimacy and efficacy of its institutions. It is, in this re­ spect, somewhat disappointing to see that an explicit reflection on Euro­ pean Values gets only a marginal role on the European agenda. 5 Discussion and conclusion: The values of European democracy ‘Cultural factors play a crucial role in the long-term survival of democ­ ratic institutions. Though the transition to democracy may be initiated by elites (or imposed by foreigners), its long-term survival is linked with relatively high levels of subjective well-being and interpersonal trust. (….) In the long run, democracy is not attained simply by making insti­ tutional changes or through clever-elite maneuvering. It survival also depends on what ordinary people think and feel.’ (Inglehart, 1999: 118– 119; see also Putnam, 2000: 414). The European Values Study was designed to empirically explore ex­ actly what Inglehart thinks is important for the future of democratic in­ stitutions: that what ordinary people think and feel. In this chapter, we have used some of the data EVS has produced to explore whether we can be somewhat more specific about the complex interrelationship be­ tween the values and the institutions of democracy. We did so because we are convinced that the study of European values could successfully be linked to emerging research agenda’s in the field of the performance of European institutions. We have, more specifically, explored some of the links between EVS and some other bodies of thinking about the nature and future of democ­ racy. Although our analysis is limited, it did show that there are good reasons to believe that those who intend to build an effective and per­ forming European Union cannot restrict themselves to a merely techni­ cal design of its institutions and the broadening of the administrative and political infrastructures. Although this is, of course, much more difficult, a ‘deepening’ of the Union, by explicitly putting the European values on the agenda, is at least as necessary in order to ensure that the European institutions perform well. It appears that the current debates on the future of Europe (that especially deal with the importance of strong bureaucra­ cies and competitive markets) have a blind spot for the specific links be­ tween the values and the related social institutions that seem to contrib­ ute to the performances of democratic and social systems. The ‘missing links’ we have tried to ‘reveal’ in this chapter suggest that it is very worthwhile to engage in further research into the values that support democracy. Changes in European values should be revealed in order to be able to design policies. Of course it will be difficult to establish whether the high levels of generalized trust and ‘qualified tolerance’ we

MISSING LINKS?

273

were able to relate to performing and consensual democracies must be seen as cause or effect. But in both instances, it is worthwhile to investi­ gate and monitor these patterns and values in order to prevent us from neglecting the social qualities and implications of our political institu­ tions. And, furthermore, these investigations and monitors will enhance our capabilities do develop a more elaborated theory about the relation­ ships between values and democracy. During a seminar devoted to the 8 future of political democracy , Hans-Dieter Klingemann has convinc­ ingly made the point that we do indeed need more specific theories about the relation between values, political culture and democracy. We are no longer satisfied with theories that look to explain the generic relation­ ships between democracy and a certain set of cultural values, but need to go further in looking for more specific relationships between different value-patterns and different types of democracy. Although there is no reason to think in terms of value-determinism, there is no reason, either, to think of value-relativism. The future of European Democracy is as much as it depends on institutions, depending on the values that support these institutions. It is in this context, that EVS might contribute to an important new research agenda. Our modest first step towards such an agenda has only tried to make clear that EVS can be helpful in designing some more specific hypothe­ ses about the relation between values, institutions and democracy. There seems to be a kind of elective affinity between the consensual nature of the democratic institutions and the polyarchical nature of the broader so­ cial context. Strong democracies seem to go along with substantial levels of civic engagement. Both of them seem to flourish on a fertile soil of specific value patterns: patterns that can be characterized by high levels of (mutual) trust, qualified tolerance and political involvement. A further development of a democratic and effective European Union seems to re­ quire, as much as it needs the further development of institutions, a (po­ litical) culture that expresses these values. One could infer from these data, that mutual trust, more than treaties, lies at the basis of the further deployment of European democracy. This should inspire European lead­ ers to think of the notion of a further ‘expansion’ of the European Union beyond the mere territorial and technocratic dimensions that have domi­ nated the recent discussions about Europe’s future. A successful strategy of a broadening of the European Union seems to presuppose, more than ever, a ‘deepening’ of the Union by explicitly discussing the cultures and values that go along with strong democracies: Trust, Tolerance and In­ volvement. ______________ 8

Tilburg University, 20 November 2002.

274

CHAPTER NINE

References Almond, G.A. & S. Verba 1963. The Civic Culture. Princeton: Princeton University Press. Barber, B.R. 1984. Strong Democracy. Berkeley: University of California Press. Beer, S., A.B. Ulam, S. Berger & G. Goldman 1973. Patterns of Government. The Major Po­ litical Systems of Europe (third edition). Random House: New York. Beugelsdijk, S. & A.B.T.M. van Schaik (2003). Participation in Civil Society and European Regional Economic Growth. Chapter five in this volume. Bourdieu, P. 1980. Le Capital Social: Notes Provisoires. Acted de la Recherche en Sciences So­ ciales 31: 2–3. Coleman, J.S. 1990. Foundations of Social Theory. Cambridge: Harvard University Press. Dekker, P., P. Ester & H. Vinken (2003). Civil Society, Social Trust and Democratic Involve­ ment. Chapter eight in this volume. Douglas, J.A.T. 1987. Political Theories of Nonprofit Organization. Pp. 43–54 in W.W. Powell (ed.), The Nonprofit Sector. A Research Handbook. New Haven: Yale University Press. Fine, B. 2001. Social Capital versus Social Theory. Political Economy and Social Science at the Turn of the Millennium. London/New York: Routledge. Hofstede, G. 2001. Culture’s Consequences. Comparing Values, Behaviors, Institutions and Organizations Across Nations. London/New Delhi: Sage: Thousand Oaks. Inglehart, R. 1999. Trust, Well-being and Democracy. Pp. 88–120 in M.E. Warren (ed.), De­ mocracy and Trust: Cambridge: Cambridge University Press. —— 2000. Culture and Democracy. Pp. 80–97 in L.E. Harrison & S.P. Huntington (eds.), Cul­ ture Matters. How Values Shape Human Progress. New York: Basic Books. Fukuyama. F. 2000. Social Capital. Pp. 98–111 in L.E. Harrison, S.P. Huntington (eds.), Cul­ ture Matters. How Values Shape Human Progress. New York: Basic Books. Knack, S. & P. Keefer 1997. Does Social Capital Have an Economic Pay-off? A Cross Country Investigation. Quarterly Journal of Economics 112: 1251–1288. Lijphart, A. 1999. Patterns of Democracy. Government Forms and Performance in Thirty-Six Countries. New Haven: Yale University Press. Newton, K. 2001. Social Capital and Democracy. Pp. 225–234 in B. Edwards, M.W. Foley & M. Diani (eds.), Beyond Tocqueville. Civil Society and the Social Capital Debate in Comparative Perspective. Hanover: University Press of New England. Pharr, S., R.D. Putnam (eds.) (2000). Disaffected Democracies. What’s Troubling the Trilateral Countries? Princeton: Princeton University Press. Putnam, R.D., R. Leonardi & R.Y. Nanetti 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press. —— 2000. Bowling Alone: the Collapse and Revival of American Community. New York: Simon & Schuster. Warren, M.E. (ed.) 1999. Democracy and Trust. Cambridge: Cambridge University Press. Salamon, L.M., H.K. Anheier, R. List, S. Toepler, S. Wojciech & Associates 1999. Global Civil Society. Dimensions of the Nonprofit Sector. Baltimore: The Johns Hopkins Center for Civil Society Studies.

CHAPTER TEN

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE? WIL ARTS, LOEK HALMAN & WIM VAN OORSCHOT 1 Introduction According to some, there is a story to be told in contemporary Western Europe of diminishing social capital and decreasing informal solidarity and civic morality. This story does have a villain, and the villain is truly detestable: it is the welfare state. According to others the welfare state is sooner the hero of this story. Wherever social capital, informal solidarity and civic morality are still in good condition it is thanks to the welfare state. This difference of opinion between welfare state pessimists and opti­ mists has all the characteristics of earlier disputes about alleged moral decline. History abounds in descriptions of times and places in which some people were lamenting the loss of public spirit and the decay of morals, while others shrugged their shoulders as to so much lack of un­ derstanding of the spirit of the age. Today we can easily conclude, with the benefit of hindsight, that sometimes the complaints of the pessimists were correct and sometimes they were wide off the mark. In the past also the optimists have been now and then in the right and sometimes in the wrong. So, invoking the help of history does not help us settling the dispute. Glancing rapidly through the literature does not help us either. For, at least at first sight, it looks as if the dispute produced more heat than light. Both sides in the dispute have an inclination to claim the moral high ground and to preach for the converted. What we want to do in this chapter is to stay aloof from the ideological positions in the dispute and choose a more sophisticated approach instead. We will treat the stand­ points of welfare state optimists and pessimists as theoretical conjectures that can be tested empirically. By doing so we will try to shed new light on the dispute. What we more specifically want to do in this chapter is to reconstruct first of all the debate on the ‘moral evil effects of the welfare state’ the­ sis. At the center of this debate is the question of whether the welfare state has had unintended adverse effects on the amount of social capital, the level of civic morality and the degree of informal solidarity specific

276

CHAPTER TEN

to contemporary Western European societies. Next we will use this theo­ retical reconstruction as a heuristic tool for formulating informative hy­ potheses. Then we will put these hypotheses to the empirical test by ana­ lyzing data from the 1999/2000 wave of the European Values Study sur­ vey. We will limit the countries to be analyzed to the member states of the European Union. Next, we will explore the personal and contextual characteristics that influence people’s informal solidarity and civic mo­ rality. Finally, we will conclude whether the welfare state is indeed the villain or sooner the hero of the piece or that the relation between wel­ fare statism, social capital, civic morality, and informal solidarity is a much more complex one than both sides in the dispute realized. 2 The ‘moral evil effects of the welfare state’ debate The history of the development of European welfare states is a long and complicated one (Flora & Heidenheimer, 1981; Mommsen, 1981; De Swaan, 1988). The take-off of the welfare state as a type of society sui generis occurred as early as the end of the nineteenth century. It took, however, till the middle of the twentieth century before the golden age of the welfare state was suddenly upon us. Its heyday lasted for some­ what more than two decades. What followed thereafter was an appar­ ently never-ending crisis that gave rise to recesses of welfare state re­ trenchment. Boje (1996) has argued that in the heyday of European wel­ fare states during the 1960s and 1970s civil society was gradually shut out of welfare provision because of strong and comprehensive welfare state intervention. Welfare state institutions took over in large numbers the obligations of support previously located in civil society and family networks. Welfare state ideologues and politicians were, however, ai­ ming too high at that time. Welfare state institutions appeared to cost the earth and proved to be uncontrollable at the governmental and adminis­ trative level. A crisis ensued. By the end of the 1980s a new shift took place, but now in the balance between the market and the state. Welfare state retrenchment took place by means of the privatization and marketi­ zation of public services. The state retreated gradually from welfare pro­ vision. Boje argues that the political efforts to de-commodificate indivi­ duals via welfare state intervention in the 1960s and 1970s and to re­ commodificate them by strengthening the market mechanism in the late 1980s and the 1990s has to some extent led to a destruction of commu­ nity and family networks and the isolation of individuals. Also according to others, neo-liberal social policy-making threatens social solidarity and feelings of mutual responsibility, e.g., because market-oriented provision runs the risk of exhausting the moral legacy of citizenship and trust on

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

277

which they depend (Goodin, 1996; LeGrand; 1997; Taylor-Gooby, 1998). For others, however, only the welfare state itself is the real villain of the piece. Wolfe (1989) for example not only argues that comprehensive welfare states have squeezed families, communities, and social net­ works. He also claims that an historical irony may exist: when social ob­ ligations become public, intimate ties will weaken and so will distant ones, thus undermining the very moral strength the welfare state has shown earlier on. Zijderveld (1983, 1998) too is of the opinion that wel­ fare state development has hollowed out and eroded intermediate social structures. As a result, he claims, commitment and trust declined. Just like Spieker (1986), he contends that the original welfare state ethic made place for an immoralistic ethos of consumerism and people’s be­ havior became guided by selfishness and even hedonism. Fukuyama (2000) has advanced a more or less similar thesis. He is of the opinion that states can have a serious negative impact on the social capital of their population when they start to undertake activities that are better left to the private sector or to civil society. The ability to cooperate, he ar­ gues, is based on habit and practice. If the state gets into business of or­ ganizing everything, people will become dependent on it and lose their spontaneous ability to work with one another. According to De Swaan (1988) the modern welfare state has eroded interpersonal feelings of solidarity, and has replaced individually felt responsibility with the idea that it is the state that has to care. Finally, also Schuyt (1995) acknowl­ edges that the developed welfare state has severe unintended social and moral consequences. He refers among others to the withering away of care obligations for family members and friends, to a growing misuse of welfare state benefits and services, and the genesis of an immoralistic welfare ethic, in which hedonism and selfishness have become dominant attitudes. Taking these ideas and opinions together, two different lines of argu­ ment seem to show up. One line argues that welfare state provision is ‘crowding out’ civil society institutions, informal networks and sponta­ neous activities, by making them superfluous. And a second line of ar­ gument stressing that the welfare state has an effect of ‘moral corrup­ tion’ on citizens, who lose feelings of mutual responsibility and solidar­ ity, and become selfish, hedonistic, and uncaring. The problem with these two lines of argument is that they are mainly based on theoretical conjectures, anecdotal evidence and ad hoc interpre­ tations. The proponents of the ‘evil effects’ thesis, i.e., the welfare state pessimists, have hardly ever tried to empirically test their thesis. It will therefore not come as a surprise that the opponents have criticized the

278

CHAPTER TEN

thesis on empirical grounds. One argument against the crowding out ef­ fect to be found in the literature is that in Scandinavia the development of the encompassing welfare states has been attended by an expansion of civil society. Kuhnle and Alestalo (2000) for example argue that in Scandinavia a positive relationship exists between welfare state devel­ opment and market and third sector (civil society) welfare provision. According to Kuhnle and Selle (1990) in Scandinavia voluntary associa­ tions have often actively been searching for greater state involvement and have fully accepted to play a structural role in the overall set up of the national welfare state. Rothstein (2001) argues that research findings show that the Swedish welfare state has not eroded people’s activities in civil society and voluntary organizations. On the contrary, such activities have grown slightly in the course of the last few decades. Nor could a detrimental effect on informal social networks during this period be es­ tablished. The data shows that there has even been a strengthening of in­ formal social ties during this period. The ‘moral corruption’ line of argument too has been criticized from an empirical point of view. Review studies for example show that (higher) unemployment benefits do not corrupt the work ethic (JehoelGijsbers et al., 1995; Bryson, 1997; Gallie & Alm, 2000), nor do they inhibit unemployed people to look for a job and to accept one if offered (Atkinson, 1989; Barr, 1992); social expenditures targeted at elderly people do not undermine intra-family and inter-generational solidaristic feelings and behavior (Kohli, 1999; Attias-Donfut & Arber, 2000; Knijn, 2002); as said, voluntary work is not less in well-developed welfare states than in others, on the contrary (Kuhnle & Selle, 1990; Rothstein, 2001); and, in a well-developed welfare state as the Netherlands repeated surveys show that over a time-period from 1960 onwards no consumeris­ tic ethic has developed (Ester & Halman, 1994), nor have citizens be­ come more ‘calculating’ and selfish, nor have pro-social attitudes, trust in other people and supportive behavior in families and neighborhoods declined (Dekker & De Hart, 2000). Advocates of the ‘moral evil effects’ thesis could explain these falsi­ fications away by arguing that they offer no overall test of their thesis, because they refer only to particular countries or welfare state regimes. And even if they are considered to be proper tests they only refute the thesis of the pessimists, but do not prove that the optimists’ thesis is cor­ rect. So, a more comprehensive and systematic test is needed. Before proceeding with such a test we must first, however, derive informative hypotheses from the ‘theories’ that provided the opinions and arguments of the pessimists and optimists.

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

279

3 Hypotheses Although there is in Western Europe no such thing as the welfare state, this does not mean that welfare states are unique cases. Most students of welfare statism assume that welfare states come in types because they have developed along more or less similar lines in several countries. Those types vary considerably with regard to their accent on and their comprehensiveness of welfare provisions and their inclusiveness of so­ cial citizenship. According to Esping-Andersen (1990) individual speci­ mens cluster together in three highly distinct regime-types, each organ­ ized around its own discrete logic of organization, stratification and so­ cial integration. Each of the three types, he identifies with a deep tradi­ tion in political mobilization and political philosophy (conservatism, li­ beralism, and socialism respectively). These traditions have a link to par­ ticular features of contemporary social policy and socio-economic con­ figurations. Most other students of the welfare state fall in with EspingAndersen’s opinion that in modern advanced capitalism welfare states are not unique, but they disagree with respect to the number of distinc­ tive worlds of welfare capitalism. Arts and Gelissen (2002) have recon­ structed several welfare state typologies and have concluded that they can be merged in six ‘new-cum-old’ ideal types. Following in their foot­ steps we will assign the real welfare states of Western Europe to four ideal types. The first type is the liberal or Anglo-Saxon welfare state. The second one the conservative-corporatist or Continental welfare state. The third one the social-democratic or Scandinavian welfare state. And the last one the budding or Southern European welfare state. There are natural affinities between specific welfare state regimes and the pillars (state, market, community, and family) that support these re­ gimes (Kuhnle & Alestalo, 2000; Goodin & Rein, 2001). Some regimes sit more comfortably at top of particular pillars than others. The liberal welfare state regime has a ‘natural’ affinity with the market and the fam­ ily; the conservative-corporatist regime with family, civil society and the state; the social-democratic one with the state; and, the budding regime type with the family and civil society. The advocates of the ‘evil effects’ thesis defend a substitution hypothesis concerning the combined effects of regimes and pillar on civic morality and informal solidarity and the opponents a communicating vessels hypothesis (Komter et al., 2000). The (pessimistic) substitution hypothesis assumes that formal solidarity and morality, organized and enforced by the state, and informal solidar­ ity and morality substitute each other. The growth of the first crowds the other out or makes it at least no longer necessary. The (optimistic) com­ municating vessels hypothesis assumes, on the other hand, that if formal

280

CHAPTER TEN

solidarity and morality grow informal solidarity and morality do or at least can also increase, and if one declines the other is also threatened. Both hypotheses assume not only a direct, but also an indirect effect of welfare statism on civic morality and informal solidarity through the a­ mount of social capital. The hypotheses have in common that they ex­ pect a positive correlation between social capital on the one hand and civic morality and informal solidarity on the other. They, however, disa­ gree with regard to the answer to the question of whether the degree of welfare statism has a positive or a negative impact on social capital, ci­ vic morality, and informal solidarity. The substitution hypothesis expects a negative impact whereas the communicating vessels hypothesis ex­ pects a positive effect. Combining the two hypotheses with the welfare state typology would lead to the following predictions. The substitution hypothesis predicts that where state organized and enforced formal solidarity and morality will be strongest, as in the Scandinavian, social-democratic welfare states, social capital will be relatively low and civic morality and infor­ mal solidarity relatively weak. In the Anglo-Saxon, liberal and Southern European, budding welfare states organized solidarity is weakest and therefore social capital will be relatively high and civic morality and in­ formal solidarity relatively strong. The Continental, conservativecorporatist welfare states will take up a position in the middle. The communicating vessels hypothesis leads to predictions in the opposite direction. The problem with such very crude hypotheses is, however, that wel­ fare state theoreticians always affix a provision to them. They are as­ sumed to hold only under the condition of a ceteris paribus clause. As handy as the phrase ‘ceteris paribus’ may seem to be, to append it to a hypothesis constitutes an admission of defeat. Empirically, there is al­ ways a considerable chance that confounding variables will disturb the presumed causal relation. Therefore we should have to control for some contributory variables and to make informed conjectures about the con­ founding influence of others. However, the number of countries per type is very low (three to five depending on the type), which makes the con­ trolling for various country characteristics impossible. Another point, which can be raised regarding the effects of welfare state type, is that some authors (Goodin & Rein, 2001) observe in today’s world of wel­ fare capitalism a blurring of the demarcation that is supposed to differen­ tiate regime types and an equally pervasive mixing of pillars. There is in their opinion a development towards proliferation of hybrids discernible, which would mean that the regime types distinguished above will only very weakly, if at all, differentiate between social capital, civic morality

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

281

and informal solidarity. This would also be the case if others, like Kasza (2002), are right. They are of the opinion that few national welfare sys­ tems are likely to exhibit the internal consistence necessary to validate the regime concept, and that policy-specific comparisons may be a more promising avenue for comparative research. We could find a way around these problems by including as independent variables not only ordered regime types in our empirical analyses, but also numerical measures of welfare state effort, i.e., of the extensiveness and intensity of state inter­ vention (Gelissen, 2002). This has the additional advantage that the sub­ stitution and communicating vessels hypotheses need no longer be stated in rather crude terms of ordinal differences between groups of countries, but can be more precisely stated in terms of differences between numeri­ cally ordered separate countries. Our assumptions and expectations re­ garding aggregate-level effects can be summarily reproduced as in Fi­ gure 10.1.

+ or –

Civic morality

Welfare state: type / effort

+ or –

Social capital

+ + Social solidarity

+ or – Note: + or – = + for the communicating vessels hypothesis and – for the substitution hy­ pothesis

Figure 10.1 Hypothetical influences of welfare state type, c.q. welfare state effort, and aggregate social capital on the aggregate level of civic morality and informal social solidarity

282

CHAPTER TEN

Apart from differences at the collective level of welfare state types and countries there may be systematic variability among the citizens of coun­ tries in terms of their civic morality and informal solidarity. We will therefore also take the opportunity to explore in a second step the differ­ ences at the individual level. We will do this with reference to their be­ longing to different (sub)cultures, and their being differently located in social structures (Therborn, 1991). This means that we will combine in our analysis cultural and structural characteristics of individuals, as well as welfare regime and country characteristics. The relations we will ana­ lyze are as in Figure 10.2.

Structural and cultural personal characteristics

Civic morality

Structural and cultural context variables

Social solidarity

Figure 10.2 Hypothetical influences of contextual and individual characteristics on people’s informal civic morality and social solidarity

4 Data and measurements For both the aggregate and the individual level analyses we will use data from the 1999/2000 wave of the European Values Study survey. A much better test of the competing hypotheses would require longitudinal analyses of not only data from the 1999/2000, but also from previous EVS waves. However, the data of these earlier waves contain only very

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

283

imperfect measurements of social capital and no measurement at all of informal solidarity. Although Western Europe contains more countries than the European Union we will limit ourselves, as mentioned before, to member states of the EU1. In the 1999/2000 surveys of EVS all member states were in­ cluded. 4.1 Aggregate level analyses The crucial independent variable here is the type of welfare state regime a certain country has developed. Based on the typology discussion by Arts and Gelissen (2002), we will label the countries using the EspingAndersen typology as modified by Ferrera (1996) and analyze them as instances of different welfare state types. The United Kingdom (Great Britain and Northern Ireland) and Ireland will be treated as Anglo-Saxon welfare states. France, Germany, Austria, Belgium, and the Netherlands will be conceived off as Continental conservative welfare states. Den­ mark, Sweden and Finland will be labeled as Scandinavian, socialdemocratic welfare states. Greece, Portugal, Spain, and Italy will be called Southern European, budding welfare states. As mentioned before, we will not only use regime type as an inde­ pendent variable, but also welfare state effort. This is measured per country by the average of its public social expenditures as part of GDP over the period 1990-1998 based on Eurostat Yearbook 2002 figures. In the ‘evil effects’ thesis debate social capital is usually conceived of as an intervening variable. The problem is that the literature on social capital reveals a wide variety of definitions and measurements. Flap (1999) even comes to the conclusion that measurements of social capital are made in rather ad hoc, pragmatic and unsystematic ways. For our purposes, however, there is a quite satisfying measurement model avail­ able that contains different indicators. This model has been developed by Paxton (1999) to map the development of social capital in the USA by using data from the General Social Surveys. Janssen (2000) has adjusted this model in such a way that it can be used to analyze EVS data. The ______________ 1

We have chosen only EU member states, with exclusion of Luxembourg that is much smaller than the other ones, for both substantial and pragmatic reasons. The substantial reason is that EU members have at a transnational level an institutional and social policymaking framework in common, whereas this is not the case for the Western European non-members. Differences between members and non-members could be the consequence of this framework instead of type of welfare state regime. The pragmatic reasons are first­ ly that in Norway and Switzerland the EVS survey was not fielded in 1999/2000. Second­ ly, data for Malta and Iceland are available, but both these countries are relatively isolated islands with a tiny population at the fringe of Western Europe. It does not make much sense to include them as rather unique cases in our analyses.

284

CHAPTER TEN

three main components are trust in other people, trust in institutions, and participation in civil society. For the first of the three components of social capital, trust in other people, EVS has only one indicator. People were asked to tell on a 2­ points scale whether they would, generally speaking, say that most peo­ ple can be trusted or that you cannot be too careful in dealing with peo­ ple. The second component, trust in institutions, will be measured by us­ ing four indicators. People were asked to look at a card with institutions and tell, on a 4-points scale, how much confidence they had in them. The institutions we will include in our analyses are parliament, civil service, the social security system, and the health care system. The third compo­ nent, participation in society, is measured by using three indicators. The first one concerns belonging to and doing unpaid work for voluntary or­ ganizations. Janssen (2000) constructed a 3-points scale using an exten­ sive list of fourteen voluntary organizations. The other two indicators concern times spend with friends (4-points scale) and time spend with people in clubs or voluntary organizations (4-points scale). The social capital variable used consists of factor scores resulting from factoring the three components. The dependent variables are informal social solidarity and civic mo­ rality. In the context of welfare statism informal social solidarity refers to the sentiments of concern of people with respect to the basic target groups of welfare policies, i.e., the socio-economically weaker and (po­ tentially) neediest social groups (Van Oorschot, 2000; Van Oorschot et al., 2001). In the EVS 1999/2000 wave people were asked to tell on a five-points scale to what extent they were concerned about the living conditions of four welfare target groups in their country: the elderly, the unemployed, immigrants, and sick and disabled people. Answers regar­ ding all four groups correlate positively, but especially strong regarding the elderly and the sick and disabled (Pearson r = .67). For this reason we have taken the two latter groups together to form one group of ‘un­ abled’, next to the ‘unemployed’ and ‘immigrants’ groups. Civic morality will me measured by looking at people’s answers on a 10-points scale to the question of whether, and if so to what extent the actions of people who do not discharge their duties as citizens can be justified. Four indicators will be used: 1) claiming state benefits which one is not entitled to, 2) cheating on tax if one has the chance, 3) accep­ ting a bribe in the course of one’s duties, and 4) paying cash for services to avoid taxes. The scale-variable of civic morality results from a factor analysis on these four indicators.

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

285

4.2 Individual level analyses As said, in order to explore factors that influence people’s informal soli­ darity and civic morality we will relate the dependent variables men­ tioned above on the individual level to two sets of independent variables: a set of individual characteristics and a set of context variables. To the first belong age, household income, educational level, reli­ gious denomination, church attendance, political left-right preference, ethnocentrism (added in case of solidarity with the ‘immigrants’), and whether one is unemployed or not (added in case of solidarity with the unemployed). Age is measured in years passed since birth; level of edu­ cation is measured by age when education completed; household income is measured by a self-rating in the deciles categories of a net household income scale; unemployment is measured by the question whether peo­ ple were unemployed or not; political stance is measured through selfplacement on a 10-point left-right scale; religious denomination is meas­ ured as whether people do or do not regard themselves as Catholics; church attendance is measured in terms of frequency of attendance per week, month, year of even longer; ethnocentrism is measured as the de­ gree to which respondents do not accept other ethnic groups (immi­ grants, Muslims, Jews and gypsies) as their neighbors; the negative view of others is tapped by the respondents opinion that compatriots claim state benefits illegally, cheat on taxes and pay cash for services to avoid taxes. The set of context variables consists of the country’s welfare effort (the total of social expenditure as a percentage of GDP), as well as the following composition variables at country level (all based on respon­ dents reports): the average age; the average household income; the aver­ age level of education; the average on the left-right scale; the percentage of Catholics and frequency of church attendance; the average level of social capital. To this are added a country’s expenditures on old age cash benefits, sickness benefits, and disability benefits as percentages of GDP for the analysis of the informal solidarity with ‘unabled people’ (elderly, sick and disabled); expenditure on unemployment benefits for the analy­ sis of solidarity with the unemployed; the percentage of immigrants in a country (based on Eurostat 2000, European Social Statistics) and a country’s average score on reported ethnocentrism for the analysis of solidar­ ity with immigrants; the country’s average negative view on others and the country’s level of corruption as it appears from the transparency in­ dex, which latter is a measure of the degree of corruption in a country as perceived by businessmen (www.transparency.org).

286

CHAPTER TEN

5 Analysis and results

5.1 Aggregate level analysis Informal solidarity We start the analyses with an overview of informal solidarity as a de­ pendent variable. In Figure 10.3 we have displayed the countries’ mean scores regarding the informal solidarity with the three groups of unabled, unemployed and immigrants. The countries are grouped according to type of welfare state. 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 pt

es

it

gr

ie

gb

unabled

ni

be

unemployed

nl

fr

de

at

dk

fi

se

immigrants

Unabled: eta = .270; F = 83.049; p = .000. Unemployed: eta = .264; F = 79.333; p = .000. Immigrants: eta = .229; F = 58.465; p = .000.

Figure 10.3 Country means on concern with the unabled, the unemployed, and immigrants

The general pattern is that solidarity with the ‘unabled’ is highest in all European countries, followed by solidarity with the ‘unemployed’, which in turn is higher than people’s solidarity with the ‘immigrants’. This rank order does not come as a surprise, since it is fully in line with Coughlin’s (1979) ‘universal dimension’ of support for welfare expendi­

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

287

tures targeted at various groups of needy people. In a comparison of public opinion data from eight Western European countries Coughlin found as a recurrent pattern over time that the legitimacy of expenditures targeted at the elderly is usually highest among the public, closely fol­ lowed by the support for expenditures on the sick and disabled, and needy families, while support for unemployment benefits and services is usually (much) lower. His finding points to the existence of rather fun­ damental deservingness criteria that are applied by people when solving problems of distributing scarce resources in a welfare context. De Swaan (1988) has referred to three criteria that people have used in the course of time to distinguish the deserving from the dangerous poor: disability (capable vs. incapable), residence (resident vs. alien), and docility (law abiding vs. rebellious). Van Oorschot (2000) refers to criteria as the de­ gree to which needy people are seen as having (had) an influence on their being needy, their identity as closer to ‘us’ or to ‘them’, and the de­ gree to which they have ‘earned’ support in previous times. On grounds of these criteria one can imagine why it is that the ‘unabled’ meet more solidarity than the unemployed, and that immigrants meet least. Being old, sick and disabled are usually seen as outside people’s control. There is usually more doubt about the inevitability of unemployment, and im­ migration is usually seen as driven by personal preferences and choice (except in case of refugees). The elderly, sick and disabled are more of­ ten seen as belonging to ‘us’, since we all run these ‘risks’, while unem­ ployment usually runs in specific segments of a population, and immi­ grants quite easily and mostly are seen as a ‘them’-group. Especially elderly people are seen as having earned support after having contributed to society in their working lives, while former contribution to society is more questionable in case of some groups of unemployed people (nota­ bly the younger age groups), and immigrants as ‘newcomers’ are usually seen as not having earned anything at all yet. Figure 10.4 gives us some information on the validity of the substitu­ tion and the communicating vessels hypotheses. The substitution hypo­ thesis predicts that where state organized and enforced formal solidarity is strongest informal solidarity among the public at large will be rela­ tively weak, and vice versa. Taking welfare state type as a proxy for the degree of formal, state organized solidarity, one would expect lower lev­ els of informal solidarity in the more extended Scandinavian, than in the less extended Anglo-Saxon and the less developed Southern European welfare states. The Continental, conservative-corporatist welfare states would take up a position in the middle. The communicating vessels hy­ pothesis leads to predictions in the opposite direction, i.e., higher levels of informal solidarity in the Scandinavian type, lowest levels in the An­

288

CHAPTER TEN

glo-Saxon and Southern European types and a position in between for the Continental welfare states. 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 Southern

Anglo-Saxon

unabled

unemployed

Continental

Scandinavian

immigrants

Unabled: eta = .723; F = 4.022; p = .037. Unemployed: eta = .678; F = 3.126; p = .070. Immigrants: eta = .550; F = 1.591; p = .247.

Figure 10.4 Means per type on concern with the unabled, the unemployed, and immigrants

Figure 10.4 shows little evidence in favor of any one of these expected rank orderings. Although there is some slight support for the substitution hypothesis as far as the unabled are concerned, and some even slighter support with respect to the unemployed, the support with regard to the immigrants is negligible. What strikes one, however, is that, seen with the naked eye, the differences between welfare state regimes are very small. If one compares figure 10.4 with Figure 10.3 it seems as if the variation in informal solidarity levels within the four types is greater than it is between the four types. Figure 10.5 shows that this is actually the case for most of the solidarity scales in most of the types, which means that welfare state type is not an adequate variable to explain dif­ ferences in informal solidarity at the aggregate level. As could be expected on grounds of the previous findings an addi­ tional cluster analysis on the three scales of informal solidarity revealed that the individual countries do not cluster according to the four types of

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

289

welfare state. The Southern European welfare states cluster together, but the cluster also includes Sweden and Ireland. And France appears more similar to Finland than to the other Continental welfare states. In other words, the only indication we find is that welfare state regime types do not correlate with specific levels of informal solidarity in society, whether such solidarity concerns the needy groups of the unabled, the unemployed or the immigrants. 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0 Southern

Anglo-Saxon

unabled

C ontinental

unemployed

Scandinavian

Between types

immigrants

Figure 10.5 Variances in informal solidarity within and between welfare state types

As mentioned before, the extent of formal welfare statism can alterna­ tively be indicated, not by regime type, but by welfare state effort, measured by the total of a country’s social expenditure as a percentage of GDP. Table 10.1 shows that, using this indicator, there is a negative correlation between the level of welfare state effort and the level of in­ formal solidarity. This relation is statistically significant in case of soli­ darity with the groups of unabled and unemployed people, but it is hardly so concerning the group of immigrants. Since there are clear dif­ ferences in solidarity with immigrants between countries (as is shown in Figure 10.3), other factors than welfare statism probably play a role. Be­ cause there are only 15 countries in our analysis it is not possible to ana­ lyze the relative effect of a number of likely candidates, such as the per­ centage of immigrants among the population of a country, the degree of

290

CHAPTER TEN

ethnocentrism in the population at large, the average level of education et cetera. Table 10.1 Correlations (Pearson) between welfare state effort and informal solidarity with the unabled, unemployed, and immigrants Welfare state effort Total expenditure on social protection as % of GDP, average 1990-1998 p= N=

Unabled

Unemployed

Immigrants

–.59

–.60

–.32

.010 15

.009 15

.122 15

However, the basic message from Table 10.1 is that it yields evidence for the validity of the substitution hypothesis: countries that spend a higher percentage of their wealth on social programs show lower overall degrees of informal solidarity with needy groups as the unabled (elderly, sick, and disabled people), the immigrants, and the unemployed. The question of why this is so, is of course interesting, but not easily an­ swered. Is it the case that the welfare state corrupts more or less directly people’s feelings of compassion with the needy? Does social spending carry the implicit message that people do not have to worry about others, because this responsibility is well taken care of by the state? Or, are the correlations observed in Table 10.1 spurious and the result of the inter­ play between social spending and the socio-economic and demographic composition of a country’s population? Unfortunately, again the small number of 15 countries makes it impossible to elaborate our analysis of such questions at the aggregate level. Maybe one could say at this stage—because the correlation between welfare effort and the solidarity with immigrants is hardly statistically significant, while the correlations with solidarity with the groups of unabled and unemployed are—that the underlying mechanism might be more a matter of ‘crowding out’, than of ‘moral corruption’. The latter would expectedly lead to negative feel­ ings of solidarity towards all kinds of others, and maybe notably towards foreigners, while the first would particularly lead to less informal soli­ darity with those groups the welfare state spends most money on. Civic morality The level of civic morality, measured as a combination of people’s view on the degree to which they find avoiding taxes, accepting bribes and benefit from fraud justified, clearly differs between European countries, as is shown in Figure 10.6. Greece, Belgium and France are the coun­

291

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

tries were people are most lenient towards deviancy from norms regula­ ting public life and concerning taxes, benefits and bribes. Italy, Ireland and Denmark display the highest level of strict moral standards. 0.3 0.2 0.1 0 –0.1 –0.2 –0.3 –0.4 –0.5 –0.6 pt

es

it

gr

ie

gb

ni

be

nl

fr

de

at

dk

fi

se

Eta: .223; F = 53.266; p = .000.

Figure 10.6 Country means on civic morality

Figure 10.6 does not show a clear pattern, regarding a North-South di­ vide, richer or poorer countries, Catholic or non-Catholic countries, and the like. Figure 10.7 shows that the level of civic morality seems to dif­ fer between welfare state types, but there is no evidence that supports ei­ ther the substitution hypothesis, or the communicating vessels hypothe­ sis. The former hypothesis implies that civic morality will be lowest in Scandinavian, followed by Continental welfare states, and then the other two, whereas the latter hypothesis predicts that it would be the other way round. Furthermore, around the average level of civic morality within the four types there is much individual country variation. Additional analysis showed that the within variance in the Southern European, Con­ tinental and Scandinavian types is larger than the variance between the types. An additional cluster analysis corroborated the finding that a clear pattern is lacking. The most lenient countries, France, Belgium, and Greece, cluster together and can be distinguished from the more strict

292

CHAPTER TEN

societies, who are in two other clusters. But these lenient societies repre­ sent different welfare state regimes, while the same is true for the coun­ tries in the two other clusters. In other words, the clustering of countries according to the level of civic morality does not resemble the clustering of countries according to types of welfare state.

0.15

0.10

0.05

0

–0.05

–0.10

–0.15 Southern

Anglo-Saxon

Continental

Scandinavian

Figure 10.7 Means on civic morality per welfare type

The alternative measure of the extent of formal welfare state provision, by welfare state effort in terms of social expenditure as a percentage of GDP, does not show any relationship with civic morality either. The Pearson correlation between welfare state effort and civic morality is .093 with p = .370. Social capital Social capital, measured as a composite variable of trust in other people and in state institutions, as well as of actively participating in voluntary associations and social networks, varies considerably between European countries. Figure 10.8 shows a North-South divide, and consists of a dis­ tinction between the high-level social capital Nordic countries, and the low-level social capital countries of Portugal, Spain, Italy, and Greece. The other European countries have average levels of social capital, with the exception of the Netherlands, which links up with the Scandinavian

293

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

countries, and France, which links up with the Latin countries of the south of Europe. 1.0 0.8 0.6 0.4 0.2 0 –0.2 –0.4 –0.6 –0.8 pt

es

it

gr

ie

gb

ni

be

nl

fr

de

at

dk

fi

se

Eta = .435; F = 164.528; p = .000.

Figure 10.8 Country means on social capital

These findings suggest that there are differences in social capital be­ tween welfare state types. The levels of social capital are the lowest in the Southern European regime type, more or less average in the AngloSaxon and Continental types, and highest in the Scandinavian type. This is precisely what Figure 10.9 shows, which means that there seems to be some evidence for the communicating vessels hypothesis about the rela­ tionship between the extent of formal welfare provision and social capi­ tal. But this evidence is not final, because of the exceptions of the Neth­ erlands and France. Institutionally these Continental, insurance based welfare states, which both also have national minimum assistance schemes and large occupational social security provision especially in the field of pensions, differ in important respects from their Nordic, re­ spectively Latin counterparts. The question therefore is, whether a communicating vessels mecha­ nism would find its source in welfare state institutional differences, or in sheer extent of welfare effort, regardless of how social provisions are ac­ tually designed. It seems that there is more truth in the latter, than in the former, since we found that the Pearson correlation between social capi­

294

CHAPTER TEN

tal and welfare state effort in terms of a country’s social expenditure as a percentage of GDP is high (r = 0,81), and significant (p = .000). This re­ lation is pictured in Figure 10.10. 0.8 0.6 0.4 0.2 0 –0.2 –0.4 –0.6 Southern

Anglo-Saxon

Continental

Scandinavian

Figure 10.9 Mean level of social capital per welfare state type

However, as in the case of welfare state effort’s (negative) effect on in­ formal solidarity, the question remains through what mechanisms rela­ tive welfare state expenditure is positively correlated with social capital. It may be that social spending creates extra opportunities for people to participate in voluntary organizations, clubs and networks, and the fact that government institutions take well care of citizens might increase trust in institutions. But at this stage it is not totally excluded that the correlation found is influenced also by other, perhaps cultural factors, which may relate to the more family centered culture of the Latin Euro­ pean countries, and the more individualized, and therefore socially more open countries of the north of Europe. Clearly, multivariate analyses would on principle give us the opportunity to test alternative hypotheses in this respect, but the number of cases is again too small.

295

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

1.0 se

0.8

dk nl

0.6 fi

0.4 0.2

be

ie

at

0 ni uk

es

– 0.2

de

Social capital

it fr

– 0.4 pt

gr

– 0.6 – 0.8 20

22

24

26

28

30

32

34

36

Total expenditure on social protection

Figure 10.10

Country means on social capital and expenditure on social protection as % of GDP

The correlation of social capital with civic morality and informal soli­ darity Having assessed the effects of welfare state type and welfare state effort on civic morality, informal solidarity and social capital, we now turn to the effects of social capital on civic morality and informal solidarity. Ta­ ble 10.2 shows that there is no correlation between the average level of social capital among a country’s population and its degree of civic mo­ rality and solidarity with immigrants. Furthermore, it seems, when look­ ing at the simple correlations, that average social capital has a negative effect on the average solidarity with the unabled and the unemployed. However, holding constant for welfare state effort in terms of total social expenditure as a percentage of GDP, the effects disappear. This implies that the simple correlations are spurious, i.e., the result of the fact that welfare state effort has a positive effect upon social capital (as we have just seen) and a negative effect upon a country’s average solidarity to­ wards the unabled and the unemployed (as we have also seen above).

296

CHAPTER TEN

Table 10.2 (Partial) correlations (Pearson) of social capital with informal solidarity and civic morality (aggregate level) Social capital Social capital with ‘welfare effort’ constant

Civic morality .180 .436 *

Unabled –.488 **

Unemployed –.647 ***

–.032

–. 345

Immigrants –.162 .174

* p < .1, ** p < .05, *** p < .01.

Overall conclusion Two things are quite clear: firstly, neither the communicating vessels nor the substitution hypotheses pass the empirical test at the aggregate level with flying colors; secondly, the empirical analyses at this level produce highly confusing results. As far as the presumed direct effects are con­ cerned the proponents of the substitution hypothesis can at least boast of some slight evidence in favor of their hypothesis as far as solidarity is concerned, whereas the proponents of the communicating vessels hypo­ thesis stand empty handed in this respect. Both hypotheses failed in ex­ plaining cross-national differences in civic morality. Welfare state type can be better replaced by welfare state effort as an independent variable if one wants to explain solidarity, whereas welfare state type seems to be a satisfactory variable if one wants to explain cross-national differences in morality. Although the impact of welfare state type is, however, nei­ ther in accordance with the substitution nor with the communicating ves­ sels hypothesis. With regard to the expected indirect effects, the communicating ves­ sels hypothesis was corroborated as far as the influence of welfare state regime and welfare effort on aggregate social capital was concerned, whereas the substitution hypothesis was falsified. Both hypotheses pre­ dicted that there would be a positive correlation between the average amount of social capital in a country and civic morality on the one hand and informal solidarity on the other. The impact on morality was indeed positive, but not statistically significant. Controlled for welfare state ef­ fort the sign of the coefficient stayed positive and became statistically significant (p = .060). The impact on the three forms of solidarity with the needy was, however, negative and in two cases statistically signifi­ cant. Controlled for welfare state effort the sign for two of the cases stayed negative, but were no longer statistically significant, whereas the third form became positive, however, still not statistically significant. Because of these rather confusing results it is high time to change the level of analysis. After all it is individuals and not countries or welfare

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

297

state types who think and act solidarily, who have a sense of civic moral­ ity, who have informal ties with other people, and who trust institutions and generalized others. 5.2 Individual level analyses In Table 10.3 some results of the individual level multi-variate analyses we executed are reproduced. The dependent variables are individuals’ solidarity with the unabled, the unemployed and the immigrants, and their sense of civic morality, respectively. The independent variables are both individual and context level characteristics. In the last but one row of the table we rendered the explained variance of both individual level and context level characteristics and in the last row of the table the pro­ portion of increased explained variance of this model compared to a model with only individual level characteristics as independent varia­ bles. In all cases the individual-cum-context characteristics model added a significant proportion of explained variance to the individual charac­ teristics model. This means that people’s solidarity with needy others and their sense of civic morality is not only a matter of personal values, beliefs and interests, but also shaped by the wider cultural and insti­ tutional settings that they live in. There are, however, differences with respect to the relative impact of context and individual characteristics on each of the dependent variables. Context characteristics (.029) explain solidarity with the immigrants to a much lesser degree than individual characteristics (.090) do. The same applies to civic morality (.038 vs. .073). In the case of solidarity with the unabled the proportions ex­ plained variance by individual and context variables are more or less equal (.054 vs. .056). Context variables explain solidarity with the un­ employed to a higher degree (.073) than individual characteristics (.057). In the other rows of Table 10.3 only the factors that have a statisti­ cally significant impact (p < .01) on the dependent variables are inclu­ ded. The cells that we have left in blank contain ‘hidden’ coefficients that are not statistically significant. The cells with an x in it refer to vari­ ables that are included in some of the rows (analyses), but not in these ones. Informal solidarity Firstly, within the set of individual characteristics a common pattern is found that solidarity is higher among those who have a higher level of social capital, i.e., those who have more trust in other people and in in­ stitutions, and who participate more actively in informal social networks and voluntary organizations. This kind of pro-social attitudes and beha­

298

CHAPTER TEN

vior tends to go hand in hand with stronger feelings of solidarity with needy groups, irrespective of the character of the group. Table 10.3 Regression of individual and context level characteristics on informal solidarity and civic morality Unabled

Unemployed Immigrants

Civic morality B Beta 3.427

B Beta B Beta B Beta 9.852 2.973 (Constant) 3.545 Personal characteristics Age .007 .135 .008 .133 .003 .050 .012 .189 Income .010 .060 Educational level Catholic Church attendance .031 .086 .034 .081 .004 .098 .034 .080 Political preference –.017 –.039 –.056 –.113 –.079 –.154 –.018 –.036 Social capital .045 .053 .076 .075 .131 .127 .037 .038 .308 .063 x Unemployed x x x x x Ethno-centrism x x x x –.143 –.168 x x Negative view on others x x x x x x –.294 –.160 Context characteristics Old age ben. exp. –.0004 –.117 x x x x x x Disability ben. exp. –.001 –.115 x x x x x x x x x x x x Sickness ben. exp. .001 .167 x x x x x x Unemployment ben. exp. x x –.008 –.272 x x x x Unemployment rate 1998 x x x x –3.610 –.093 x Percentage immigrants x Mean ethno-centrism x x x x x x Mean negative view on others x x x x x x 1.426 .270 x x x x x x –.110 –.190 Corruption level Welfare effort –.089 –.312 –.032 –.112 –.034 –.124 Mean age –.060 –.075 –.098 –.104 –.046 –.050 Mean income .152 .127 –.424 –.303 –.122 –.086 –.165 –.120 .056 .146 .157 .349 .039 .087 Mean educational level .255 .080 –.497 –.132 Mean political stand Catholic* church attendance

–.0004 –.084 .0005 .096 .0008 .138 .0006 .107 Mean social cap.

–.786 –.415 –.653 –.295 .646 .299 .130 .119 R2

.111 .110 R2 change vs. individual level .073 .029 .056 .038 model Notes: * all coefficients significant at p < .01; x = variables not included in regression.

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

299

Secondly, older people as a rule tend to have stronger solidaristic feel­ ings towards needy groups. It is not certain whether in the European so­ cieties that we analyzed the age-effect really is a matter of life cycle, which would mean that, when people get older, they become more soli­ daristic, e.g., because they have learned over their lives that fate or bad luck can knock on every one’s door, or because older age brings the fu­ ture risk of personal neediness more closely to the mind. It is also possi­ ble that the age-effect observed actually is a cohort-effect, or a genera­ tion effect, which could mean that the correlation between age and in­ formal solidarity would change once the present cohorts and generations will become older. But, both would also mean that all the cohorts or generations in all of the 15 countries in our analysis would have experi­ enced highly similar events during their life courses. This might be too strong an assumption, which would favor the idea that the age-effect found is a genuine life cycle effect (growing older). Regrettably, on the basis of our cross-sectional data it is not possible to definitely test the specific character of the age-effect. Thirdly, informal solidarity is higher the more to the left an individual’s political orientation. This finding is in line with many other studies on the relation between this characteristic of people and their opinions on welfare state issues: people on the political left usually are more sup­ portive towards welfare state arrangements and expenditures, than peo­ ple on the right (Whiteley, 1981; Taylor-Gooby, 1983; Svallfors, 1997; Forma, 1999; Van Oorschot, 2000, 2002), and in our analysis they also show to be more solidaristic towards needy groups. Obviously, the left’s orientation towards social equality and social responsibility versus the right’s focus on individual interest and responsibility is to a certain de­ gree mirrored in people’s degree of solidarity. A fourth and final basic individual characteristic related to solidarity is church attendance. People who attend church more frequently have a stronger solidaristic attitude towards needy groups. As Table 10.3 shows, this relation does not depend on the answer to the question of whether people are Catholic or not, which means that it is present among all denominations. And it is also independent of age, which is signifi­ cant, since usually church attendance is higher among the older agecohorts. Since the overwhelming majority of our sample is either Catho­ lic or Protestant, the relation found may be based on a Christian moral of compassion which, by assumption, is higher among churchgoers who are often taught to have compassion with the needy during mass or service, than among others. Next to the basic personal characteristics we just discussed, Table 10.3 shows a number of influences that are specific for the type of needy

300

CHAPTER TEN

group concerned. Not surprisingly we find that unemployed people have a higher degree of concern towards the group of unemployed people, and people that are more ethnocentric (i.e., less favor people from other eth­ nic background as neighbors) are less solidaristic towards immigrants. Furthermore, and interestingly, educational level does not affect solidar­ ity towards the unabled and the unemployed, but it does have a positive effect on solidarity towards immigrants. This might indicate that people with a lower educational level lack a more abstract, more extended (but maybe less intense) solidarity with generalized others that is usually conveyed through education. Since the elderly, the sick, the disabled and the unemployed are less abstract others in our society, this interpretation might also explain why educational level has no influence on the solidar­ ity towards these groups. Interesting is also that income has no effect on informal solidarity. It is sometimes found in opinion surveys that higher income people are less supportive of welfare state arrangements, proba­ bly because they perceive welfare as less in their personal interest (Pet­ tersen, 2001; Taylor-Gooby, 2001). In any case, our findings suggest that their possible anti-welfare sentiment is not based on a lesser degree of informal solidarity towards needy groups. Among the set of contextual variables there is one variable that is more basic, in the sense that it plays a similar role in people’s solidarity towards all three groups. This is a population’s average educational level: the higher this level is, the more solidaristic individuals are to­ wards needy groups. In other words, in a higher educated population, people of all educational levels are more solidaristic than in less edu­ cated populations. We have seen that an individual’s level of education only influences solidarity towards immigrants, and not the solidarity to­ wards the groups of unabled and unemployed people. This implies that the positive effect of education works in two ways: directly at the indi­ vidual level and as a ‘spin off’ of a higher collective educational level – compared to its effect on the solidarity with unabled and unemployed people. If the basic mechanism is that education conveys a broader soli­ darity with generalized others, than living in a context in which such solidarity is more prominently present (e.g., as exposed through the me­ dia) all people will be influenced by it, more or less to the same degree. All other contextual variables differ in their influence on the solidar­ ity towards the three needy groups, either by the degree of their influ­ ence, or their direction. However, overlooking the whole there seems to be evidence for the ‘substitution effect’ hypothesis. This hypothesis most generally says that feelings of informal solidarity will be lower, if people live in a social context in which there is more formally organized solida­ rity, i.e., were needs of groups that are the object of solidarity are more

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

301

and better cared for by formal institutions and organizations in society. ‘No need to be personally concerned about those people, since society (or the state) takes care of them (…)’, might be seen as the core attitude of people towards needy groups under the substitution hypothesis. The evidence we find in Table 10.3 for such an attitude regards the facts that:

– informal solidarity is lower among people in countries that spend higher proportions of GDP on welfare: in case of solidarity with un­ abled people it is expenditure on old age and disability benefits which negatively influences solidarity, and not welfare effort itself, which can be explained by the fact that in most welfare states the largest proportion of welfare spending consists of pensions; – in poorer countries informal solidarity towards the elderly is lower and informal solidarity towards the unemployed and immigrants is higher, than in richer countries: this fits the substitution hypothesis if one recognizes that usually in welfare states benefits and provisions for the elderly and disabled are on a higher level and more univer­ sally available than those for unemployed people and immigrants, which means that in poorer countries people will be relatively less solidaristic towards the elderly and disabled and relatively more to­ wards the unemployed and immigrants, than in richer countries; – in countries with a high average level of social capital informal soli­ darity towards unabled and unemployed people is higher, than in countries with a low average level of social capital among their populations: this fits the substitution hypothesis under the assump­ tion that a high level of social capital in a society conveys a general message, and creates a general feeling, that needy groups in society are well taken care of by the work of active citizens in institutions, organizations, clubs and networks in civil society. The high trust in institutions, which is also a part of social capital, also means that one relies on their effective role in social care for needy groups. The fact that average social capital has no negative effect on solidarity to­ wards immigrants might in this line of reasoning be explained by the fact that there are no traditional civil society institutions or networks that have been directed at immigrants, since in the Western societies we analyze here this group has only recently come to the fore as a problematic and needy social group. Clearly, such substitution effects indicate that crowding out, rather than moral corruption is the basic mechanism which underlies them. Inspection of Table 10.3 additionally leads to the conclusion that there might be something like a ‘national burden effect’. Such an effect

302

CHAPTER TEN

implies that informal solidarity is on average lower, if the needy group that is the subject of solidarity is relatively larger. In such cases, infor­ mal solidarity might be tempered by a recognition of, or at least an un­ easy feeling about the costs that could be involved when meeting the needs. The findings that fit this interpretation are that:

– the informal solidarity with unemployed is lower in countries with a high unemployment rate, than in countries with less unemployment; – informal solidarity towards immigrants is lower in countries with higher percentages of immigrants; – informal solidarity with unabled people is lower in countries with an older population. Special discussion is needed for the remaining effects of individual vari­ ables. One that needs attention is the positive effect of a country’s un­ employment benefit expenditure on the informal solidarity towards un­ employed people. This is contrary to the negative effect that expenditure on old age and disability benefits has on solidarity towards unabled, and thus seems to refute the substitution hypothesis. However, there is a dif­ ference in the character of the needy groups involved. Since the propor­ tions of older, sick and disabled people in a society are relatively con­ stant over a certain time period, the relative degree of spending on old age and disability is for the largest part an indicator of the universality and generosity of such benefits. That is, of the level and adequacy of formally organized solidarity. The higher this is, says the substitution hypothesis, the lower informal solidarity among the population. Unem­ ployment, on the contrary, tends to fluctuate substantially, which means that spending on unemployment benefits is more dependent on such fluctuations, than it is on benefit universality and generosity. Since our benefit expenditure figures are averages over a period of 10 years a higher unemployment benefit expenditure figure therefore indicates higher average unemployment rates over that period, rather than a higher level and a more adequate form of formal solidarity. For this reason, no clear substitution effect would be expected. However, the positive rela­ tion between unemployment benefit expenditure and informal solidarity towards the unemployed is also at odds with the national burden effect, which would imply that informal solidarity on average will be lower if the needy group concerned is relatively larger. For the moment we have no clear solution for this interpretation problem, but it could be that people’s uneasy feelings about the size of needy groups are based more strongly on actual situations, rather than on averages or developments over time. If this is true, than we have a second argument for explaining

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

303

why the relationship at issue is not negative. But the question would re­ main why the relation is positive, and not non-existent. As for the average political stance of a country’s population, Table 10.3 shows that in more rightist countries informal solidarity with un­ abled people is higher, than in more leftist countries. But political orien­ tation has no effect on the informal solidarity towards unemployed peo­ ple and immigrants. A possible explanation for this could be that in a rightist culture more emphasis is put on deservingness criteria on which unabled people, notable pensioners, usually score (much) higher, than other needy groups. Such criteria are control, reciprocity and identity, i.e., the criteria that one cannot be blamed for one’s neediness, that one has to ‘earn’ solidarity on the basis of (former) contribution to society, and that one has to belong to ‘us’. This would not necessarily mean that unemployed people and immigrants in a rightist culture would be seen on average as clearly less deserving than they would in a leftist culture. Civic morality As far as the variation in the strictness of people’s moral standards is concerned, it appears that, if we only look for the time being at individ­ ual characteristics as explanatory variables, this variation is to a large ex­ tent determined by age. Older people are generally speaking stricter than younger people. This finding can, as before, either be interpreted as a life cycle or as a cohort effect. Does it mean that people become sadder and wiser as they grow older? Or can this finding be understood as a product of the differences in the settings in which the younger and older people of today have been raised and socialized? Today’s young people cer­ tainly have been raised and socialized in more modern, affluent settings than older people. And since affluent settings are generally regarded as fostering an individualist ethos and more or less libertarian opinions (Yankelovich, 1994), the cohort effect thesis could be true. Level of education does not add significantly to the explanation of differences in levels of civic morality, which is not in accordance with the results obtained in other studies. Higher levels of education generally produce morally progressive viewpoints and generate more acceptance of looser moral behavior linked e.g., to sexual matters (Harding, Phillips & Fogarty, 1986; Halman, 1995). The reason why people with different levels of education do not differ in their levels of civic morality here can probably be attributed to the peculiar kinds of financial and economic acts and behavior that are evaluated in our measure of civic morality. They concern benefit fraud, tax fraud, tax evasion and bribing. Catholics are not more or less strict or lenient compared with people of other denominations, and this is rather remarkable because it is often

304

CHAPTER TEN

assumed that, because of the hierarchical structure and the dogmatic cul­ ture of Catholicism, Catholics are less affected by the processes of indi­ vidualization and secularization, which usually lead to a more lenient and permissive morality (Wilson, 1997). Apparently, not so much the Catholic faith or the Catholic church institution in themselves matter, but the degree to which people of all denominations practice their religion. That is, we find that frequent church attendance is positively correlated with people’s level of civic morality. Church attending people are less inclined to accept the deviant behavior included in our measure of civic morality. Churchgoers probably are more apt to be under the guidance of the churches. While in contrast, people who do not attend religious ser­ vices, are assumed to be more inclined to favor an ethical individualism, which makes that not God but the individual is the ultimate authority in moral decisions and behavior (Crittenden, 1992). This seems to be re­ flected in a more lenient morality. The results of our analyses are in line with the usual expectations about differences in moral outlook of people on the political left and on the political right. A right wing political stance is commonly associated with conservative, traditional standpoints and people on the political right are seen as proponents of authority, order, in favor of maintaining the status quo and a restrictive moral community. People on the left, however, are regarded to be more progressive, more solidary with the poor and deprived, in favor of a just distribution of income and wealth and in moral issues more open-minded (see Halman & Heinen, 1996). The analyses do not corroborate such suggestions, most probably be­ cause of the specific forms of deviant behavior mentioned in our meas­ ure of civic morality, which all focus on an individual’s relation with the state as a tax collecting and income redistributing entity. It could be ar­ gued that rightist people have stronger negative feelings towards the re­ distributing state, than leftist people, that they emphasize personal free­ dom and autonomy in the socio-economic domain more, and that this might explain that they disapprove less of tax cheating, tax avoiding and bribing. As for social capital and trust, it is generally believed that they refrain people from pursuing sheer self-interest and pure egoism. Social capital is believed to stimulate moral behavior and that it provides the ‘glue that bind people together in the absence of enforceable contracts policed by external agents’ (Uslaner, 1999: 216). Being embedded in social groups more strongly and having stronger group attachments, might make group members more likely to have higher levels of civic morality, because it can be assumed that within a more cohesive collectivity there is stronger acceptance of rules of trustworthiness and a stronger rejection of free­

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

305

riding and improper behavior. As such social capital can be considered a moral resource (Putnam, 1993). Our analyses provide evidence for these arguments both at the individual and aggregate level: not only people with more social capital appear stricter, also societies that are richer on social capital disapprove more strongly of fraud and bribing. Remarkable at first sight seems our finding that people who have strong negative views on the moral behavior of their fellow-citizens, i.e. who perceive more strongly that cheating on taxes and benefits, fraud and bribing are wide spread, have a less strict civic morality. That is, they are more lenient towards such behavior. However, this could be un­ derstood as, maybe a fatalistic, adjustment of individual people’s orien­ tations to perceived circumstances, as an adjustment of morals to per­ ceived practices. Individual people might adapt to a situation in which they think that cheating, fraud and bribing is more or less ‘the standard’. However, at the aggregate level things are different. In societies were it is generally believed that tax and benefit cheating, fraud and bribing is wide spread, the collective reaction seems to be a more strict civic mo­ rality towards these kinds of behavior. There is a clear difference in the relation between the public’s perception of deviant behavior and civic morality, compared to the relation between the corruption perception by businessmen and civic morality. Where the first is positive, as we have just seen, the latter is negative. That is, in countries were businessmen see much corruption civic morality is more lenient. As for welfare state effort, Table 10.3 shows a negative effect on civic morality. Since our measure of civic morality focuses on people’s moral behavior towards the state as a tax collecting and welfare redis­ tributing entity, the relation might indicate that high spending, and thus high taxing welfare states arouse a moral counter reaction, in the sense that people become less averse of cheating on taxes and benefits. This would be proof of a ‘moral corruption’ effect. Our finding that civic morality is lower in countries with a more rightist population can be understood on grounds of our previously dis­ cussed finding that rightist individuals are less strict towards tax cheat­ ing and bribing. But such a direct relation is not present in case of age, income and educational level. At the individual level we find that older people are stricter, but at the aggregate level we find that people living in older populations are less strict. At the individual level we find no rela­ tion between income and civic morality, but people in richer societies are less strict. And, at the individual level we find no relation between educational level and morality, but at the aggregate level we see that people in a higher educated society are less strict. We do not see imme­ diate explanations here, but it seems that mean age, mean income and

306

CHAPTER TEN

mean educational level are indicative of a latent, underlying factor that has to do with a country’s complex of demographic, socio-economic and cultural development. 6 Conclusions In this chapter we have examined two hypotheses on welfare state dy­ namics: the substitution hypothesis, advanced by welfare state pessi­ mists, and the communicating vessels hypothesis, defended by welfare state optimists. The former hypothesis predicted an ongoing erosion of social capital, informal solidarity and civic morality as a result of the so­ cial structures and culture created by the welfare state. The latter one predicted that the institutional matrix of the welfare state would stimu­ late people to invest in social capital and would offer incentives for in­ formal solidarity and civic morality. We have tried to test both hypothe­ ses as thoroughly as possible by using EVS data. One of the problems we encountered was, however, that although the EVS survey was fielded in three waves (1981, 1990, 1999), we could only use data from the 1999 wave, because the earlier waves did not include all the information we needed. This was a severe draw back. The competing hypotheses refer to processes in time – welfare state and public opinion dynamics – and all we can do is cross-sectional analyses. We decided to persevere, how­ ever, considering that a bird in the hand is worth two in the bush. We will first reiterate the results of the aggregate level analyses and then proceed with the outcomes of the individual level ones. Striking was first of all how small the observed between countries variation of average levels of informal solidarity and civic morality was. There seems to be, at least in this respect, more unity than diversity discernible in Western Europe. As far as variation between welfare states could be observed, the grouping together of these states in regime types did not explain this variation satisfactorily. If welfare states come in types, as many students of the welfare state assume, the hypothetical clustering predicted in the literature is not discernible from our aggregate analyses. Between types variation was most of the time smaller than within type variation. Whether this is a result of the blurring of regime types and the proliferation of hybrids in the last few decades, as Goodin and Rein (2001) assume, is impossible to establish empirically because we have no satisfactory time-series data at our disposal to test this idea. What is interesting, however, is that there is a high positive correlation between a country’s level of expenditure on social protection and the amount of so­ cial capital. So, the degree of extensiveness and intensity of a welfare state seems to determine the amount of social capital. ‘Seems’, because

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

307

we do not know what came first. Does welfare state development stimu­ late the building up of social capital, or is it the other way round? Neither the substitution nor the communicating vessels hypothesis was convincingly corroborated or falsified by our aggregate level analy­ ses. Although the communicating vessels hypothesis stood its ground fairly well as far as the relation with social capital was concerned, the di­ rect and indirect relations with informal solidarity and civic morality were more complex The findings of the aggregate analyses were, as far as informal solidarity was concerned, to at least some degree in accor­ dance with the substitution hypothesis, but these findings were at vari­ ance with the predictions about social capital and civic morality. One could argue therefore that the empirical competition between both hy­ potheses resulted in a draw, but a more reasonable interpretation of the aggregate findings is that they resulted in confusion. It would be naïve to suppose that people’s attitudes about informal solidarity and civic morality are entirely determined by the welfare state regime or the welfare effort of the country that they live in. Research findings (Arts & Gelissen, 2001) make it immediately apparent that in­ formal solidarity and civic morality will not be completely determined by contextual factors. If that were the case, we would find a nearly gen­ eral consensus on issues of informal solidarity and civic morality within welfare states. We did presume, however, that there would be systematic variance at the individual level not only caused by contextual factors, but also by people’s situational location and their cultural belongingness, i.e., individual characteristics. What were the results of the individual level analyses? As expected, such characteristics played an important part in explain­ ing informal solidarity and civic morality at the individual level. The impact of individual characteristics was greater than that of context characteristics in case of civic morality and informal solidarity with im­ migrants, smaller with respect to solidarity with the unemployed and more or less equal with regard to solidarity with the unabled. As far as the direct effects of welfare effort are concerned, the results of the indi­ vidual level analyses, controlled for the individual characteristics, sooner corroborated the substitution hypothesis than the communicating vessels hypothesis. The more affluent a country is and the higher its welfare ef­ fort the lower the informal solidarity of its inhabitants towards the un­ employed and the immigrants and the lower their civic morality. As far as the indirect efforts are concerned, both hypotheses predicted a positive effect of the average amount of social capital in a country on both informal solidarity and civic morality. The hypotheses hold only true with respect to civic morality and not with regard to informal soli­

308

CHAPTER TEN

darity. As an individual characteristic, however, social capital is posi­ tively correlated with both informal solidarity and civic morality. This means that social capital has the predicted effect on the individual level, but only to a certain extent at the country level. Next to this, some general patterns were found of individual and of country characteristics influencing positively people’s informal solidar­ ity with unabled people, unemployed people and immigrants alike, as well as their degree of civic morality. At the individual level, we found that older people, leftist people and people who attend church frequently are usually more concerned with the living conditions of needy groups and have a higher civic morality, than their counterparts. Regarding con­ textual factors we found evidence for the existence of a ‘national burden’ effect, which implies that informal solidarity towards a needy group is tempered if this group is relatively larger. Informal solidarity towards all needy groups is higher, however, in countries with a higher educated population. All in all, our findings indicate that high welfare spending is related to high aggregate levels of social capital, and that it tends to have a nega­ tive effect on people’s informal solidarity towards the groups that are cared for by the welfare state and its institutions. The latter not as a re­ sult of some form of moral corruption brought about by the welfare state, but as a crowding out effect: ‘(...) no need to worry about needy people, since the state takes well care of them (...)’ seems to be the core attitudinal reaction to a well developed welfare system. However, the re­ lations analyzed here have shown to be more complex than either wel­ fare state optimist or pessimist assumed. There seems to be more be­ tween their sweeping statements and our empirical findings than dreamed of in their philosophies. References Atkinson, A.B. 1989. Poverty and social security. Hemel Hempstead: Harvester Wheatsheaf. Arts, W. & J. Gelissen 2001. Welfare states, solidarity and justice principles: Does the type really matter? Acta Sociologica 44: 283–299. —— 2002. Three worlds of welfare capitalism or more? A state-of-the-art report. Journal of European Social Policy 12: 137–158. Attias-Donfut, C. & S. Arber 2000. Equity and solidarity across the generations. Pp. 1–12 in S. Arber & C. Attias-Donfut (eds.), The myth of generational conflict: the family and state in ageing societies. London: Routledge. Barr, N. 1992. Economic theory and the welfare state: a survey and interpretation. Journal of Economic Literature 30: 741–803. Boje, T. 1996. Welfare state models in comparative research: do the models describe reality? Pp. 13–27 in B. Greve (ed.), Comparative welfare systems: the Scandinavian model in a period of change. Houndmills: Macmillan Press.

THE WELFARE STATE: VILLAIN OR HERO OF THE PIECE?

309

Bryson, C. 1997. Benefit claimants: villains or victims? Pp. 73–88 in R. Jowell, J. Curtice, A. Park, L. Brook, K. Thomson & C. Bryson (eds.), British Social Attitudes, the 14th report. Aldershot: Ashgate. Coughlin, R. 1979. Social policy and ideology: public opinion in eight rich nations. Compara­ tive Social Research 2: 3–40. Crittenden, J. 1992. Beyond individualism. Reconstituting the liberal self. Oxford: Oxford Uni­ versity Press. Dekker, P. & J. de Hart 2000. Wie is mijn naaste? Pp. 22–38 in R. Hortulanus & J. Machielse (eds.), Wie is mijn naaste? ‘s Gravenhage: Elsevier. De Swaan, A. 1988. In care of the state: state formation and collectivization of health care, education and welfare in Europe and America in the modern era. Oxford: Polity Press. Esping-Andersen, G. 1990. The three worlds of welfare capitalism. Cambridge: Polity Press. Ester, P. & L. Halman 1994. Consumentisme en waardenoriëntaties. Is er sprake van een con­ sumentistisch ethos in Nederland? Pp. 11–38 in P. Ester & L. Halman (eds.), De cultuur van de verzorgingsstaat. Tilburg: Tilburg University Press. Ferrera, M. 1996. The ‘Southern’ model of welfare in social Europe. Journal of European So­ cial Policy 6: 17–37. Flap, H. 1999. Creation and returns of social capital. Revue Tocqueville 20: 5–26. Flora, P. & A. Heidenheimer 1981. The historical core and changing boundaries of the welfare state. Pp. 17–34 in P. Flora & A. Heindenheimer (eds.), The development of welfare states in Europe and America. New Brunswick, New Jersey: Transaction. Forma, P. 1999. Interests, institutions and the welfare state: studies on public opinion towards the welfare state. Helsinki: Stakes. Fukuyama, F. 2000. Social capital and civil society. Washington: IMF working paper. Gallie, D. & S. Alm 2000. Unemployment, gender and attitudes to work. Pp. 109–133 in D. Gallie & S. Paugam (eds.), Welfare regimes and the experience of unemployment in Europe. Oxford: Oxford University Press. Gelissen, J. 2002. Worlds of welfare, worlds of consent? Public opinion on the welfare state. Leiden: Brill. Goodin, R. 1996. Institutions and their design. Pp. 1–53 in R. Goodin (ed.), The theory of insti­ tutional design. Cambridge: Cambridge University Press. —— & M. Rein 2001. Regimes on pillars: alternative welfare state logics and dynamics. Public Administration 79: 769–801. Halman, L. 1995. Is there a moral decline? A cross-national inquiry into morality in contempo­ rary society. International Social Science Journal 145: 419–440. —— & T. Heinen 1996. Left and right in modern society. Pp. 35–58 in L. Halman & N. Nevitte (eds.), Political value change in Western democracies. Tilburg: Tilburg Univer­ sity Press. Harding, S., D. Phillips & M. Fogarty 1986. Contrasting values in Europe. London: MacMil­ lan. Janssen, C. 2000. De smeerolie van de samenleving. Een cross-nationale en historische studie naar sociaal kapitaal in Europa. Tilburg: Unpublished MA dissertation Tilburg Univer­ sity. Jehoel-Gijsbers, G., H. Scholten & A. Vissers 1995. Sociale zekerheid en arbeidsparticipatie. Den Haag: Organisatie voor Strategisch Arbeidsmarktonderzoek. Kasza, G. 2002. The illusion of welfare ‘regimes’. Journal of Social Policy 31: 271–287. Knijn, T. 2002. Bijstandsmoeders: burgerschap en de grondslagen voor ondersteuning. Pp. 69– 82 in R. Hortulanus & J. Machielse (eds.), Modern burgerschap. Den Haag: Elsevier Be­ drijfsinformatie. Kohli, M. 1999. Private and public transfers between generations: linking the family and the state. European Societies 1: 81–104. Komter, A., J. Burgers & G. Engbersen 2000. Het cement van de samenleving: een verkennen­ de studie naar solidariteit en cohesie. Amsterdam: Amsterdam University Press. Kuhnle, S. & M. Alestalo 2000. Introduction: growth, adjustments and survival of European welfare states. Pp. 3–18 in S. Kuhnle (ed.), Survival of the European welfare state. Lon­ don & New York: Routledge.

310

CHAPTER TEN

Kuhnle, S. & P. Selle 1990. Meeting needs in a welfare state: relations between government and voluntary associations in Norway. Pp. 165–184 in A. Ware & R. Goodin (eds.), Needs and welfare. London: Sage. LeGrand, J. 1997. Knights, knaves or pawns? Human behaviour and social policy. Journal of Social Policy 26: 149–169. Mommsen, W. (ed.) 1981. The emergence of the welfare state in Britain and Germany. Lon­ don: Croom Helm. Paxton, P. 1999. Is social capital declining in the United States? A multiple indicator assess­ ment. American Journal of Sociology 105: 88–127. Pettersen, P. 2001. Welfare state legitimacy: ranking, rating, paying: the popularity and support for Norwegian welfare programmes in the mid 1990s. Scandinavian Political Studies 24: 27–48. Putnam, R.D., R. Leonardi & R.Y. Nanetti 1993. Making democracy work: civic traditions in modern Italy. Princeton: Princeton University Press. Rothstein, B. 2001. Social capital in the social democratic welfare state. Politics & Society 29: 207–241. Schuyt, C. 1995. Tegendraadse werkingen: sociologische opstellen over de onvoorziene gevol­ gen van verzorging en verzekering. Amsterdam: Amsterdam University Press. Spieker, M. 1986. Legitimitätsproblem des Sozialstaates. Bern: Haupt. Svallfors, S. 1997. Worlds of welfare and attitudes to redistribution: a comparison of eight Western Nations. European Sociological Review 13: 283–304. Taylor-Gooby, P. 1983. Legitimation deficit, public opinion and the welfare state. Sociology 17: 165–184. —— 1998. Markets and motives: implications for welfare. Canterbury: CCWS working paper, Darwin College, University of Kent. —— 2001. Sustaining state welfare in hard times: who will foot the bill? Journal of European Social Policy 11: 133–148. Therborn, G. 1991. Cultural belonging, structural location and human action. Explanation in sociology and in social science. Acta Sociologica 34: 177–191. Uslaner, M. 1999. Morality plays. Social capital and moral behavior in Anglo-American de­ mocracies. Pp. 213–239 in J.W. van Deth et al. (eds.), Social capital and European de­ mocracy. London: Routledge. Van Oorschot, W. 2000. Who should get what, and why? On deservingness criteria and the conditionality of solidarity among the public. Policy and Politics 28: 33–49. —— 2002. Individual motives for contributing to welfare benefits in the Netherlands. Policy and Politics 30: 31–46. ——, D. Houtman, A. Komter & L. Halman. 2001. Wij en Zij in Europa: De solidariteit van Nederlanders in Europees perspectief. Utrecht: AWSB, Universiteit van Utrecht. Whiteley, P. 1981. Public opinion and the demand for social welfare in Britain. Journal of So­ cial Policy 10: 453–476. Wilson, J.Q. 1997. The moral sense. New York: Free Press. Wolfe, A. 1989. Whose keeper? Social science and moral obligation. San Francisco: University of California Press. Yankelovich, D. 1994. How changes in the economy are reshaping American values. Pp. 16–53 in H. Aaron, T. Mann & T. Taylor (eds.), Values and public policy. Washington: The Brookings Institution. Zijderveld, A. 1983. Transformatie van de verzorgingsstaat. Pp. 195–223 in P. Idenburg (ed.), De nadagen van de verzorgingsstaat: kansen en perspectieven voor morgen. Amsterdam: Meulenhoff Informatief. —— 1998. The vertical division of the European welfare state. Society 35: 62–71.

CHAPTER ELEVEN

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES: CULTURAL AND ECONOMIC EXPLANATIONS MATTHIJS KALMIJN 1 Background and research questions Attitudes about the appropriate roles of men and women in society have changed considerably in the past fifty years. Empirical indicators show that in virtually all western societies, there is decreasing support for a rigid division of labor between husbands and wives, that acceptance of married women’s work has increased rapidly, and that men are expected to become more involved in traditional female activities like child rear­ ing (Davis, 1984; Myers & Booth, 2002; Scott, Braun & Alwin, 1993; Thornton & Young-DeMarco, 2001). On a more abstract level, these changes consist of three related developments. The first is an increase in the acceptance of individual discretion in the way men and women be­ have. This means that men and women are increasingly supposed to be free to engage in activities that have traditionally been defined as belonging to the role of the opposite sex. The second component is normative emphasis on non-specialized roles. It is not only that women are supposed to be free to work for pay if they want to, for example, it is also believed that women should work. Similarly, it is not only believed that men are allowed to engage in household labor if they want to, men are also expected to contribute. In other words, traditional norms have eroded and have gradually been replaced by new, modern norms. The third and last component of the revolution is increased commitment to the principle of equal rights for men and women. This means that women are believed to have the same social, economic, and political op­ portunities in society as men have. The sex-role revolution occurs in vir­ tually all western societies, but the timing and speed of the trend differs across countries. In the most developed segments of the western world, the revolution appears to be more or less over (Van den Akker, Halman & De Moor, 1994), while other countries are in the middle of the trans­ formation. Like other value changes, the sex-role revolution has been explained from two different theoretical perspectives: a cultural and an economic perspective (Cherlin, 1992; Haller & Hoelinger, 1994; Inglehart, 1997;

312

CHAPTER ELEVEN

Lesthaeghe, 1983). The cultural perspective links attitude changes to other value changes in society. In this perspective, the increase in egali­ tarian sex-role attitudes is believed to be part of a broader value change, which consists of secularization, individualization, and the erosion of as­ criptive and particularistic evaluation principles. Secularization is proba­ bly one of the most important elements in this theory. Most religions promote the idea of sex-role differentiation and spread norms supporting the reproductive role of women and the breadwinner role of men. Not all religions and denominations promote traditional gender norms to the same extent: the Catholic church is more traditional than the Protestant church, but orthodox denominations within the Protestant church are the most traditional. On average, religion is believed to promote traditional views and secularization is therefore regarded as an important reason why sex-role attitudes have changed. The causal order is not certain in the theory, however, since churches may also lose support as a result of gender attitudes becoming more liberal. Nonetheless, it is often argued that religious changes are more fundamental in nature and that changing sex roles are a specific outcome of this process (Inglehart, 1997). In other words, the causal order is believed to run from secularization to liberalizing gender norms. A second explanatory approach argues that value changes are a re­ flection of economic changes in society. Economic and technological in­ novations are the driving force behind behavioral changes and these changes in turn lead to new or modified attitudes that are more in line with the new conditions. In the present context, the hypothesis is that changing sex-role attitudes are a reflection of changes in women’s em­ ployment. In virtually all western societies, there has been a rapid in­ crease in the relative number of (particularly married) women who work. According to the economic perspective, this change was to a large extent governed by an increase in the demand for women’s work (e.g., England & Farkas, 1986; Oppenheimer, 1967; Van der Lippe & Van Dijk, 2001). The expansion of the service sector resulted in a considerable growth of traditionally female occupations and this pulled more and more (mar­ ried) women into the labor market. This change was combined with ris­ ing wages in female occupations, which made economic activity more and more attractive to women. Although a pure demand-side explanation of the rise of women’s work is too simple, it is an important factor, par­ ticularly for explaining the beginning of the sex-role revolution. A tell­ ing illustration in this context is that the practice of hiring married women in some sectors of the labor market (teaching for example), be­ gan at a time when norms (and sometimes laws) were strongly opposed to doing so (Oppenheimer, 1967).

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

313

The economic explanation argues that rising employment of married women resulted in greater acceptance of their employment and in more support for egalitarian sex-role attitudes in general as well. Two mecha­ nisms play a role here. One mechanism lies in the social-psychological notion of attitude ad­ justment (Petty & Wegener, 1998). If a person’s behavior is inconsistent with his or her attitudes, a state of cognitive dissonance emerges which creates a pressure to adjust one’s attitudes or one’s behavior. The latter is less likely than the former and attitudes will be adapted in the direc­ tion of a person’s behavior. The resulting hypothesis is that married wo­ men who work, will become more liberal in their attitudes toward sex roles than married women who do not work for pay (Jansen & Kalmijn, 2000; Thornton, Alwin & Campburn, 1983). At the macro-level, this implies that when more women work, society as a whole becomes more liberal – a compositional adjustment effect. A second, and less often recognized mechanism is that attitudes can also be adjusted in response to seeing other people behave in different and novel ways. When people observe an increase in previously ‘forbid­ den’ behavior around them, they tend to become more accepting of such behavior. One reason for this is that people prefer to have a harmonious view of the world around them. Another reason is that observing previ­ ously forbidden behavior leads to new thoughts and ideas about such be­ havior which provides an incentive to reconsider one’s original views. The hypothesis is that people will become more liberal in their views about sex roles when the relative number of married women who work in their surroundings increases. This effect is believed to operate after controlling for the effect of women’s individual employment status. In other words, the hypothesis implies a contextual adjustment effect, over and above a compositional effect. The two effects both contribute to value change and together produce a macro-level correlation between female employment and sex-role attitudes. In this chapter, I answer the following research questions: 1. How do European societies differ in their attitudes about sex roles? 2. To what extent can these differences be explained by differences in religious and labor force characteristics of individuals on the one hand, and by religious and labor force characteristics of societies on the other hand? The questions are answered by analyzing survey data from the European Values Study (EVS) on about 36 thousand men and women in thirty European countries in 1999. The analyses are guided by hypotheses at

314

CHAPTER ELEVEN

the contextual level and hypotheses at the individual level. The hypothe­ ses are tested with multilevel regression models which make it possible to analyze individual and contextual variables simultaneously. Residual differences between countries will also be analyzed in order to assess to what extent other country-level explanations are potentially important. Note that differences between countries are an indirect way of testing theories about change. Countries differ in the speed and timing of the sex-role revolution so that the position of a more conservative country can be taken as a proxy for the position of a more liberal country some decades ago. This assumption can be (and often is) questioned, but even if one does not believe in such an one-dimensional perspective on change, it is plausible that theoretical mechanisms that are relevant for change will also be relevant for cross-national differences. Country differences in sex-role attitudes have been analyzed before. Van den Akker, Halman and De Moor (1994) analyzed differences in family attitudes in 15 countries using the 1981 and 1990 waves of the EVS. They show that most countries are moving toward more individu­ alized attitudes although traditional attitudes have not (yet) disappeared and in some cases seem to make a come-back. Another descriptive study comes from Scott, Braun and Alwin (1993) who compare the family at­ titudes of Britain, the United States, West Germany, and the Irish Re­ public using data from the International Social Survey Programme (ISSP). A theoretically oriented study is done by Haller and Hoelinger (1994) who analyze eight countries in the ISSP data. Similar to my ap­ proach, Haller and Hoelinger look at religion and female employment and focus on both micro- and macro-level differences, although without distinguishing contextual and compositional effects. They also use the two theoretical perspectives discussed above, but the labels are slightly different. An interesting conclusion in their study is that at the macrolevel, traditional gender views are strongly correlated with religiosity, supporting the cultural perspective, while correlations with the female employment rate are surprisingly weak. My study tests these notions again but uses different data and renews the design by making a separa­ tion between contextual and compositional effects of cultural and eco­ nomic characteristics on individual attitudes. 2 Hypotheses To test the cultural theory, I focus on church attendance. More specifi­ cally, I expect that church attendance at the individual level is associated with more traditional attitudes on gender. I also expect that church mem­ bers are more traditional than non-church members. Recent work sug­

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

315

gests that church attendance and membership are reinforcing factors in changing attitudes. Next, I expect differences among ‘religions,’ with the Orthodox church being most conservative, the Protestant church most liberal (but more traditional than non-church members), and the Catholic church located in between. Another, and more interesting hypothesis is that church attendance at the country level will lead to more traditional sex-role attitudes, even af­ ter the individual effect of church attendance is controlled for. The rea­ son to look at both levels simultaneously is that the contextual effect of­ fers a better test of the cultural perspective. The individual effect of church attendance is not necessarily causal in nature but may reflect se­ lection. After all, persons who are more liberal will be less attracted to the church so that the association does not reflect a causal effect of re­ ligion. An individual’s attachment to a church may simply be a reflec­ tion of his or her pre-existing attitudes. When the association at the indi­ vidual level is controlled for, selection is also controlled for implicitly so that the effect of church attendance at the macro-level will be more causal in nature. The individual effect is due to both selection and cau­ sation, the (net) contextual effect is primarily due to causation. To test the economic theory, the focus is on the employment of women. The hypothesis can again be formulated at two levels. The first (and imperfect) test is obtained by looking at economic characteristics of individual women. We expect that women who work for pay are more liberal in their attitudes towards women’s economic role in society. This association may reflect an adjustment of attitudes to behavior. Since atti­ tudes also lead to behavior, the employment effect is biased by selection (liberal women being more likely to enter the labor force). Without panel data, the effect of value adjustment cannot be separated from the selec­ tion effect. A more attractive test is obtained at the macro-level. I expect that people will be more liberal in their views on gender when the relative number of women who are employed in their country is higher. Because women’s employment is controlled for on the individual level, the pos­ sible effect of this aggregate variable in the multilevel model taps the degree to which people adjust their attitudes to the employment of mo­ thers around them. In one sense, this is a purely contextual effect, reflec­ ting the way people’s attitudes are formed by the people who surround them. An important additional attraction of this approach is that while the adjustment effect of individual employment on attitudes is biased by selection, this is less plausible for the contextual effect of employment on attitudes. The individual effect reflects adjustment and selection, the (net) contextual effect primarily reflects adjustment. The contextual ef­

316

CHAPTER ELEVEN

fect therefore yields a more stringent test of the economic theory on value change than the individual effect. For men, the story is a little dif­ ferent since the effect of female employment may also reflect adjustment to the behavior of the wife. Next to testing economic and cultural hypotheses, I consider several other characteristics which are relevant for explaining differences in sexrole attitudes at the individual level. Following several earlier individual level studies, I expect that life course experiences change the attitudes that people have (Jansen & Kalmijn, 2000; Moors, 1997; Peplau & Wegener, 1998; Thornton, Alwin & Campburn, 1983). Marriage and children tend to make people more conservative, in part because mar­ riage often turns out to be a more traditional arrangement than people expect when they are young. It is plausible therefore, that people adjust their attitudes in a conservative direction when they marry and have children. I also expect that people (and especially women) become more liberal after experiencing a divorce. For more arguments about the rea­ sons for value adjustment in response to life course transitions, I refer to individual level studies. Another important variable is education. Following the classic work of Hyman and others, I expect that better educated persons will be more liberal in their outlook on gender relations (Hyman & Wright, 1979). This association is often attributed to a causal effect: universities and higher vocational schools are institutions that promote norms of (gender) equality. Whether educational institutions really change people’s attitu­ des remains topic of considerable debate, but the present research design does not allow me to go into this debate more deeply. Birth cohort is also included (which is here equal to age). According to Inglehart’s thesis (1997) of cohort effects and cohort replacement, we should find that younger persons are more liberal than older persons. If cohort effects are weak, a period explanation is more plausible. In the period explanation, all age groups are affected by the sex-role revolution to the same extent. Because life course effects are controlled, we tend to believe that the cohort effect will not strongly be affected by pure aging effects. 3 Data and measurement The present analysis uses the 1999 wave of the EVS survey. A total of thirty countries will be used in the analyses, covering all regions of the European continent. We cover Scandinavian countries, Western Euro­ pean countries, Southern European countries, and formerly communist countries. East and West Germany are treated separately. In each coun­

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

317

try, about a thousand respondents were interviewed at home using faceto-face interview schedules. 3.1 Dependent variables I focus on the following six attitude items pertaining to the role of women in society: 1. A pre-school child is likely to suffer if his or her mother works. 2. A working mother can establish just as warm and secure a relation­ ship with her children as a mother who does not work. 3. Being a housewife is just as fulfilling as working for pay. 4. A job is alright but what most women really want is a home and chil­ dren. 5. Having a job is the best way for a woman to be an independent per­ son. 6. Both men and women should contribute to the household income. Respondents could answer (1) ‘fully disagree’, (2) ‘disagree’, (3) ‘a­ gree’, and (4) ‘fully agree’. A neutral statement was not offered but a ‘don’t know’ response was permitted and often used (5-10 percent in some countries). I decided to treat the ‘don’t know’ response as a neutral answer and located it numerically in between ‘agree’ and ‘disagree’ (us­ ing a score of 2.5). Some countries did offer a middle category in ad­ vance and because this leads to incomparable results, these countries were not included in the analyses1. The frequency distributions of the items are presented in Table 11.1. The table shows that a small majority of Europeans believe that a child will suffer when the mother works for pay. At the same time, however, we see that a large majority of women believe that a mother who works can establish just as warm a relationship with her children as a mother who does not work. In other words, beliefs about women’s employment are mixed. When looking at the third and fourth items, we see that about half of the respondents think that being a housewife is just as fulfilling ______________ 1

The item about husband’s and wife’s contribution to the household income yields ex­ ceptional results for the Netherlands. The percentage of Dutch respondents who disagree with the notion that both husband and wife should contribute to the household income was exceptionally high (64 percent, against 19 percent for Europe as a whole). Since nei­ ther other items nor other data sources do point to such a traditional value climate in the Netherlands (Social and Cultural Planning Bureau, 2000), it is hard to believe these re­ sults and I decided to drop the Netherlands when analyzing this item. Note that the phras­ ing of the item in the Dutch language may unintentionally have suggested that spouses should contribute equally to the household income. There are obviously not many people who agree with that.

318

CHAPTER ELEVEN

as working for pay. Similarly, about half of the respondents think that most women prefer the homemaker role to the economic role. The eco­ nomic role of women receives support of the majority of respondents. More than 70 percent of respondents think that both husband and wife should contribute to the household income, and more than 70 percent of respondents think that having a job is the best way for a woman to be in­ dependent. Table 11.1 Frequency distribution of six items on sex roles in 1999 Fully dis­ agree

Dis- Agree Fully No Don’t Total agree agree answer know

A pre-school child is likely to suffer if his or her mother works

8

31

41

16

0

4

100

A working mother can estab­ lish just as warm and secure a relationship as a nonworking mother

4

19

42

34

3

3

100

Being a housewife is just as fulfilling as working for pay

8

29

38

16

1

8

100

A job is alright but what most women really want is a home and children

8

27

41

16

1

7

100

Having a job is the best way for a woman to be an independent person

4

20

44

26

1

5

100

Both the husband and wife should contribute to household income

3

15

46

32

1

3

100

Source: EVS survey 1999 (own calculations). N = 36,285.

The items do not lead to a good scale when treated as one dimension. Cronbach’s alpha is .54 overall, and in some countries the reliability is below .20. Closer inspection of the items suggests that three underlying dimensions of sex-role attitudes are involved. The first two items are in fact not attitudes but beliefs, i.e., beliefs about the consequences of mar­ ried women’s employment. The third and fourth items are attitudes, and more specifically, they are evaluations of the traditional housekeeping role of women. The last two items are attitudes as well, but they focus on the economic role of women. These three dimensions will be correlated,

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

319

but not necessarily strongly. It is possible, for example, that people favor the economic role of women without rejecting the housekeeping role. Similarly, people may be opposed to the economic role of women in the household, without believing that mother’s work is bad for children. In all analyses, I focus on the three subdimensions rather than on an overall scale: a. negative beliefs about the consequences of women’s employment (the sum of item 1 and 2); b. attitudes favoring the housewife role (the sum of item 3 and 4); c. attitudes against the economic role of women (the sum of item 5 and 6). Correlations between these dimensions are .27 (beliefs with housewife), .17 (beliefs with economic role), and .10 (housewife with economic role). All items are coded in a traditional direction which means that higher scale scores point to more traditional views. 3.2

Independent variables

I present a series of multilevel regression analyses in which the three subscales are the dependent variables. Analyses are done for men and women separately and the regression models are cumulative: the first model contains individual level variables, the second set of models adds contextual variables. The variables are defined below (see also Table 11.2):

– being enrolled in school; – working more than 30 hours a week, working less than 30 hours a week (not working is the reference); for men a single dummy is used for being employed; – living in a couple relationship; – children, in three mutually exclusive groups: currently children at home, ever had children but none living at home (empty nest), and never had children (reference); – whether the person ever divorced; – church visit (at least monthly); – religious denominations (Protestant, Catholic, Orthodox, other, no religion is the reference); – level of education (proxied by the age at leaving full-time educa­ tion);

320

CHAPTER ELEVEN

– percentage of women 18-65 who are employed in a given country (obtained from Van der Lippe & Van Dijk (2001) and amplified by EVS data if no population estimate is available in that source); – percentage of respondents in a country who visit church at least once a month (from EVS data). Table 11.2 Means of independent variables Year of birth - 1900 (in decades)

5.39

Full-time employed (women) Part-time employed (women) Employed (men) Enrolled in school

.37 .09 .60 .06

Married/cohabiting Ever experienced divorce Children under 18 at home Empty nest stage

.67 .06 .42 .37

Age at leaving last school Monthly church attendance Member Catholic church Member Orthodox church Member Protestant church Member other church

18.46 .28 .41 .13 .14 .05

Source: EVS survey 1999 (own calculations). N = 36,285.

When analyzing men, we also need a measure for wife’s employment because that would enable us to separate the influence of the spouse from the influence of the wider social context. In the present design, the female employment rate at the country level in the male model may re­ flect an effect of the spouse. The survey did ask for the employment of the spouse, but only when the respondent is not the chief wage earner. Due to this unfortunate omission, we cannot link male attitudes to varia­ tions in wife’s employment. 4 Results I start with descriptive results for the thirty countries and subsequently present associations between attitudes and characteristics at the macro level. These descriptive analyses are followed by a discussion of the multilevel regression models in which the micro- and macro-level are analyzed simultaneously.

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

321

4.1 Gross country differences In Table 11.3, I present the average scores for each country on the three subdimensions, expressed as deviations from the overall mean. This analysis gives a first impression of how different sex-role attitudes in Europe are. Because the subdimensions yield different rankings, it is practical to see first how the overall ranking looks like. When simply taking the average: The most liberal countries are: – East Germany; – Denmark; – Sweden; – the Netherlands.

The most conservative countries are: – Malta; – Poland; – Lithuania; – Russia; – Italy.

On average, Scandinavian countries appear in the most liberal segment in all columns. Formerly communist countries do not appear clustered on either side and the Southern European countries do not look very tra­ ditional (Italy is conservative, for example, but Spain is liberal). The ta­ ble also shows that country differences are not very large. A simple ANOVA test for group differences is statistically significant, but the proportion of variance explained by country differences is 10 percent at most. In other words, there is variation across countries, but there is much more variation within countries. The subdimensions yield different rankings. To assess whether there are patterns in these differences, I calculated so-called inconsistency scores. These are the standard deviations of the three means (deviations) within each country. Finland and Denmark have the highest inconsis­ tency score, which is due to the fact that these countries are liberal about the effects of mother’s employment, while being (surprisingly) con­ servative about the contribution of the wife to the household income. It is unclear whether this is a real effect or whether there are also differ­ ences in question wording at work. Another pattern that emerges is that several formerly communist countries have inconsistent scores. In many of these countries, respondents think that being a housewife is an attrac­ tive role while they also believe that women should contribute to the household income. Examples are Bulgaria, Hungary, Romania, Slovakia, the Czech Republic, and Poland. These countries appear to have a mixed gender ideology – modern in some respects but traditional in others.

322

CHAPTER ELEVEN

Table 11.3 Deviations from means on traditional sex-role attitudes in 30 countries Overall (mean of 5 items)

Negative Positive Negative Inconsis­ beliefs about views on views on tency mother’s homemaker economic score work role role (sum of (sum of (sum of 2 items) 2 items) 2 items) France –.05 –.16 .22 –.31 .22 United Kingdom .01 .02 –.28 .70 .41 West Germany .02 .60 –.38 .21 .40 East Germany –.39 –.47 –1.22 –.34 .39 Luxemburg –.05 .07 .01 .13 .05 Netherlands –.13 –.31 –.58 .24 Belgium –.08 –.30 .05 –.04 .15 Denmark –.39 –1.01 –.75 .10 .48 Sweden –.30 –.72 –.56 –.47 .10 Finland –.07 –.74 .16 .49 .52 Iceland .00 –.48 –.06 .21 Estonia .09 .30 .14 .09 .09 Latvia .02 .28 –.10 –.12 .18 Lithuania .19 .27 .62 .15 .20 Poland .21 .77 .33 –.12 .36 Czech Republic .07 –.14 .40 –.13 .25 Slovakia .04 –.09 .32 –.17 .21 Hungary .07 .12 .35 –.28 .26 Romania –.03 –.29 .25 –.20 .24 Bulgaria –.01 .02 .10 –.35 .19 Russia .15 .21 .45 .22 .11 Ukraine .13 .19 .43 .05 .16 Belarus .10 –.11 .42 .10 .22 Croatia –.02 .09 –.18 –.17 .12 Slovenia –.06 –.24 .02 –.27 .13 Greece –.01 .29 –.12 –.34 .26 Italy .14 .66 .05 .15 .27 Spain –.09 –.10 –.30 .09 .16 Portugal –.02 .38 –.39 –.11 .32 Malta .41 .94 .59 .88 .15 Mean 2.39 4.60 5.36 3.88 St. deviation .48 1.33 1.37 1.26 Anova F-test 143.1* 140.5* 115.6* 71.0* R-square .10 .10 .09 .05 N of countries 30 30 30 28 Source: EVS survey 1999 (own calculations). N = 36,285. Note: Positive values indicate more traditional than average, negative values indicate more liberal than average. Inconsistency score is the standard deviation of the deviations on the three subdimensions. * p < .05.

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

323

So far, the scores of men and women were combined. In Figure 11.1, I present the difference in male and female scores for the overall index and I plot this difference against the average index. The picture first shows that women in virtually all countries are more liberal than men, a difference which has been noted before (Myers & Booth, 2002). More interesting is that gender differences in gender attitudes are systemati­ cally related to the average level on the index. More specifically, we see that in countries where the average is low, the gender difference is large. Sweden, for example, is among the most liberal of countries, and in Sweden, women are much more liberal than men. Italy, in contrast, is a conservative country and there men and women are more in agreement. A longitudinal interpretation of this pattern is that gender differences in­ crease when countries become more liberal. In other words, women are forerunners. 0.3

gre

swe

e-g 0.2

Male - female score

net den

bef lux ice cro w-g fra fin slo por

cze

pol

est uk slo ita rus lat ukr lit bul rom

0.1

mal

spa hun

bru

0

1.8

2.0

2.2

2.4

2.6

2.8

Traditional attitudes (index)

Figure 11.1 Traditional gender attitudes and gender differences

3.0

324

CHAPTER ELEVEN

4.2 Associations at the macro level We now turn to the role of religion and women’s work in explaining the differences. We start with the association at the macro-level. In Figure 11.2, the horizontal axis displays the female labor force participation rate and the vertical axis displays attitudes. For the sake of presentation, I limit this to women’s scores on the first dimension (beliefs about the harmful effects of employment). The figure shows that there is a nega­ tive relation: countries where more women are working for pay have less traditional views on women’s work. The correlation is –.62. When an outlier is removed (Malta), the correlation is still high, –.55. 6.0

5.5

mal pol

Traditional beliefs of women

ita w-g 5.0 gre

lat hun cro

bul bru uk

lux

spa

4.5

por ukr lit rus est

slk cze fra slo

rom

bef

e-g

net 4.0

ice fin swe

3.5

3.0 0.3

den

0.4

0.5

0.6

0.7

0.8

Female labor force participation rate

Figure 11.2 Women’s work and sex-role attitudes

Figure 11.3 presents a scattergram for the association between gender beliefs and religiosity. Here we see a strong positive correlation. The more people who attend church in a country, the more traditional the av­

325

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

erage beliefs about gender. The correlation is .61. After removing highly religious countries, the correlation is still substantial, .43. 6.0

mal

5.5 pol ita w-g

Traditional beliefs of women

5.0

por rusest lat ukr hun bul bru uk

4.5

cze fra

lit gre cro lux

spa

slk

slo

rom

bef net

e-g 4.0 ice fin swe den

3.5

3.0 0

0.2

0.4

0.6

0.8

1.0

%attending church weekly

Figure 11.3 Church attendance and sex-role attitudes

The two figures can in theory be explained by either compositional or contextual effects (or a combination of both). Countries can be more lib­ eral because women who work are more liberal, or countries can be more liberal because both working and non-working women have more liberal views when there are more women working in their society. A similar reasoning holds for religion. If the macro-pattern in the figures is only driven by compositional effects, there is nothing specific about countries that needs to be explained and the problem is reduced to the individual level. In the remainder of this chapter, I try to separate indi­ vidual (compositional) and contextual effects by analyzing the data with multilevel models.

326

CHAPTER ELEVEN

4.3 Individual effects Table 11.4 shows the multilevel regression analyses for women, Table 11.5 shows the results for men. Individuals are the units and they are clustered within countries. Table 11.4

Multi-level regression analyses of women’s traditional sex-role attitudes in 30 countries

Year of birth

Negative beliefs about mother’s work –.052 *

Positive views on homemaker role –.035 *

Negative views on women’s economic role .020 *

Full-time employed Part-time employed Enrolled in school

–.271 * –.264 * –.136 *

–.343 * –.337 * –.338 *

–.279 * –.221 * –.223 *

Married/cohabiting Ever experienced divorce Children under 18 at home Empty nest stage

.021 .009 –.038 –.091 *

.049 * –.010 .120 * .055

.125 * –.070 .097 * –.087 *

Years of completed schooling Monthly church attendance Member Catholic church Member Orthodox church Member Protestant church Member other church

–.033 * .135 * .040 –.014 –.070 .108 *

–.053 * .264 * .023 .013 .053 .129 *

–.010 * .149 * .025 –.015 .050 .051

Intercept Mean of Y Standard deviation of Y R-square within countries R-square between countries R-square overall N of respondents N of countries

5.451 4.5 1.3 .043 .353 .064 19,529 30

6.364 5.3 1.4 .076 .002 .059 19,445 30

3.830 3.8 1.3 .020 .111 .023 18,441 28

Source: EVS survey 1999 (own calculations). * p < .05 (two-sided).

The effects of birth cohort are as predicted for two of the three dimen­ sions for women. Negative beliefs about the consequences of women’s employment become less common in the younger cohorts and support for the housewife role declines as well. For egalitarian views on the eco­ nomic role, there is a positive effect, suggesting that these attitudes are more conservative among the young, a result which is difficult to inter­

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

327

pret. For men, the results are similar. For the first two dimensions, we notice a clear tendency for conservative responses to become less com­ mon in younger cohorts. The third dimension is again an exception and reveals no trend. We conclude that for the first two dimensions, Inglehart’s cohort explanation of change is supported, although longitudinal data are required for a conclusive test. Table 11.5 Multi-level regression analyses of men’s traditional sex-role attitudes in 30 countries

Year of birth

Negative beliefs about mother’s work –.061 *

Positive views on homemaker role –.069 *

Negative views on women’s economic role .016

Employed Enrolled in school

–.089 * –.042

–.119 * –.134 *

–.024 –.014

Married/cohabiting Ever experienced divorce Children under 18 at home Empty nest stage

–.025 .111 * .015 .060

.028 .061 .031 .033

–.032 –.060 .087 * .003

Years of completed schooling Monthly church attendance Member Catholic church Member Orthodox church Member Protestant church Member other church

–.027 * .117 * .043 .104 * .035 .265 *

–.038 * .227 * .086 * .091 * .031 .271 *

–.000 .148 * –.009 –.122 * –.046 .213 *

Intercept Mean of Y Standard deviation of Y R-square within countries R-square between countries R-square overall N of respondents N of countries

5.522 4.7 1.3 .030 .136 .035 16,589 30

6.415 5.4 1.3 .053 .001 .045 16,469 30

3.928 4.0 1.2 .007 .105 .009 15,533 28

Source: EVS survey 1999 (own calculations). * p < .05 (two-sided).

The link between employment and gender attitudes at the individual level is quite strong. For all items, we see that women who are working are more liberal. The effect of part-time employment is not weaker than the effect of full-time employment. The size of the effects is substantial: employment is associated with a change in attitudes of about 20 to 25 percent of the standard deviation of the dependent variable (e.g.,

328

CHAPTER ELEVEN

.271/1.3). This finding is consistent with the notion that attitudes are a reflection of behavior, but it is also consistent with the notion that atti­ tudes lead to behavior. We observe that employment on the part of men has a negative effect on gender attitudes as well. Men who are employed have more liberal beliefs and have more negative views of the housekeeping role of women. These effects are not due to retired men being more conserva­ tive, because when we leave out men over 60, we still find a significant effect of employment (–.10 for the first dimension and –.10 for the sec­ ond dimension). Hence, the effects must reflect the position of unem­ ployed men. A possible interpretation is that unemployment on the part of men leads to more negative views on the employment of women. Un­ der those circumstances, women’s employment is often needed to solve financial problems in the household and may therefore be considered a threat to the male breadwinner role. Note, however, that the effects are small in size. School enrolment has a strong liberalizing effect on all items for both men and women, although the effects are more consistent for women. This replicates the well-known effect that students in higher education are more liberal on all sorts of issues. Educational attainment, which is measured here by the age at which people left school—an internationally comparable measure—has a significant negative effect on all dimensions for women and on two dimensions for men. That the better educated are more liberal in their sex-role attitudes is consistent with a host of earlier findings. The magnitude of the educational effect can best be assessed by calculating the standardized coefficient. For women, the bèta­ coefficients are .10, .15, and .03. For men, the coefficients are .08 and .12. In comparison to earlier studies on education and values, these are rather weak effects. Sex-role attitudes are also affected by life course variables, but the ef­ fects are not always significant and are modest in size. Differences be­ tween married (or cohabiting) women and single women are found for two dimensions. In line with expectations, marriage makes women more conservative, at least with respect to how they think about the economic and the housekeeping role that women can play. This can be interpreted as a case of value adjustment: when women marry, they see that it is dif­ ficult for them to continue to contribute to the household income, they become more realistic, and hence they adjust their attitudes. Marital status differences are absent for men. Attitudes are also associated with having children. Children in the household have a positive effect on the last two dimensions for women and on the last dimension for men. The effects suggest that children make people somewhat more conservative

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

329

in their views about sex roles. Effects of divorce are weak and can only be found for the first dimension and only for men. We expected a liberalizing influence on women, but this is not borne out by the data. People who attend church regularly have more conservative views on sex roles than people who do not go to church or who go only inciden­ tally. This applies to both men and women and is observed for all three dimensions. The effects are not trivial: church attendance is associated with an increase in traditional values of (on average) 13 percent of the standard deviation. When church attendance is controlled for, small ad­ ditional effects of the type of religion emerge, but primarily for men. The reference group consists of non-members, and most ‘religions’ have a significant effect. When a ranking has to be made, the order is (from liberal to conservative): Protestant, Catholic, Orthodox, and others (in­ cluding Muslims). When we simplify the model by including a single dummy variable for church attendance and a single dummy variable for membership, we find generally stronger effects of attendance than of membership (.14 versus. .08, .25 versus .10, and .18 versus –.02; for women). This suggests that it is active membership that counts and not membership per se. When all the individual-level variables are included, the percentage of variance that can be explained is low. The percentage varies from 2 to 6 percent for women and is even lower for men, from 1 to 4 percent. In part, this result is understandable in light of the fact that the explained variance in research on values is always on the low side in comparison to other research areas (Kalmijn & Kraaykamp, 1999). The values re­ ported here are nonetheless lower than usual for value research. Another explanation lies in formerly communist countries. When leaving out these countries, the percentage explained variance increases substan­ tially. Negative beliefs about women’s work provide a good example. The explained variance of this scale is 12 percent for women in capitalist countries, but only 1 percent for women in formerly communist coun­ tries. More generally, it appears very difficult to explain individual value differences in Eastern Europe with the standard social and demographic variables. The models also show how much of the differences between coun­ tries can be explained by individual variables. The explained variance (for women) is highest for the belief scale (35 percent), somewhat lower for the economic scale (11 percent), and absent for the housekeeping scale. For men, country differences are explained to a lesser extent but the order is the same and differences can best be explained for beliefs. The conclusion is that only a part of the variation among countries is due to compositional effects, i.e., differences in the composition of the

330

CHAPTER ELEVEN

population with respect to the individual characteristics that are included in the model. Interesting to observe is that country differences in beliefs can better be explained in the capitalist part of Europe (53 percent for women and 20 percent for men) than in the formerly communist part of Europe (0 for women and 11 for men). 4.4 Contextual effects In Table 11.6 and 11.7, I add contextual variables to the multilevel re­ gression models. Because the number of contexts is small, I follow a lib­ eral strategy by looking at contextual effects one-at-a-time before inclu­ ding combinations of independent variables. We note that multilevel models take into account that the number of units at the macro-level is small and the results therefore will reveal lower levels of significance than OLS regression. The correlation between the proportion employed women and the proportion church attendance at the macro level is rather high, –.64. The combination of a low N and a high intercorrelation be­ tween predictor variables implies that it will be difficult to estimate the contextual effects simultaneously. Table 11.6

Contextual effects on women’s traditional sex-role attitudes in 30 countries

Country characteristics Negative beliefs about women’s work % employed women in country % church attendance in country R-square between Positive views about homemaking % employed women in country % church attendance in country R-square between Negative views about economic role % employed women in country % church attendance in country R-square between

Model I

Model II

Model III

1.12 * .42

–1.56 * .56 .49

–.09 .00

–1.21 –.52 .03

.15 .10

–.13 .10 .10

–2.20 * .46 –.62 .02 –.25 .11

Source: EVS survey 1999 (own calculations).

Note: Multi-level regression analyses. Individual variables included (see Table 11.4).

* p < .05.

The first and most important finding is that the proportion of employed women has a negative effect on sex-role attitudes. Controlling for the ef­ fect of women’s individual employment on individual attitudes, it ap­ pears that women are more liberal when there are more women who

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

331

work in their country. This contextual effect shows up only for the first dimension of attitudes. In other words, when there are more women who work, other women become less negative about the consequences of fe­ male employment. We regard this as evidence for value adjustment at the macro level. Seeing others behave in untraditional ways leads to less support for traditional views. It is also interesting to see that the effect of women’s employment is absent for the other scales. In other words, when more women are working, other women do not change their evaluations of the homemaker and the employment role, they only adjust their beliefs. The size of the coefficient is 2.2, and because the indepen­ dent variable ranges from .33 to .78, this means that the maximum effect is almost one point on the attitude scale (.45 x 2.2). This is a substantial effect. Table 11.7

Contextual effects on men’s traditional sex-role attitudes in 30 countries

Country characteristics Negative beliefs about women’s work % employed women in country % church attendance in country R-square between Positive views about homemaking % employed women in country % church attendance in country R-square between Negative views about economic role % employed women in country % church attendance in country R-square between

Model I

Model II

Model III

1.15 * .39

–1.22 * .72 * .46

.10 .01

–1.05 –.27 .04

.11 .09

–.46 –.04 .11

–2.01 * .39 –.75 .04 –.41 .11

Source: EVS survey 1999 (own calculations).

Note: Multi-level regression analyses. Individual variables included (see Table 11.5).

* p < .05.

Because this is an important finding, I checked whether the results are the same for the capitalist and formerly communist parts of Europe. The results show that the effect only shows up in the capitalist part (the coef­ ficient is –2.7). Moreover, in this part we also see an effect of the female employment rate on attitudes about women’s economic role (the coeffi­ cient is –1.2). Hence, in the western part of Europe, women become more liberal about their economic role as well if the number of women working increases.

332

CHAPTER ELEVEN

Another question that arises is whether the contextual effect applies to different subgroups to the same extent. When splitting the sample into working women and other women, we find that the female employment rate has an effect in both groups. However, the effect is stronger for non­ working women (–2.3) than for working women (–1.8). Both effects are significant, though. In other words, not only non-working women be­ come more liberal when other women are working, working women also become more liberal. When we look at the equation for men, we find social adjustment ef­ fects as well. For the belief scale, the effect is negative and significant, showing that men become more liberal in response to larger numbers of working women. There are also negative effects on the other two dimen­ sions, but only one of them is significant and only in the full model. As discussed before, adjustment effects for men may be the result of two forces: the employment of women in society at large, and the employ­ ment of the spouse. Because there is no information on the spouse’s em­ ployment in the male sample, we cannot separate these effects. Cultural effects at the contextual level are present as well. We find that people are more conservative when the relative number of church at­ tendars in their society is larger, even after controlling for the individual effect of church attendance. These effects are found for both men and women but they are again limited to the belief scale. These findings are in line with a causal interpretation of the cultural perspective: people be­ come more traditional when religious ideologies in society are stronger. The effects of aggregate church attendance are stronger in capitalist countries than in Central and Eastern Europe. When we limit the sample to the west, we find significant effects of aggregate church attendance on attitudes about women’s economic role, the third dimension (.82 for women and .84 for men). In other words, the more people there are in a society who attend church, the more traditional people become in their views about the economic role of women. More detailed analyses indicate that the effect of aggregate church at­ tendance is equally strong for people who attend church often, as for people who do not go to church. The coefficients are 1.2 and 1.1 for men and women who go to church, and 1.0 and 1.2 for men and women who do not go to church (for the first dimension). In other words, in religious countries, both religious and non-religious persons are more conserva­ tive than in less religious countries. This is again evidence of a contex­ tual cultural effect. When looking at the two contextual variables in combination, we note that the effect of female labor force participation becomes weaker, but it is still negative and significant. The effect of church attendance, how­

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

333

ever, becomes insignificant for women, suggesting that for women, eco­ nomic explanations are more important than cultural explanations (at the aggregate level). 4.5 Residuals In Table 11.8, I present residuals from all full regression models. The re­ siduals are averaged for each country and will tell us whether there are any remaining country differences that are left unexplained. This analy­ sis may also point to other, macro-level explanations that have not been covered in the models. We first observe that Scandinavian countries are more liberal than expected given the model. This applies to the first di­ mension most clearly, but also to the other two dimensions. This may point to the often quoted role of public policies in Scandinavian coun­ tries which tend to be highly supportive of equal rights for men and women (Sainsbury, 1996). Such ‘socialization by the state’ explanations are problematic, however, since the reverse influence is more likely. Egalitarian laws and policies will reflect pre-existing liberal attitudes in the population. Without good dynamic data, the effect of public policy cannot be assessed conclusively and institutional explanations remain speculative. Table 11.8

France United Kingdom West Germany East Germany Luxemburg Netherlands Belgium Denmark Sweden Finland Iceland Estonia Latvia Lithuania Poland Czech Republic Slovakia Hungary Romania

Regression residuals of prediction of women’s sex-role attitudes by country Beliefs about mother’s work –.12 .16 .38 –.18 –.04 –.38 –.49 –.70 –.35 –.32 –.14 .62 .36 .37 .41 .01 –.21 .06 –.75

Views on homemaker role .22 –.22 –.62 –1.38 .01 –.58 –.05 –.67 –.28 .45 .32 .31 –.16 .84 .35 .41 .44 .20 –.03

Views on women’s economic role –.27 .81 .16 –.27 .17 . –.12 .19 –.32 .61 . .18 –.04 .14 –.23 –.06 –.26 –.32 –.26

334

CHAPTER ELEVEN

Table 11.8 (continued) Bulgaria Russia Ukraine Belarus Croatia Slovenia Greece Italy Spain Portugal Malta

.41 .49 .39 .28 –.15 –.17 .19 .19 –.45 .14 –.02

.22 .49 .46 .63 –.18 .07 –.25 –.15 –.59 –.54 .25

–.30 .29 .12 .13 –.26 –.25 –.48 .05 .07 –.20 .72

Source: EVS survey 1999 (own calculations). N = 36,285.

Note: Residual score is observed score minus predicted score from multi-level regression.

Hence, positive values indicate more traditional than predicted, negative values indicate

more liberal than expected.

We also observe that formerly communist countries have positive re­ siduals on the first two dimensions and negative residuals on the third dimension. Hence, these countries are more traditional in their beliefs and more traditional in their views about the homemaker role than we would expect given the high levels of participation in these countries. In a more speculative sense, we can say that these countries are more egali­ tarian in their behavior than in their values. At the same time, however, views about the contribution that women should make to the household income in formerly communist societies are more egalitarian than ex­ pected. 5 Conclusions There are modest differences between European countries in sex-role at­ titudes. In 1999, East Germany, Denmark, Sweden, and the Netherlands are the most liberal. The most conservative countries are Malta, Poland, Lithuania, Russia, and Italy. Scandinavian countries are generally lib­ eral, but Southern European countries are not always very traditional. Formerly communist societies do not have a very clear or consistent po­ sition. Individual employment and church attendance are the strongest corre­ lates of sex-role attitudes: working women are more liberal, people who go to church are more conservative. These two characteristics also have contextual effects, supporting both cultural and economic theories on value change. The economic perspective argues that people adjust their attitudes, not only in response to what they themselves experience, but

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

335

also in response to what they observe that other people are doing in their surroundings. This hypothesis is supported by the finding that women are more liberal in their sex-role attitudes when the number of women who work in a country is higher, after controlling for whether they them­ selves are employed. A similar effect is found for men. The cultural per­ spective argues that people base their attitudes on the cultural climate in society. This hypothesis is supported by the finding that men and women are more conservative in their sex-role attitudes when the church is more important in society, even after controlling for whether they themselves are attending church. In both tests, individual controls of the parallel characteristic are important because by including these measures, selec­ tion effects are to a large extent ruled out. Hence, contextual effects are not as strongly biased by selection effects as individual effects and this results in a more stringent test of the cultural and economic theories. A caveat is that overall, it is difficult to explain much of the individ­ ual variance in sex-role attitudes. At the individual level, the percentage explained variance is lower than usual in this type of research. This ap­ plies in particular to formerly communist countries: in these countries, standard sociological determinants explain almost nothing. When look­ ing at capitalist societies, the percentage of explained variance is more acceptable, at least for women. When we focus on country-level differences, we see that some of these can be explained by compositional differences in individual char­ acteristics. This applies in particular to country differences in traditional beliefs. The two contextual factors (church attendance and female em­ ployment) add an important part to the macro-level explanation. The fi­ nal model, which incorporates both compositional and contextual fac­ tors, explains about half of the differences in women’s and men’s beliefs about gender between countries. The contextual effects we find are not only relevant for understanding country differences, they are also relevant from a theoretical point of view. While it is interesting to find that a person’s own behavior affects his or her attitudes, it is sociologically more interesting to find that an individual’s attitudes are affected by the behaviors of other people. Ef­ fects like these point to social influence processes and such processes are usually assessed in the psychological laboratory. Contextual analyses of­ fer a way to discover such processes in survey data and thereby help building the bridge between micro-level processes on the one hand, and macro-level phenomena on the other hand. While this approach has resulted in positive findings in the present analysis, it is also important to discuss some of its limitations. The most important limitation is that the national context is a rather distant proxy

336

CHAPTER ELEVEN

for the context in which individuals are embedded. A more direct test would be obtained by looking at measures of employment and church at­ tendance in the social groups of the respondent, but such measures are not easy to construct. It would nonetheless be interesting for future re­ search to experiment with aggregate measures of church attendance and female employment in age-groups and educational groups within coun­ tries. Another disadvantage is that national contexts differ in many other respects from each other and that the present approach has only looked at two of these. And while these factors appear important, the test is not conclusive. A better test can be obtained by looking at a range of charac­ teristics of national contexts, but such a design is hampered by the relati­ vely small number of contexts that we have available. Distinguishing so­ cial groups within countries may be a solution for this problem as well. References Cherlin, A. 1992. Marriage, divorce, remarriage. Cambridge: Harvard University Press. Davis, K. 1984. Wives and work: Consequences of the sex role revolution. Population and De­ velopment Review 10: 397–417. England, P. & G. Farkas 1986. Households, employment, and gender. New York: Aldine. Haller, M. & F. Hoellinger 1994. Female employment and the change of sex roles: The con­ flictual relationship between participation and attitudes in international comparison. In­ ternational Sociology 9: 87–112. Hyman, H.H. & C.R. Wright 1979. Education’s lasting influence on values. Chicago: Univer­ sity of Chicago Press. Inglehart, R. 1997. Modernization and postmodernization: cultural, economic and political change in 43 societies. Princeton: Princeton University Press. Jansen, M., & M. Kalmijn 2000. Emancipatiewaarden en de levensloop van jong-volwassen vrouwen. Sociologische Gids 47: 293–314. Kalmijn, M. & G. Kraaykamp 1999. De verklaarde variantie verklaard: Een vergelijking van sociologische onderzoeksartikelen in de periode 1975-1998. Mens en Maatschappij 74: 166–180. Lesthaeghe, R. 1983. A century of demographic and cultural change in Western Europe: an ex­ ploration of underlying dimensions. Population and Development Review 9: 411–435. —— (ed.) 2002. Meaning and choice: Value orientations and life course decisions. The Hague: NIDI/CBGS Publications. Moors, G. 1997. The dynamics of value-based selection and value adaptation. Brussel: Vrije Universiteit Brussel, Centrum voor Sociologie. Myers, S.M. & A. Booth 2002. Forerunners of change in nontraditional gender ideology. Social Psychology Quarterly 65: 18–37. Oppenheimer, V.K. 1967. The interaction of demand and supply and its effect on the female la­ bour force in the United States. Population Studies 21: 239–259. Peplau, L.A., C.T. Hill & Z. Rubin 1993. Sex-role attitudes in dating and marriage: a 15-year follow-up of the Boston Couples Study. Journal of Social Issues 49: 31–52. Petty, R.E. & D.T. Wegener 1998. Attitude change: multiple roles for persuasion variables. Pp. 323–390 in D.T. Gilbert, S.T. Fiske & G. Lindzey (eds.), The handbook of social psy­ chology. Vol. 1. Boston: McGraw-Hill. Sainsbury, D. 1996. Gender, equality, and welfare states. Cambridge: Cambridge University Press. Scott, J., M. Braun & D.F Alwin 1993. The family way. Pp. 23–48 in R. Jowell, L. Brook & L. Dowds with D. Ahrendt (eds.), International Social Attitudes. Dartmouth: Hants.

COUNTRY DIFFERENCES IN SEX-ROLE ATTITUDES

337

Social and Cultural Planning Bureau 2000. Sociaal en Cultureel Rapport 2000. Rijswijk: Staatsuitgeverij. Thornton, A. & L. Young-DeMarco 2001. Four decades of trends in attitudes toward family is­ sues in the United States. Journal of Marriage and the Family 63: 1009–1037. ——, D.F. Alwin & D. Camburn 1983. Causes and consequences of sex-role attitudes and atti­ tude change. American Sociological Review 48: 211–227. Van den Akker, P., L. Halman & R. de Moor 1994. Primary relations in Western societies. Pp. 97–128 in P. Ester, L. Halman & R. de Moor (eds.), The individualizing society: Value change in Europe and North America. Tilburg: Tilburg University Press. Van der Lippe, T. & L. van Dijk (eds.) 2001. Women’s employment in a comparative perspec­ tive. New York: Aldine de Gruyter.

This page intentionally left blank

CHAPTER TWELVE

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE: EFFECTS OF CULTURAL, STRUCTURAL AND COMPOSITIONAL FACTORS JOHN GELISSEN 1 Introduction Within different welfare states, the family is commonly, although in varying degrees, perceived as the bedrock of society. The family func­ tions as an important ‘locus of solidarity’ and, according to conservative political views, it is even the nucleus of social integration. However, both family structure and behavior have undergone, and are still under­ going, dramatic changes. Most importantly, whereas the standard nu­ clear family model appeared to be strongly institutionalized in most countries in the immediate post-war period, the one-earner family has become more and more an exception to the rule in recent decades. The stable one-earner family is no longer standard but atypical, and alterna­ tive stable forms of living-arrangements, such as cohabitation and single-person households are becoming more and more the norm. Women are now more able to pursue a career of their own. Finally, children are more likely than in the past to grow up with parents who both work, or with a single parent. They are also more likely to experience divorce or separation of their parents. Despite these important and relatively general changes in family structure and behavior, we can observe marked differences between countries in the popularity of certain family types and forms of demo­ graphic behavior. For example, Kiernan (2000) shows for the countries of the European Union that there is a great deal of intra-European diver­ sity in the extent of cohabiting unions, marital unions and being single. On the ideational level, Lesthaeghe and Meekers (1986) found signifi­ cant differences between eight European countries with respect to famil­ ialism and non-conformism with respect to marriage and procreation. This suggests that not only actual behavior, but also public views about normative demographic behavior may differ in important ways between countries. Esping-Andersen’s study (1999) suggests that public consent to divorce and women’s employment are important values to consider in this respect, because it is on these two dimensions—family instability

340

CHAPTER TWELVE

and female employment—that modern welfare states are being chal­ lenged. Divergent public values about these issues may lie at the heart of actual demographic outcomes and a closer examination of these family values may enhance our understanding of actual demographic behavior. For example, the public’s adherence to traditional family values—such as the non-acceptance of divorce—may be stronger in some countries than in others and this may lead to specific demographic outcomes in a country. In addition, from a societal point of view it is interesting to know to what extent traditional and non-traditional views about the fam­ ily are supported: dwindling legitimacy for more traditional forms of family life may necessitate appropriate policy responses; however, these may be more necessary in some countries than in others. Given the scientific and societal relevance for studying family values, the objective of this chapter, then, is to focus on one of the salient di­ mensions of public family values, namely the degree to which individu­ als see divorce as legitimate1. Prior studies (e.g., Palomba & Moors, 1998) have cross-nationally compared public preferences about living arrangements and the reconciliation of work and the family, but little is known about how public acceptance of divorce is structured crossnationally and how differences and similarities in these public views can be explained. Therefore, we will examine cross-national differences and similarities of public consent to divorce, using data from the European Values Study 1999. These data include a measure for the public’s toler­ ance towards divorce for 33 Western and Eastern European countries. What we will offer in this chapter is, however, not only a description of the cross-national differences in public consent to divorce. We will go beyond mere descriptions and seek explanations for observed differences in public consent to divorce between and within countries, by relating people’s views about the legitimacy of divorce to structural and cultural characteristics of countries, and to their specific historical circumstances, on the one hand, and individual background characteristics, on the other. Svallfors (1997) has suggested that in order to better understand crossnational patterns of values and attitudes, institutional characteristics of welfare states should be taken into account. Welfare state arrangements create a context, in which not only people’s values and attitudes about the welfare state itself are formed, but also values about other major so­ cietal institutions—such as the family—as the family and the welfare state are traditionally strongly interlocked. The leading hypothesis of this chapter is that institutional characteristics of the welfare state, the actual ______________ 1

I would like to thank Matthijs Kalmijn, Paul de Graaf and the other members of the seminar ‘Divorce in The Netherlands’ for many helpful comments on an earlier draft of this chapter.

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

341

incidence of divorce, a nation’s dominant belief system—reflecting the cultural dimension—and historical experiences with respect to familial behavior affect attitudes towards the legitimacy of divorce. In particular, we expect that de-familializing regimes will be characterized by higher public consent to divorce than familializing regimes. Countries which are characterized by de-familializing regimes are expected to create normative frames in which non-conformism to traditional family values is more easily accepted, whereas familializing regimes will create nor­ mative frames in which traditional family values are emphasized. These differences in regime characteristics, in turn, will be reflected in people’s values, beliefs and attitudes about divorce. In addition, we will argue that one of the important dimensions of the dominant belief system of a country, namely religion, will also affect how people feel about divorce. Finally, it is expected that how family life has been organized and val­ ued in the past will also have its impact on how people currently feel about divorce. Of course, differences in family values will not merely occur between countries. Many prior studies have shown norms, values and attitudes about the family to differ along social positions and ideological dimen­ sions within countries. Countries vary in composition according to social demographic and ideological characteristics of their citizens, and this may also affect how citizens of a given country evaluate divorce as justi­ fied. Therefore, we will also investigate the extent to which public con­ sent to divorce is related to individual characteristics which reflect com­ positional differences between countries. 2 Theory and hypotheses Relating attitudes to social structure poses an enduring challenge to so­ ciological thought because the nature and extent of the relationship be­ tween micro and macro levels of analysis need to be specified. Socio­ logical theories of attitudes follow two lines: structural and cultural. Structural theories assume that attitude similarity within groups stems from the common life situations they encounter as a result of their rela­ tionships to social structure and institutions. Structural theories contrast with cultural norm theories, which explain attitudes as inherent in groups’ values and transmitted during socialization. Cultural norm theo­ ries adduce social learning, social acceptance, and differential associa­ tion as determinants of attitudes. The two types of factors are not always easily separable, and explanations frequently draw on both (Kiecolt, 1988). Although admittedly the relationship between the two types of factors is complex, in what follows we will incorporate both kinds of

342

CHAPTER TWELVE

explanation when reasoning about between and within country variations in public consent to divorce. We start with the question, how differences between countries in their public’s consent to divorce can be explained by macro-level characteristics. 2.1 Contextual-level characteristics The general mechanisms through which contextual factors affect the atti­ tudes of individuals are complex. Here, we argue that contextual factors only indirectly exercise their influence on the attitudes of individuals (Arts & Gelissen, 2001). First, people must become accustomed to the contextual factors and their social situations via learning and habit for­ mation. Only if institutional and other contextual characteristics pertain­ ing to the family in general and to divorce in particular have been around for some time citizens have had the chance to gather individual and col­ lective knowledge necessary to act in the socially approved manner. Secondly, how people feel about the legitimacy of divorce is strongly af­ fected by the way they frame the situation. Those frames are shaped by the socio-economic and institutional context of the past. National laws and customs concerning divorce affect people’s definition of the situa­ tion at hand and pre-structure their ideas about the legitimacy of divorce. Frames describe rules that are thought to be a social fact and can serve as generalized standards whereby individuals eventually develop expecta­ tions for rewards in specific situations. As a consequence of beliefs about what is typically the case, expectations are formed about what one can legitimately claim ought to be the case. A similar mechanism is pro­ posed by framing theory: ‘Any stable state of affairs tends to become ac­ cepted eventually, at least in the sense that alternatives to it no longer readily come to mind’ (Kahneman & Tversky, 1981). Following this line of reasoning, a first macro-level characteristic we expect to affect people’s attitude towards divorce by means of habit for­ mation and framing is the extent to which divorce actually occurs within different countries. The divorce rate in a country may be related to pub­ lic consent to divorce in two competing ways. On the one hand, it can be argued that if the divorce rate is high in a country, people are more likely to experience or have experienced divorce themselves. Moreover, be­ cause divorce more frequently occurs around them, people may learn to accept divorce as a legitimate way out of an unsatisfying marriage. Fi­ nally, the divorce rate reflects, to some extent, each nation’s laws and customs regarding divorce (Trent & South, 1989): Higher rates reflect more liberal laws and liberal divorce laws are presumably positively re­ lated to public permissiveness of divorce. This leads to the hypothesis that a higher divorce rate will have a positive impact on the degree of

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

343

overall public consent to divorce. On the other hand, it can be argued that higher divorce rates in a country are associated with negative public views about the consequences of divorce (e.g., negative consequences for the children of divorcees), and this will lead to a higher degree of public rejection of divorce (Dafoe Whitehead, 1997). In addition, it can be argued that in countries where divorce it not common, the call for di­ vorce as a legitimate way out of an unsatisfying marriage may be stronger than in countries were divorce occurs more frequently, because in the former case more people are forced to stay in a marriage because divorce is not socially accepted. This leads to the hypothesis that a higher divorce rate will be negatively related to the overall level of con­ sent to divorce. It is not immediately clear which of the competing ex­ planations is more valid; therefore, we propose the conservative hypo­ thesis that the divorce rate will have a (positive or negative) effect on the public acceptance of divorce, ceteris paribus. Another explanation is offered by Esping-Andersen’s concept of wel­ fare regime. Esping-Andersen (1990) argues that when we focus on the principles embedded in welfare states, we discover that variations are structured by three highly diverse regime-types, each organized around its own discrete logic of organization, stratification and societal integra­ tion. We can identify three models, or ideal-types of welfare states: con­ servative, liberal, and social-democratic. Such a bold hypothesis met, of course, not only acclaim but also much criticism (see, for an overview of comments on Esping-Andersen’s typology, Arts & Gelissen (2002)). Es­ pecially feminist researchers pointed out that Esping-Andersen’s original discussion of welfare regimes seriously lacked attention to the family in general and to such characteristics as work and welfare incentives to women in particular (Sainsbury, 1994). Esping-Andersen (1999) recognized this shortcoming and responded with a more thorough discussion of gender and family issues within the framework of welfare regimes. Although the standard nuclear family is becoming extinct, it is, as Esping-Andersen (1999) observes still the linchpin of social policy in most welfare states. Although the family is a fundamental and integrated part of all welfare states, he argues that even among mature welfare states attention for the family varies in terms of family policies and, accordingly, that welfare states can be ranked on the regime-characteristics of de-familialization (which is closely related to his notion of de-commodification). De-familialization refers to the de­ gree to which households’ welfare and caring responsibilities are relaxed – either via welfare state provision, or via market provision. Rising di­ vorce rates imply more potentially precarious lone parent families, al­ most all of which are headed by women, and active de-familialization

344

CHAPTER TWELVE

reduces the risk for these families of falling into poverty. In other words, a high level of de-familialization ensures that women are able to set up independent households. Real welfare states can be typified according to the degree of de-familialization by which they are characterized, and the resulting classification comes close to the original Esping-Andersen’s typology of welfare states based on the de-commodification criterion (Esping-Andersen, 1990). According to Esping-Andersen, a high level of de-familialism, characterizes liberal welfare states; the same holds for the Social-democratic Scandinavian countries. On the other hand, low levels of de-familialization characterize the conservative welfare states and the Mediterranean welfare states. Here, we assume that welfare regimes also encompass, to some ex­ tent, national laws and customs with respect to divorce. Goode’s (1993) study of cross-national differences in divorce patterns and divorce laws uses a classification of countries which is remarkably close to EspingAndersen’s classification. Specifically, Goode considers differences be­ tween Western European countries, Nordic countries, Southern Euro­ pean Catholic countries and the Anglo-Saxon countries. This suggest that welfare regimes may also be, to some extent, ‘divorce regimes’. Specifically, unlike the Southern European Catholic countries, the West­ ern European countries have all permitted divorce for many generations. The Nordic countries have not only legally permitted divorce for a long time, but they even provide public programs for dealing with the conse­ quences of divorce. Finally, the Anglo-Saxon countries followed the common law tradition of viewing family disputes, and thus the problems of divorce, as adversarial proceedings in which the rights of the two par­ ties were to be decided on the basis of an accumulating body of individ­ ual cases, guided by folk principles (Goode, 1993). If welfare regimes also systematically differ in divorce patterns and legal traditions with re­ spect to divorce, we should expect that these differences will also have an impact on the extent to which people from different nations are per­ missive of divorce. Unfortunately, Esping-Andersen does not include post-communist countries in his analysis. Eastern European countries are characterized by a relatively high fertility along with high employment rates among mothers, which has been explained by a stronger emphasis on accom­ modating employment to childbearing through high levels of child-care services and family benefits. Thus, it can be argued that a high level of de-familialization, relatively similar to the Scandinavian welfare states, also characterizes the welfare states of the former East-bloc countries. All in all, in terms of de-familialization, the worlds of welfare capitalism appear to be bipartite, with the Social-democratic, Liberal, and Post­

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

345

communist welfare regimes being characterized by high, though varying levels of de-familialization in their welfare arrangements, and the Con­ servative and Mediterranean welfare states being characterized by lower levels of de-familialization. Following Knudsen and Waerness (1999: 171–172), we expect public opinion in nations that have experienced the collapse of the communist regime to be systematically affected by the experience and the legacy of the communist ideology. Countries of the Soviet Union had distinctive gender relations in legal and social welfare provisions. The regimes sup­ ported women as paid workers while sustaining traditional patterns of unpaid work in the family (Pascall & Manning, 2000). After the col­ lapse, when women in these nations felt free from the former dictate of double duties, support for women to stay at home grew. In addition, the domestic sphere and family life during the Communist Era, even by the double-burdened women came to be cherished as a site of psychological and moral refuge. And even if the political landscape has changed in the post-soviet era, this evaluation of the domestic sphere does not seem to have changed (Knudsen & Waerness, 1999). Typically, this return to the home was expressed as positive compensation after hard work in the la­ bor market and in the household. Therefore, it can be expected that fam­ ily values in former East-bloc nations differ from the traditional feminist position in the West. As Knudsen and Waerness (1999: 171–172) further point out, despite strong rhetoric and political attempts in the early Soviet era to under­ mine the historical functions of family and marriage, Socialist societies after World War II did not base their policies specifically on Marxist ideology in this area. Policies and practices more often seemed to have been guided by cultural and religious traditions. Against this background it is not obvious what patterns to expect on attitudes towards divorce. Since we interpret liberal attitudes as an expression of the general pro­ cess of individualization, one might expect former Soviet satellite na­ tions to have been shielded from these influences. If this was the case, attitudes towards marriage and divorce, too, should be more conserva­ tive in Eastern than in Western European countries. However, the countries of the former East-bloc probably do not form a homogeneous cluster with respect to the degree of de-familialization in social policy arrangements, their relative level of modernity and their level of acceptance of a capitalist culture and structure. For example, Mellens (1999) distinguishes between an Intermediate regime and a Post-totalitarian regime. The first regime consists of former communist countries in Central Europe, which are on the one hand relatively mo­ dern, but on the other hand less wealthy than the countries in Northern,

346

CHAPTER TWELVE

Western, and Southern Europe. In addition, a substantial number of peo­ ple in these countries still hold neo-communist views. The Post-totalitarian regime consists of countries, which are characterized by an in­ complete transition to a capitalist structure in the sense that it is not rooted in a firm legal and cultural structure (e.g., Russia). Female par­ ticipation in the economy is high in these countries, but this does not lead to the acceptance of non-traditional gender roles and non-traditional family values. If we combine all the arguments offered above we arrive at the fol­ lowing hypothesis: with respect to the differences in public consent to divorce among welfare regimes, we expect that consent will be highest in countries, which belong to the social-democratic welfare regime. The publics of Conservative, Mediterranean, Intermediate and Post-Totalitarian welfare regimes will show lower levels of consent, with the Con­ servative regime taking second place, the Mediterranean regime taking third place, the Intermediate regime taking fourth place, and the popula­ tions of the Post-totalitarian regime showing the lowest levels of public consent to divorce, ceteris paribus. Siaroff (1994) observes that the historically dominating religion of a country may play a part on its own in shaping characteristics of the wel­ fare state and people’s interpretations of it. Siaroff found, in agreement with previous findings, a direct relationship between Protestantism and increasing labor force participation. This relationship is explained by Protestantism’s greater stress on individual rights, and to the fact that Catholicism (and other traditional religions) has had stricter views that the place of women should be in the home. Such reasoning would sug­ gest that one would expect systematic cross-country differences accor­ ding to dominant belief systems, not just with respect to gender role atti­ tudes, but perhaps even more so concerning marriage. A reasonable hy­ pothesis is that typical Roman Catholic nations should have a culture less favorable toward divorce (Knudsen & Waerness, 1999). The Pope’s strict stand against abortion, contraception and divorce has been debated even in traditional Catholic nations. As in earlier periods, tensions often exist between the official doctrines of the Catholic church and the prac­ tice and opinion of its followers. Notwithstanding, Catholic nations are expected to be more robust in their resistance to modern changes in fam­ ily life, gender roles, and marriage. Thus, we would expect that the church’s strong defense of marriage as a sacred sacrament instituted by God will affect public attitudes about permissiveness with respect to family life. The same could be expected from nations with large Ortho­ dox populations, although the effect may be less strong as for countries with large Catholic populations. As Ware (1993) points out, Orthodoxy

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

347

also regards the marriage bond as in principle lifelong and indissoluble, and it sees the breakdown of marriage as a tragedy due to human weak­ ness and sin. But while condemning the sin, the church still desires to help suffering humans and to allow them a second chance. When, there­ fore, a marriage has entirely ceased to be a reality, the Orthodox church does not insist on the preservation of a legal fiction. Our first hypothesis with respect to religion at the aggregate level is that, all else being equal, the higher the percentage of Catholics or Or­ thodox persons is in a country, the lower the degree of public consent to divorce will be. Furthermore, since religious attendance is often found to be a better indicator for religiousness than simply asking people’s de­ nomination and because it will be especially the people who regularly at­ tend church who will oppose divorce, we furthermore expect at the ag­ gregate level that, all else being equal, the higher the degree of atten­ dance of religious services is in a country, the lower the degree of public consent to divorce will be. 2.2 Effects of individual-level variables There are two reasons why to incorporate individual characteristics into an explanation of public consent to divorce. In the first place, apart from differences between countries with respect to public consent to divorce, we like to explain individual differences in consent to divorce. More­ over, we like to consider the possibility of so-called compositional ef­ fects (Snijders & Bosker, 1999). If individual characteristics explain, to some extent, an individual’s consent to divorce and if these individual characteristics are unequally distributed across countries, then they also explain, to some extent the differences in public consent across coun­ tries. Several individual characteristics have been identified in the litera­ ture, which may affect an individual’s attitude towards divorce. Thorn­ ton (1985: 857–858) points out that the causal connection between mari­ tal dissolution experience and attitudes is an important theoretical issue. According to him, couples with strong moral objections to separation and divorce are probably less willing than others to use divorce in re­ solving an unsatisfactory marital relationship. There may even be a sub­ stantial number of persons who oppose separation and divorce so strongly that they would not themselves consider terminating a marriage. It is also likely that unpleasant marital relations and experience with a marital disruption lead to a re-evaluation of the acceptability of separa­ tion and divorce. Unhappy marriages may cause people who initially disapprove of divorce to decide that, at least in their own cases, termina­ tion of the marriage is not only acceptable, but perhaps the best alterna­

348

CHAPTER TWELVE

tive. It is likely that people who adjusted well to the divorce situation and who perhaps entered a happy second marriage have positive opi­ nions of divorce, which may be quite different from their attitudes when they were first married. Thus, it can be expected that people who have ever experienced a divorce are more likely to consent to divorce than persons who have never experienced a divorce. Thornton (1985) further discusses religious involvement as a source of traditional family values. Individuals with a high commitment to reli­ gious activities and values have been shown to have fewer divorces. It is also probable that they have less approving attitudes toward separation and divorce. One’s denomination may also play a significant role here. The Roman Catholic church has taken strong positions on several family issues, including divorce and remarriage, and this official church posi­ tion is probably reflected in the attitudes of individual church members. These considerations lead to the expectation that Catholics disapprove of separation and divorce more than the members of other religious de­ nominations. Yet another factor influencing consent to divorce, discussed by Thornton, lies in the magnitude of the relative advantage of marriage over life as a single person. Here, Thornton refers to Becker (1981) who has argued that, as the earning power and labor force participation— especially that of women—increase and fertility declines, gender based division of labor becomes less advantageous and the gain from marriage is reduced. Those persons with greater earning capacity—because of higher educational attainments or more experience in the labor force— have enhanced potential for financial independence and can afford to look more favorably on the prospects of terminating a marriage. Educa­ tion and labor force participation may also increase the acceptability of separation and divorce because they represent access to new ideas and values, which can be at odds with traditional orientations. Thus, we ex­ pect that higher educational attainment and labor force participation will positively affect consent to divorce. Whether a person lives in a couple relationship is also expected to af­ fect his or her stance towards the legitimacy of divorce. Persons who live in a couple relationship will most likely be dedicated to the relation­ ship and this will cause them to see divorce as less legitimate than per­ sons who are not involved in a couple relationship. Large families can also reduce the ability of individuals to support themselves, and children may be more valuable in the marriage that pro­ duced them, causing those with large families to be negatively disposed to the possibility of marital dissolution. In addition, it can be expected that if children are living in the household, this will also negatively af­

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

349

fect people’s disposition towards the legitimacy of divorce. People with children living with them will probably feel responsible for their chil­ dren and this feeling of responsibility will act as a barrier for seeing di­ vorce less easily as a legitimate way out of a marriage. The process of individualization implies threats to established social forms as known and cherished by the older generations. To the members of these generations, the fact that divorce is becoming more common does conflict with the established social forms, which they are accus­ tomed to. On the other hand, the younger generations in many of the countries we analyze are being socialized with the idea that divorce is a legitimate way to end an unhappy marriage. In the Scandinavian coun­ tries, for example, this situation for most young people is taken for granted. Nevertheless, for those in these countries who lived a great part of their lives in another era, when public norms and values associated with marriage were more traditional, today’s development might appear immoral and undesirable. It implies a grave challenge to the rationality and value basis in which they built their own lives. From such arguments follows the hypothesis that younger people will be more accepting of di­ vorce than older people, ceteris paribus. Finally, we expect that men will be more skeptical towards divorce than women. Bernard (1972) hypothesized that men benefit more from marriage than women. Specifically, she argued that men have higher re­ turns from marriage than women in terms of the quality of life and care by the spouse. Another argument for the presumed more positive stance towards divorce by women may lie in women’s increased earnings and labor market participation. Because these developments have reduced women’s gain from marriage more than for men, women may also look more favorably on the prospects of terminating a marriage than men. 3 Data, measurements, and method

3.1 Data To test the above hypotheses, we use data from the European Values Study (EVS). The project provides standardized cross-national measures of people’s perspectives and views in a broad range of important areas of life, such as religion, morality, family, marriage, gender roles, and poli­ tics. The first wave of surveys was conducted in 1981 in ten West Euro­ pean countries. Researchers from other countries later joined the project, resulting in a 26-nation study project. In order to monitor changes in va­ lues, a second wave of surveys was organized in 1990 in all European countries, including Switzerland and Austria and countries in Central

350

CHAPTER TWELVE

and Eastern Europe, as well as in the US and Canada. A third wave of surveys was conducted in 1999/2000, which covers 33 countries in Western, Central, and Eastern Europe (Halman, 2001). In this chapter, we use data from the third wave to test our hypotheses. Data come from the following 32 countries: France, United Kingdom, West Germany, East Germany, Austria, Italy, Spain, Portugal, the Netherlands, Belgium, Denmark, Sweden, Finland, Iceland, Northern Ireland, Ireland, Estonia, Latvia, Lithuania, Poland, the Czech Republic, Slovakia, Hungary, Ro­ mania, Bulgaria, Croatia, Greece, Russia, Luxembourg, Slovenia, Ukraine, and Belarus. Note that we omit Malta from the multivariate analyses. For this country, a reliable and comparable estimate of the di­ vorce rate was not available. 3.2 Measurements The dependent variable in this study—public consent to divorce—was measured by asking the following question to respondents: ‘Please tell me for the following statement whether you think it can always be justi­ fied, never be justified, or somewhere in between: Divorce’ (1 = never ­ 10 = always). Thus, higher scores on this scale indicate stronger appro­ val of divorce. At the individual level, we include the following demographic and social background characteristics. Sex has been precoded into 2 catego­ ries (women = 0 and men = 1). We include a respondent’s year of birth, and we also distinguish between 5 birth cohorts: 1901-1939, 1940-1949, 1950-1959, 1960-1969, and 1970-1985. The oldest birth cohort is the reference category. Four dummy variables indicate a respondent’s activ­ ity status: Whether one works more than 30 hours weekly, whether one works less than 30 hours weekly, whether one is a student, or whether one is not gainfully employed, the latter group being the reference cate­ gory. Characteristics of family life and marital history are operationally defined, first, by a dummy variable which indicates whether one lives in a couple relationship (reference is ‘not living in a couple relationship’), whether one has ever been divorced (reference is ‘not been divorced previously’), whether the respondent has children, whether the respon­ dent has children younger than 18 living with him or her, and whether the respondent has children who do not live in the household. For assess­ ing the impact of church attendance a dummy variable is included with those attending church less than monthly as the reference category. Reli­ gious denomination is indicated with 6 dummy variables: not religious (the reference category), Catholic, Protestant, Muslim, Orthodox, and other religion. Finally, educational attainment is operationally defined as the age at which one completed his or her full-time education.

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

351

Explanatory variables at the contextual or country-level include, first of all, the average divorce rate per 100 marriages over the period 1995­ 2000 (United Nations, 1999) and its quadratic term. The drawback of the divorce rate per 100 marriages is that only a few of those who divorce in any given year were also married in that year. Especially in countries where divorce may take several years, it is likely that none came from the same population (Trent & South, 1989). Our efforts to find a more valid measure, such as the number of divorces per 1000 married couples, or men or women, proofed unsuccessful, primarily because of the large number of countries for which data have to be available. To get a more precise estimate of the incidence of divorce in every country, we calcu­ lated the average divorce rate per 100 marriages over a longer period of time, namely over the period from 1995 to 2000 (if available). However, as this is not the best measurement of the incidence of divorce, our re­ sults should be interpreted cautiously. Furthermore, we include these ag­ gregated variables based on the EVS data, namely the percentage of people regularly attending church, the percentage of people who say that they are Catholic, and the percentage of people who say they are ortho­ dox. Because of substantial skewness in these variables, they are used here in logarithmic form. Finally, we include a classification of countries along welfare regime characteristics as proposed by Arts and Gelissen (2002). Two remarks must be made with respect to the classification we use here. First, we also include former East-bloc countries and we argue that these can be classified into an Intermediate and a Post-totalitarian regime type, as proposed by Mellens (1999). Second, we do not include a liberal regime type, since we lack countries, which can be labeled with­ out doubt as liberal (e.g., the United States). This means that we classify hybrid cases as the United Kingdom to the conservative regime. In these countries, the family plays an important role in the provision of welfare besides welfare provision by the market. In contrast, the role of the state in providing welfare to the family is limited (Kuhnle & Alestalo, 2000). Thus, these countries can be regarded as scoring relatively low in terms of de-familialization. We distinguish the following five regime-clusters of countries: 1) Social-democratic: Denmark, Finland, Iceland, Norway, and Sweden; 2) Conservative: Austria, Belgium, France, United King­ dom, Ireland, Northern Ireland, East Germany, West Germany and Lux­ embourg; 3) Mediterranean: Greece, Italy, Portugal, and Spain; 4) In­ termediate: Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovak Republic, and Slovenia, and 5) Post-totalitarian: Belarus, Bulgaria, Romania, Russia, and Ukraine. Descriptive statistics of the variables used in the analysis are given in Table 12.1.

352

CHAPTER TWELVE

Table 12.1 Descriptive statistics of variables Variable

Max.

38448 39753

1 0

10 1

5.540 .459

Cohort 1901-1939

39761

0

1

.242

Cohort 1940-1949

39761

0

1

.151

Cohort 1950-1959

39761

0

1

.191

Cohort 1960-1969

39761

0

1

.196

Cohort 1970-1985 Children living with respondent

39761

0

1

.221

39761

0

1

.413

Respondent has children (ref. = no) Children do not live with respondent

39050

0

1

.740

39761

0

1

.372

Not working

39576

0

1

.413

Working more than 30 hours weekly

39576

0

1

.460

Working less than 30 hours weekly

39576

0

1

.069

Enrolled in school

39576

0

1

.058

Church attendance (ref. < monthly)

39416

0

1

.312

Not religious

39761

0

1

.209

Catholic

39761

0

1

.472

Protestant

39761

0

1

.160

Muslim

39761

0

1

.008

Orthodox

39761

0

1

.132

Other religion

39761

0

1

.019

Ever divorced (ref. = no)

27104

0

1

.085

Lives in couple relationship (ref. = no) 39761

0

1

.659

Age of completing full-time education

38698

0

25

18.346

Conservative regime

38759

0

1

.289

Southern regime

38759

0

1

.137

Intermediate regime

38759

0

1

.293

Post-totalitarian regime

38759

0

1

% Church attendance (ln)

38759 –2.401

–.248

–1.418

.645

% Catholic (ln)

38759 –6.928

–.032

–1.887

2.172

% Orthodox (ln) Mean divorce rate per 100

38759

.668

.106

.192

Divorce rate squared

N

0

Mean

St. dev.

Min.

Approval of divorce Sex (ref. = women)

2.907

3.935

.177

38759 8.000 53.850 32.718 13.575 38759 64.000 2899.822 1254.754 841.030

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

353

3.3 Method We estimate several random coefficient models to estimate the effects of contextual and individual-level variables on individuals’ consent to di­ vorce. The random coefficient model is a probability model, which con­ ceives of the unexplained variation within groups and the unexplained variation between groups as random variability (Snijders & Bosker, 1999). For the present study, this means that not only unexplained varia­ bility between citizens of countries, but also unexplained variation be­ tween countries is regarded as random variability. Multilevel models have several advantages over standard OLS regression analysis with nested data. First, estimates derived from a multilevel model yield tests of hypotheses that are appropriately more conservative. Multilevel mo­ dels also allow for a decomposition of variance in the outcome of inter­ est into that part attributable to differences between individuals located in different contexts and that part related to variation between individu­ als within the same context. This decomposition of variance into be­ tween (attributable to context) and within (attributable to individuals) components yields some indication of the relative importance of vari­ ables from the different levels of aggregation on the outcome under con­ sideration. Furthermore, conditional on the measured covariates, each of these variance components can be decomposed into an explained and an unexplained portion (Teachman & Crowder, 2002). 4 Results In Figure 12.1, we present the average level of public approval to di­ vorce for the countries of the EVS. The highest levels of approval are found in Sweden, Denmark and Finland. Relatively high levels of support are further found in the Ne­ therlands, Iceland, France, Greece and Slovenia. On the other hand, two groups of countries show relatively low levels of support. First, there is the group of countries with a large Catholic population: Italy, Northern Ireland, and Ireland. Second, there is little support for the justification of divorce in several of the former East-Bloc countries, namely Poland (which is also strongly Catholic in orientation), Latvia, Lithuania, Hun­ gary, Romania, Croatia and Ukraine. To gain insight in the ways in which the explanatory contextual varia­ bles are related to overall levels of public approval to divorce, we re­ gressed the overall level of consent to divorce on the crude divorce rate, its quadratic term, the percentage of Catholics in the population, the ex­ tent of church attendance, and the clustering of countries according to welfare state regime characteristics. The results of these aggregate-level

354

CHAPTER TWELVE

regressions are presented in Table 12.2. We also provide plots of the various explanatory variables against the dependent variable, which gives insight how specific countries are positioned within the relation­ ship.

9

Average level of public approval to divorce

8

7

6

5

4

3 A U S B E LR B

CR

BUL

DK CZ

E ST EG

FI F

HU GR

LA T N IR L

IR L IC E

IT

L IT

PO L

NL

RO M

PO R

RUS

S

SLVN

S LV K

SP

UK

WG

UKR

Figure 12.1 Average levels of public approval of divorce, across countries.

Given the relatively limited number of countries, in the following we are more interested in the strength and direction of particular associations, rather than in strict significance tests. The scatter-plot of the divorce rate against the overall level of public approval to divorce suggests that the linear association between these variables is positive. Indeed, according to Model 1A, the linear effect of the crude divorce rate on the overall level of public consent to divorce is .026. To investigate the possibility of a non-linear association in the plot, we added a quadratic term of the crude divorce rate (Model 2A). The partial effect of the crude divorce ra­ te is now –.015 and the quadratic term for the crude divorce rate is .001; however, both effects are not significant (see Figure 12.3). As one would expect, Figure 12.4 suggests a strong negative association between the degree of church attendance in a country and the level of public approval of divorce. Estimation of the effect of church attendance in Model 3A corroborates this, the effect being –.660 (which is significant at the 1 per

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

355

Table 12.2 Aggregate-level regression of public consent to divorce on contextual characteristics Model Model Model Model Model Model Model Model 1A 2A 3A 4A 5A 6A 7A 8A Mean divorce rate (per .026** –.015 100 marriages) (.011) (.060) Mean divorce rate squared

–.055

–.019

(.063)

(.054)

.001

.001

.000

(.001)

(.001)

(.001)

Percentage church at- tendance (ln)

–.660***

–.732*

–.570

(.220)

(.414)

(.361)

–.075

–.009

.048

(.074)

(.110)

(.093)

Percentage Catholics (ln) Percentage Orthodox (ln)

–1.104 –1.412

.408

(.807) (.956)

(1.130)

Welfare regime: Social-democratic regime (ref.) Conservative regime

–.414*** –.877* (.400) (.483)

Southern regime

–1.376***–.640 (.471) (.595)

Intermediate regime

–1.883***–1.445***

Post-totalitarian regime

–2.093***–1.887*** (.447) (.633)

(.394) (.438)

Constant

4.782*** 5.297*** 4.708*** 5.489*** 5.743*** 5.480*** 7.116*** 6.075*** (.376) (.828) (.337) (.210) (.175) (.810) (.333) (.845)

32

32

R-squared .17 .18 .23 .03 .06 .33 .51 Standard errors in parentheses; * p < .1, ** p < .05, *** p < .01 (two-tailed test).

.64

Observations

32

32

32

32

32

32

cent significance level). According to Models 4A and 5A, the linear as­ sociation between the proportion of Catholics in a country and the level of consent is less pronounced, and the same holds for the proportion of Orthodox religious persons: none of the estimated effects reaches sig­ nificance. We also regressed the overall level of consent to divorce si­ multaneously on the crude divorce rate, its quadratic term, the proportion

356

CHAPTER TWELVE 10.0

Average level of public approval to divorce

9.0 S

8.0 DK

7.0 6.0 IT

5.0

IRL

CR POL

FI

NL ICE

SLVN

F

GR SP

EG CZ AUS

WG POR SLVK BUL NIRL

BELR UK B

EST

RUS LAT

ROM

LIT HU UKR

4.0 3.0 2.0 1.0 0

10

20

30

40

50

60

Mean divorce rate (per 100 marriages)

Figure 12.2 Linear relationship between public approval of divorce and the divorce rate

10.0

Average level of public approval to divorce

9.0 S

8.0 DK

7.0

NL ICE

SLVN GR SP

6.0 IT IRL

5.0

CRPOL

WG POR SLVK BUL NIRL ROM

LAT

FI F EG CZ AUS

BELR UK B RUS

EST

LIT HU UKR

4.0 3.0 2.0 1.0 0

10

20

30

40

50

60

Mean divorce rate (per 100 marriages)

Figure 12.3 Non-linear relationship between public approval of divorce and the divorce rate

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

357

10.0

Average level of public approval to divorce

9.0 S

8.0

DK FI ICE F BELR EG CZ

7.0 6.0 RUS

NL

AUS

WG

B

UK

EST

SLVN GRSP

POR SLVK IT NIRL CR ROM

BUL

5.0

LIT

HU UKR

LAT

IRL POL

4. 0 3.0 2.0 1.0 –2.5

–2.0

–1.5

–1.0

–0.5

0

Percentage church attendance (ln)

Figure 12.4 Linear relationship between public approval of divorce and church attendance

10.0

Average level of public approval to divorce

9.0 S

8.0 DK

7.0

FI

NL

GR

6.0

BELR EG UK

EST RUS BUL

5.0

UKR

LAT

ROM

SLVNICE F SP AUS B POR SLVK IT NIRL LIT IRL CR POL HU CZ WG

4.0 3.0 2.0 1.0 –7

–6

–5

–4

–3

–2

–1

0

Percentage Catholics (ln)

Figure 12.5 Linear relationship between public approval of divorce and the percentage Catholics

1

358

CHAPTER TWELVE 10.0

Average level of public approval to divorce

9.0 S

8.0

FI SLVN F

7.0 CZ

6.0

EG UKWG

IT

5.0

CR POL HU

AUS B SLVK NIRL

BELR EST LIT

GR

RUS BUL LAT

UKR ROM

4.0 3.0 2.0 1.0 –8

–6

–4

0

–2

Percentage Orthodox (ln)

Mean average level of public approval to divorce

Figure 12.6 Linear relationship between public approval of divorce and the percentage Orthodox persons

10.0 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 Social -democratic

Southern Conservative

Post-totalitarian Intermediate

Welfare regime

Figure 12.7 Public approval of divorce, across welfare regimes

2

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

359

of Catholics and Orthodox persons in a country, as well as the extent to which church attendance occurs (Model 6A). In this model, we only find evidence for a negative effect of church attendance on public approval of divorce. In addition, we tested the extent to which a classification of countries according to welfare regime types would yield expected results. The proposed classification of countries fits the data relatively well (Model 7A and Figure 12.7). The Conservative cluster of countries deviates sig­ nificantly from the Social-democratic cluster. As expected, lower levels of consent are also found in countries, which belong to the Mediterra­ nean cluster. Moreover, in accordance with our hypothesis, we find that the countries of the former East-bloc show the highest degree of disap­ proval of divorce. The findings also suggest a distinction within the for­ mer East-bloc cluster, namely between the former communist countries which are relatively modern and the Post-totalitarian countries which have made only an incomplete transition to capitalism. The latter cluster of countries shows the lowest level of public approval of divorce. In the final Model 8A we entered all prior contextual characteristics simultaneously to see whether the effects of the dummy-variables for the welfare regimes in Model 7A would be affected by controlling for the other contextual characteristics. In this model, the previously found dif­ ferences between the Conservative and Southern regime on the one hand and the social-democratic regime on the other hand have partially or completely disappeared, whereas the Intermediate and the Posttotalitarian regime still differ markedly from the Social-democratic re­ gime. The aggregate-level analyses discussed above do not take into ac­ count the possibility that compositional differences may also explain contextual-level effects. Adding individual background variables is a way to control for the compositional differences across countries – that is, differences in sex, age, educational attainment et cetera. If the regres­ sion coefficients of the contextual variables do not approach zero and/or lose their statistical significance when individual background character­ istics are controlled for, one can conclude that the contextual characteris­ tic has a unique effect on individual attitudes that is not accounted for by the compositional differences across countries (Kamano, 1999). To in­ vestigate the effects of compositional differences, we performed several random intercept regressions. The results of these multivariate analyses are presented in Table 12.3. Model 1B is equivalent to a one-way analysis of variance with ran­ dom effects. This model functions as a baseline model, because it gives estimates of the maximal within-context variance � e) and the between­

360

CHAPTER TWELVE

context variance� e), which exists in public consent to divorce. It also allows us to estimate the intraclass correlation, � u�� u– e), which is an indicator of the relative importance of context, which larger values indi­ cating a greater relative impact of context. Model 1B indicates that the grand mean for public consent to divorce is 5.633 with � u) =.88 and � u) = 2.76. These values indicate an intraclass correlation of .24 =.88/(.88 + 2.76). This is a modest intraclass correlation and indicates that most of the variation in public consent to divorce lies at the individual level or within countries, which represent the context in these analyses. In model 2B, we control for some basic demographic and status char­ acteristics of individuals. Model 2B shows, first of all, that men see di­ vorce as less legitimate than women. In addition, those who belong to the birth-cohort 1940-1949, 1950-1959, 1960-1969, and 1970-1985 are more likely to view divorce as legitimate compared to those who belong to the oldest cohort2; thus, as expected, younger cohorts appear to be more tolerant towards divorce than older cohorts. In addition, we find that those who are gainfully employed, either full-time or part-time, are more likely to consent to divorce than persons who are not gainfully em­ ployed. Those still enrolled in school are also more likely to approve of divorce than those who are not enrolled in school. According to Model 3B, those persons who belong to the oldest birth cohort are significantly less likely to approve of divorce than the mem­ bers of more recent birth cohorts. We hypothesized that a person’s mari­ tal status, his or her marital history and child status also would affect one’s opinion about the legitimacy of divorce. As Model 3B shows, those individuals who live in a couple relationship are not less likely to favor divorce than those who are not in a couple relationship. On the other hand, if persons have experienced divorce in the past, they are more tolerant about divorce than those who have never been divorced. Finally, Model 3B shows that the child status of a respondent does not affect one’s attitude towards divorce. In Model 4B, we also account for the influence of church attendance, denomination and educational attainment, in addition to the effects of demographic and status variables. Compared to Model 3B, we now find a significant difference between persons who live in a couple relation­ ship and persons who do not: the latter are more tolerant about divorce than the former. Furthermore, this model shows that those who regularly attend church are, as expected, less likely to see divorce as legitimate as those who do not regularly attend church. Furthermore, we observe that ______________ 2

Tests for the differences between the included dummy variables showed that the co­ hort 1960-1969 did not differ significantly from the cohort 1970-1985. The other effects of the consecutive dummy variables differed significantly at p < .05.

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

361

compared to those persons who see themselves as not religious, Muslims are least permissive about divorce, followed by those who reckon them­ selves to other religions. Catholics and Protestants are also less in favor of divorce than non-religious persons. However, there is no significant difference in consent to divorce between Orthodox religious persons and the non-religious. Finally, we observe that educational attainment has, in accordance with our prediction a positive effect on the attitude towards divorce. If we compare the R-squared values of Model 4B to the baseline Model 1B, we see that the compositional differences controlled for in Model 4B account for 10 per cent of the variance within countries. In addition, the compositional differences controlled for in Model 4B ac­ count for 15 per cent of the variance between countries. To what extent do contextual variables still have an effect on an individual’s stance towards divorce after we account for differences in the composition of the populations under study? To answer this question, we estimated Models 5B to 12B, in which we estimate the additive effect of several country-level characteristics on the average level of public con­ sent to divorce, holding constant for compositional differences. Note that since adding the contextual characteristics to the individual-level vari­ ables of Model 4B does not substantially alter the effects of these individual-level variables in Models 5B to 12B, we focus on the effects of the contextual variables in our discussion. In addition, as collinearity in the contextual-level part of the models might occur, we also estimate separate models, one for each contextual variable. Model 5B shows that the divorce rate has a weak positive effect on the public’s approval of divorce—controlling for compositional differences—but the effect is not significant. This result is contrary to our earlier finding of a significant positive effect in the aggregate-level re­ gression of public consent to divorce on the crude divorce rate. In Model 6B we add the quadratic term of the crude divorce rate. Both the partial effect as well as the departure from linearity are not significant. From Models 7B and 8B, we infer that neither the percentage Catho­ lics in a country affects the overall approval of divorce, nor the degree to which people go to church on a regular basis in a country. However, the results of Model 9B show that the percentage Orthodox religious persons is negatively related to the level of public approval of divorce. In Model 10B, we entered the divorce rate, its quadratic term, the percentage of Catholics and Orthodox persons and the degree of church attendance in a country. Again, the only contextual characteristic which is directly re­ lated to public approval of divorce is the percentage of Orthodox reli­ gious persons in a country. Model 11B includes the clustering of coun­

362

CHAPTER TWELVE

Table 12.3 Parameters from random intercept models for public consent to divorce Model 1B Model 2B Model 3B Model 4B Individual-level characteristics: Sex (ref.=women) Cohort 1901-1939 (ref.)

Cohort 1940-1949

Cohort 1950-1959

Cohort 1960-1969

Cohort 1970-1985

Employment status (ref.=not working)

Working more than 30 hours weekly

Working less than 30 hours weekly

Enrolled in school

Lives in couple relationship (ref.=no)

Ever divorced (ref.=no)

Children living with respondent (ref.=no)

Respondent has children (ref.=no)

Children do not live with respondent (ref.=no)

Church attendance (ref.=less than monthly)

Denomination (ref.=not religious)

Catholic

Protestant

Muslim

–.144*** (.029)

–.104*** (.036)

–.272*** (.036)

.711*** (.049) .925*** (.049) 1.066*** (.049) 1.155*** (.050)

.628*** (.056) .802*** (.064) .929*** (.070) .832*** (.083)

.469*** (.056) .487*** (.064) .564*** (.070) .455*** (.083)

.409*** (.037) .490*** (.061) .755*** (.072)

.454*** (.045) .508*** (.074) .799*** (.234) –.074 (.052) .740*** (.061) –.152 (.231) –.199 (.225) –.152 (.234)

.301*** (.045) .354*** (.073) .672*** (.236) –.141*** (.052) .643*** (.060) –.089 (.236) –.177 (.229) –.130 (.239) –1.243*** (.042) –.238*** (.060) –.161** (.071) –1.142*** (.200)

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

363

Table 12.3 (continued) Model 5B Model 6B Model 7B Model 8B Model 9B Model 10B Model 11B Model 12B –.272*** (.036)

–.272*** (.036)

–.272*** (.036)

–.272*** (.036)

–.272*** (.036)

–.272*** (.036)

–.273*** (.036)

–.273*** (.036)

.470*** (.056) .488*** (.064) .565*** (.070) .455*** (.083)

.469*** (.056) .487*** (.064) .564*** (.070) .454*** (.083)

.470*** (.056) .487*** (.064) .565*** (.070) .455*** (.083)

.469*** (.056) .486*** (.064) .563*** (.070) .454*** (.083)

.469*** (.056) .485*** (.064) .560*** (.070) .452*** (.083)

.469*** (.056) .485*** (.064) .561*** (.070) .452*** (.083)

.468*** (.056) .485*** (.064) .562*** (.070) .455*** (.083)

.468*** (.056) .484*** (.064) .560*** (.070) .454*** (.083)

.302*** (.045) .355*** (.073) .674*** (.236) –.140*** (.052) .642*** (.060) –.088 (.236) –.180 (.229) –.129 (.239) –1.242*** (.042)

.301*** (.045) .354*** (.073) .672*** (.236) –.139*** (.052) .642*** (.060) –.088 (.236) –.178 (.229) –.129 (.239) –1.243*** (.042)

.301*** (.045) .355*** (.073) .673*** (.236) –.141*** (.052) .642*** (.060) –.089 (.236) –.178 (.229) –.129 (.239) –1.242*** (.042)

.301*** (.045) .354*** (.073) .674*** (.236) –.140*** (.052) .643*** (.060) –.087 (.236) –.179 (.229) –.128 (.239) –1.243*** (.042)

.300*** (.045) .354*** (.073) .668*** (.236) –.139*** (.052) .643*** (.060) –.072 (.236) –.191 (.229) –.115 (.239) –1.242*** (.042)

–.230*** (.060) –.153** (.071) –1.142*** (.200)

–.232*** (.060) –.158** (.071) –1.142*** (.200)

–.234*** (.060) –.159** (.071) –1.143*** (.200)

–.245*** –.257*** –.254*** –.260*** –.265*** (.060) (.060) (.060) (.060) (.060) –.163** –.180** –.181** –.202*** –.202*** (.072) (.072) (.072) (.072) (.072) –1.140*** –1.131*** –1.131*** –1.137*** –1.134*** (.200) (.200) (.200) (.200) (.200)

.300*** .299*** .299*** (.045) (.045) (.045) .354*** .351*** .351*** (.073) (.073) (.073) .665*** .660*** .660*** (.236) (.236) (.236) –.137*** –.140*** –.140*** (.052) (.052) (.052) .642*** .643*** .643*** (.060) (.060) (.060) –.071 –.093 –.082 (.236) (.236) (.236) –.192 –.168 –.178 (.229) (.229) (.229) –.114 –.133 –.124 (.239) (.238) (.239) –1.241*** –1.241*** –1.240*** (.042) (.042) (.042)

364 Table 12.3

CHAPTER TWELVE

(continued)

Orthodox Other religion Age of completing full-time education

Model 1B Model 2B Model 3B Model 4B –.074 (.074) –.449*** (.132) .087*** (.005)

Contextual characteristics: Mean divorce rate (per 100 marriages) Mean divorce rate squared Percentage church attendance (ln) Percentage Catholics (ln) Percentage Orthodox (ln) Welfare regime (ref.=social-democratic) Conservative regime Southern regime Intermediate regime Post-totalitarian regime Constant

Observations Number of countries R-squared within R-squared between

5.633*** 4.679*** (.157) (.139)

4.910*** 4.309*** (.130) (.163)

37446 37263 25260 24470 32 32 32 32 .00 .05 .05 .10 .00 .16 .20 .15 .88 .76 .57 .60 u 2.76 2.69 2.69 2.61 e Wald X2 1294.51 1867.53 1211.14 2621.46 Df 0 8 13 20 Standard errors in parentheses. * p < .1, ** p < .05, *** p < .01 (two-tailed test)

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

Table 12.3

365

(continued)

Model 5B Model 6B Model 7B Model 8B Model 9B Model 10B Model 11B Model 12B –.078 –.076 –.075 –.067 –.041 –.040 –.055 –.046 (.073) (.074) (.074) (.074) (.074) (.074) (.074) (.074) –.445*** –.447*** –.448*** –.450*** –.459*** –.458*** –.471*** –.472*** (.132) (.132) (.132) (.132) (.132) (.132) (.132) (.132) .087*** .087*** .087*** .087*** .087*** .087*** .087*** .087*** (.005) (.005) (.005) (.005) (.005) (.005) (.005) (.005) .004 (.007)

–.032 (.042) .001 (.001) –.081 (.166) .047 (.054) –1.786*** (.599)

–.051 (.049) .001 (.001) –.198 (.324) –.020 (.086) –2.027*** (.758)

4.168*** (.281)

4.622*** (.594)

4.197*** (.282)

4.398*** (.193)

4.493*** (.176)

4.962*** (.644)

24470 32 .10 .17 .54 2.61 2622.44 21

24470 32 .10 .17 .57 2.61 2622.77 22

24470 32 .10 .16 .58 2.61 2621.81 21

24470 32 .10 .15 .62 2.61 2622.05 21

24470 32 .10 .28 .63 2.61 2631.01 21

24470 32 .10 .31 .61 2.61 2633.39 25

–.004 (.052) .000 (.001) –.047 (.346) .040 (.090) –1.024 (1.094) –.877** –.921** (.365) (.464) –.648 –.469 (.431) (.572) –1.560*** –1.555*** (.360) (.422) –1.912*** –1.407** (.410) (.609) 5.432*** 5.459*** (.330) (.823) 24470 32 .10 .61 .60 2.61 2655.40 24

24470 32 .10 .64 .59 2.61 2657.20 29

366

CHAPTER TWELVE

tries according to welfare state regimes. Holding constant for composi­ tional differences between countries, we find that in comparison to coun­ tries, which belong to the Social-democratic regime type, the public’s consent to divorce is lowest in countries, which belong to the Intermedi­ ate regime and the Post-totalitarian regime. Countries, which can be rated among the conservative welfare states, take a position in between the Social-democratic and Post-totalitarian regime. Note that countries, which belong to the Mediterranean type of welfare states, do not deviate significantly from the Social-democratic regime type. This, presumably, is the consequence of controlling for compositional differences with re­ spect to religion and church attendance. In terms of explained variance between countries, the classification of countries according to welfare regime characteristics performs relatively well: It explains 61 per cent of the variance between countries in overall levels of public approval to­ wards divorce. Of course, this is not surprising given the fact that there is far more within-country variation than between-country variation in the dependent variable; then, classifying 32 countries into a relatively large number of regime types and using these to explain the aggregatelevel variation in the dependent variable will lead to a high explained variance. In the final Model 12B, we entered all contextual characteristics si­ multaneously. The previously found differences between the Conserva­ tive, Intermediate and Post-totalitarian regime on the one hand and the Social-democratic regime on the other remain, but the differences be­ tween the Southern regime, on the one hand, and the Social-democratic regime, on the other, now have disappeared. None of the other contex­ tual variables significantly affects the degree of public permissiveness of divorce. 5 Conclusion and discussion The purpose of this chapter was, first, to demonstrate that cultural char­ acteristics as well as welfare state regime characteristics of a country play a significant role in affecting individual’s approval of divorce, and, second, to show that individual background characteristics are also im­ portant determinants of the public consent to divorce. Below we sum­ marize our findings. But first we need to draw attention to some short­ comings of our analysis. The dependent variable of this study, the degree of approval to di­ vorce, is measured only by one item and this, of course, constitutes a far from comprehensive measure of the acceptance of divorce. How crossnationally valid and reliable our measure is, is due to its one-item nature

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

367

very difficult to assess. Moreover, the cross-sectional design of our study makes it impossible to reach causal conclusions about the association be­ tween variables. For example, we cannot assess whether individuals with strong moral objections to separation and divorce are less willing than others to divorce, or that the experience with divorce leads to a reevalu­ ation of the acceptability of divorce (Thornton, 1985). Here, we assumed that the latter explanation holds true, but ideally we would need prospec­ tive data to reach definite conclusions about this causal connection. Likewise, it could be argued that differences in policies and divorce rates do not affect how permissive people are about divorce, but that public permissiveness affects policy development and the prevalence of di­ vorce. Here we have chosen to investigate the former relationship, but since both interpretations are valid, future research also needs to con­ sider the effects of public values on policy outcomes. Finally, we already noted that our contextual measures suffer from shortcomings, which would necessitate the collection of more reliable macro-level data. For example, we used the number of divorces per 100 marriages, but a more valid measure would have been the number of divorces per 1000 married couples, or married men or women, but this base figure is not easy to ob­ tain for a relatively large number of countries (Goode, 1993). More reli­ able macro-level data would give more substance to future analyses of the impact of cultural and structural difference on individual attitudes. In an aggregate-level regression analysis we first examined the effects of the divorce rate, the percentage of Catholics and Orthodox persons, the extent of church attendance in a country and the type welfare regime on the overall level of public approval of divorce. Three of the included contextual characteristics were significantly related to the public’s ap­ proval of divorce: The extent of church attendance, the divorce rate and the type of welfare regime. In accordance with our hypotheses, we found that as church attendance is more common in a country, people are less likely to see divorce as legitimate. In addition, we found that the com­ monality of divorce in a country matters for people’s tolerance about this issue: as divorce occurs more often in a country, people tend to become more tolerant towards it. However, based on the findings of the subse­ quently estimated random intercept models, we must conclude that the aggregate-level effects of church attendance and the divorce rate are sooner due to compositional differences rather than that they are true contextual effects: When controlled for compositional effects, the linear effects of church attendance and divorce rate on public approval of di­ vorce disappeared. In addition, we found a clear ordering of countries according to the classification proposed by Esping-Andersen (1990, 1999) and its exten­

368

CHAPTER TWELVE

sion to the former East-bloc countries. The publics of countries, which can be described as Social-democratic—with a high degree of defamilialization—showed a high acceptance of divorce. The populations of other types of welfare states showed a lower level of consent to di­ vorce, in the following ordering: Mediterranean welfare states, Conser­ vative welfare states, Intermediate welfare states and Post-totalitarian welfare states. Thus, our findings suggest that the degree of de­ familialization is not linearly related to the public acceptance of divorce as Esping-Andersen argues, especially if we extend Esping-Andersen’s original classification to the former East-bloc countries. Then, countries, which are at first sight relatively similar in levels of de-familialization (Pascall & Manning, 2000), show differing degrees of public consent to divorce. In particular, Social-democratic welfare states which can be characterized as highly de-familialized show high levels of public ap­ proval of divorce, whereas the Intermediate and Post-totalitarian coun­ tries, which can also be characterized as highly de-familialized take a less favorable stance towards divorce. All in all, the one-dimensional approach with the de-familialization criterion is quite insightful, but our analysis also suggest that other national characteristic, such as a nation’s specific historical developments—as in the case of the former East-bloc countries—or the prevalence of divorce may more directly affect public permissiveness about divorce. After all, the differences found between the Conservative and Southern welfare regimes on the one hand and the Social-democratic regime on the other substantially weakened or even disappeared once we controlled for other contextual characteristics. Thus, differences in the permissiveness of divorce between the popula­ tions of these regimes exist because these regimes also differ systemati­ cally in the prevalence of divorce and the other contextual characteristics and not so much because they differ in the degree to which households’ welfare and caring responsibilities are relaxed via state or market provi­ sion. In this sense, these welfare regimes can better be denoted as ‘di­ vorce regimes’. We then argued that compositional differences across countries—that is differences in sex, age, the level of educational attainment, or the de­ gree to which individuals in a country have experience divorce themselves—might also account for differences in levels of acceptance of di­ vorce. When analyzed in a multivariate random intercept regression, several of the demographic and social background characteristics showed significant effects, which were according to expectation. Women appeared to be more accepting of divorce than men. Further­ more, members of younger birth cohorts were more accepting of divorce than members of older birth cohorts. Whether this is an effect of age or a

CROSS-NATIONAL DIFFERENCES IN PUBLIC CONSENT TO DIVORCE

369

cohort effect could not be investigated with the data at hand. Several of the work status variables—as well as variables relating to family life— also significantly affected individual’s attitude towards divorce. People who are gainfully employed and students were more approving of di­ vorce than people who are not gainfully employed. Moreover, people living in a couple relationship were found to be less approving of di­ vorce than single people, whereas if a person has ever experienced di­ vorce one is more likely to approve of divorce. We furthermore found that highly schooled individuals are more approving of divorce than lowly schooled individuals. Finally, religious variables were also signifi­ cant predictors of the non-acceptance of divorce. As persons attend church more frequently, they are less approving of divorce. Moreover, when compared to persons who do not belong to a religion, Catholics, Islamic people, Protestants and persons belonging to other religions are all less tolerant about divorce. These compositional differences also explained a considerable amount of variation in an individual’s level of approval of divorce, not only within but also between countries. This suggests that compositional differences across countries are also important determinants of crossnational differences in public consent to divorce. Here, we found that compositional differences explain the aggregate-level effects of church attendance and the divorce rate on public permissiveness of divorce. In addition, once the compositional differences between countries were held constant, we found that the proportion of Orthodox persons was negatively related to the extent of public acceptance of divorce. Finally, after controlling for these compositional effects, the contextual effects of the type of welfare regime and a country’s historical experience with re­ spect to family life arrangements found in the aggregate-level regression were also largely found in these analyses. This suggests that these con­ textual variables have a unique effect on individual attitudes towards di­ vorce. But the final question, left to be answered by future research is whether this effect will also be detected if more direct measures of these characteristics are used. References Arts, W. A. & J. Gelissen 2001. Welfare States, Solidarity, and Justice Principles: Does the Type Really Matter? Acta Sociologica 44: 283–300. —— 2002. Three Worlds of Welfare Capitalism or More? A State-of-the-Art Report. Journal of European Social Policy 12: 137–158. Becker, G.S. 1981. A Treatise on the Family. Cambridge: Harvard University Press. Bernard, J. 1972. The Future of Marriage. New York: World Publishing. Dafoe Whitehead, B. 1997. The Divorce Culture. New York: Alfred Knopf. Esping-Andersen, G. 1990. The Three Worlds of Welfare Capitalism. Oxford: Polity Press.

370

CHAPTER TWELVE

—— 1999. Social Foundations of Post-Industrial Economies. Oxford: Oxford University Press. Goode, W.J. 1993. Worlds Changes in Divorce Patterns. New Haven: Yale University Press. Halman, L. 2001. The European Values Study: Source Book of the 1999/2000 European Values Study Surveys. Tilburg: EVS, WORC, Tilburg University. Kahneman, D. & A. Tversky 1981. The Framing of Decisions and the Psychology of Choice. Science 211: 453–458. Kamano, S. 1999. Comparing Individual Attitudes in Seven Countries. Social Science Research 28: 1–35. Kiecolt, K. J. 1988. Recent Developments in Attitudes and Social Structure. Annual Review of Sociology 14: 381–403. Kiernan, K. 2000. European Perspectives on Union Formation. Pp. 40-58 in L. J. Waite, C. Bachrach, M. Hindin, E. Thomson & A. Thornton (eds.). The Ties that Bind: Perspectives on Marriage and Cohabitation. New York: Aldine de Gruyter. Knudsen, K. & K. Waerness 1999. Reactions to Global Processes of Change: Attitudes toward Gender Roles and Marriage in Modern Nations. Comparative Social Research 18: 161– 195. Kuhnle, S. & M. Alestalo 2000. Introduction: Growth, Adjustments and Survival of European Welfare States. Pp. 3–18 in S. Kuhnle (ed.), Survival of the European Welfare State. London & New York: Routledge. Lesthaeghe, R. & D. Meekers 1986. Value Changes and the Dimensions of Familialism in the EEC. IPD-Working Paper. Brussels: Vrije Universiteit. Mellens, M. 1999. Uniformity and Diversity Defined. Pp. 33–44 in J. de Beer & L. van Wissen (eds.), Europe: One Continent, Different Worlds: Population Scenarios for the 21st Cen­ tury. Dordrecht: Kluwer Academic Publishers. Palomba, R. & H. Moors 1998. Population, Family, and Welfare: A Comparative Survey of European Attitudes. Oxford: Clarendon Press. Pascall, G. & N. Manning 2000. Gender and Social Policy: Comparing Welfare States in Cen­ tral and Eastern Europe and the Former Soviet Union. Journal of European Social Policy 10: 240–266. Sainsbury, D. 1994. Gendering Welfare States. London: Sage Publications. Siaroff, A. 1994. Work, Welfare and Gender Equality: A New Typology. Pp. 82–100 in D. Sainsbury (ed.), Gendering Welfare States. London: Sage. Snijders, T. & R. Bosker 1999. Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling. London: Sage. Svallfors, S. 1997. Worlds of Welfare and Attitudes to Redistribution: A Comparison of Eight Western Nations. European Sociological Review 13: 283–304. Teachman, J. & K. Crowder 2002. Multilevel Models in Family Research: Some Conceptual and Methodological Issues. Journal of Marriage and the Family 64: 280–294. Thornton, A. 1985. Changing Attitudes toward Separation and Divorce: Causes and Conse­ quences. American Journal of Sociology 90: 856–872. Trent, K. & S. J. South 1989. Structural Determinants of the Divorce Rate: A Cross-Societal Analysis. Journal of Marriage and the Family 51: 391–404. United Nations 1999. Demographic Yearbook. New York: United Nations.

CHAPTER THIRTEEN

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE VEERLE DRAULANS & LOEK HALMAN 1 Introduction It is almost commonplace to state that the importance of religion has di­ minished both in society and for the individual. A large body of litera­ ture has developed yielding evidence that religion in general and churches and church leaders in particular have lost their once dominant and prominent position in contemporary Europe. However, although these ideas are commonly shared among European scholars, some Americans are less convinced. American studies often yield opposite re­ sults from European studies, and some Americans therefore conclude that all European ideas about secularization are simply wrong (Stark, 1997, 1999; Stark & Iannaconne, 1994), or conclude that the European developments with regard to religion are an exceptional case (Martin, 1996; Davie, 2002). Most often evidence in Europe comes from declin­ ing levels of church attendance and changes in traditional religious be­ liefs. In the USA, levels of church attendance and adherence to religious beliefs have not changed dramatically over the last decades, and the claim that religion is decreasingly popular cannot be substantiated in the USA. European sociologists therefore consider the USA the exceptional case. This is not the place to rehash this debate, particularly not because we confine ourselves in this chapter to contemporary Europe where the fallen levels of church attendance and religious beliefs seem to suggest that indeed religion and churches are no longer the most powerful social forces guiding people’s lives. A classic explanation of this gradual de­ cline of the significance of religion is that the various societal spheres have become separated from one another and became autonomous spe­ cialized domains with their own logic, their own rationale and own set of guiding principles, norms and modes of conduct. In such a functionally differentiated society, religion is just one ‘subsystem alongside other subsystems’ (Dobbelaere, 1993: 24; Wilson, 1982, 1998). Such a decline in the importance of religion is further presumed to be closely linked with a process of individualization, which can be described as the social

372

CHAPTER THIRTEEN

and historical process by which people gradually become liberated from structural and social constraints (Ester et al., 1994: 7; Beck, 1992: 2). In­ dividualization is as far reaching as differentiation, for it implies that the authority of churches is seriously questioned. It is an ongoing process and can be traced back to the time of the Renaissance and Reformation when several Protestant theologies accentuated that in the personal rela­ tionship of men and women with God, mediation of institutional churches was no longer a necessity. The prevalent view is that individu­ alization indicates that people have become freer and more autonomous individuals, who decide for themselves and whose religious beliefs are more and more dependent upon personal desires, ideas and preferences. To this can be added the process of rationalization, which made the world more and more controllable and calculable. The scientific worldview that developed made the sacred elements of traditional faith super­ fluous and the necessity of religion and churches to provide meanings and interpretations questionable. All this is not to say that religiosity as such is necessarily on the de­ cline. What is on the decline is the significance of institutional religion and churches. Today’s people’s beliefs and religious convictions are in­ creasingly grounded in private interests and considerations, and not so much on what is offered or provided by collective authorities. And being a private affair implies diversity in convictions and beliefs because some will accept the traditional religions and their beliefs and practices, whereas others will experiment with new forms, while others will reject them. This issue of religious pluralism has been less prominent and often neglected in the discourses and studies on religious changes that are con­ fined to religious developments in terms of declining levels of religious beliefs and practices. The decline of traditional, institutional religion is often assumed to have had major consequences also for morality in society. The moral guidance of religion and the churches have come under strong pressure, particularly with regard to sexual and ethical issues. However, the issue of the impact of religion on morality has been debated for a long time. For example, MacIntyre strongly opposed the suggestion that the changes in moral and social life are a consequence of a religious decline. According to him ‘it is not the case that men first stopped believing in God and in the authority of the church, and then subsequently started behaving differently.(...) Social changes and with it moral change is chronologically prior to the loss of belief effected by intellectual argu­ ment, except where a very small minority are concerned’ (MacIntyre, 1967: 54–55). Following such arguments, moral decline would imply re­ ligious decline. Recently, Rodney Stark stimulated this debate again and

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

373

argued that the ‘functionalist ‘law’ that religion sustains the moral order must be amended’ (Stark, 2001: 619). He concluded that religion has this effect only ‘as it is based on belief in powerful, active, conscious, morally-concerned gods. Contrary to Durkheim’s claims, participation in religious rituals per se has little independent impact on morality and none when done on behalf of gods conceived as unconscious essences, or as conscious gods of small scope and lacking moral concerns’ (Stark, 2001: 619). Since European Christian tradition offered such a morally concerned God, it can be assumed that moral beliefs were once indeed strongly un­ der the influence of religion. The actual decline of the significance of re­ ligion implies that rigid moral standards, imposed by religious leaders are no longer accepted as taken for granted. Instead each individual be­ comes his or her own moral guide. As a consequence, diversity in moral convictions and beliefs also will have developed. Some people prefer to remain strict and will follow the moral guidelines of the authorities, while others are more lenient towards all kinds of behaviors formerly re­ jected by the churches. Thus, according to the dominant view, in modern society religion and church leaders are decreasingly able to dictate people’s decisions in eve­ ryday life. The taken for granted status in consciousness and beliefs has disappeared and in secularized society, religious and moral beliefs and practices are said to be dependent upon personal desires and preferences and constantly open to debate, reformulation and change. In this chapter we further explore these claims that contemporary (European) society is religiously and morally pluralistic. Using the data from the European Values Study, the patterns of religious and moral plu­ ralism will be explored. Of course the pattern will reveal differences and similarities in Europe. Analyses of the previous EVS data collected in 1981 and 1990 yielded evidence of remarkable similarities and differ­ ences in the religious and moral profiles of European societies. Such dif­ ferences seem to indicate that there is not one unique trajectory of reli­ gious change but that the trajectory has many faces and did not advance in all European societies in a similar speed and pace (Halman & de Moor, 1994; see also Dobbelaere, 1995). Religious beliefs and activities seem to have declined faster and more extensively in Northern Europe than in Southern Europe. These findings have been attributed to 1) dif­ ferences in longstanding cultural and religious traditions, 2) differences in church-state relationships, and 3) the particular (religious) situation and circumstances that emerged from it (Martin, 1978a; Therborn, 1995; Davie, 2002). In this regard, Catholic and Protestant traditions seem to have produced enduring national cultures that persist nowadays and that

374

CHAPTER THIRTEEN

have made secularization an ‘uneven process’ in Europe. The theoretical explanation for this uneven process is partly found in the theological dif­ ferences between Catholicism and Protestantism. Also David Martin (1978a) emphasized the crucial role of these religious traditions in the secularization in Europe, but he added the issue of the degree of compe­ tition between denominations in a country to understand and clarify the degree to which secularization has proceeded in a society. In the next section we will elaborate on his secularization perspective and formulate some expectations about differences and similarities in degrees of reli­ gious and moral pluralism in Europe. Theories of cultural modernization provide other explanations for the religious decline and offer suggestions for expecting varieties in degrees to which European societies are morally and religiously pluralistic. One of these theories stresses the importance of economic development and increased levels of existential security for the loss of the social signifi­ cance of religious institutions, activities and consciousness. It is argued that because in economically advanced societies modern welfare states have developed, the need ‘for the reassurance that has traditionally been provided by absolute belief systems’ (Inglehart, 1997: 80) has gradually declined. Thus, increasing levels of prosperity and economic welfare provide high levels of existential security that is counter-parted with a religious decline. Consequently, differences in economic situations are likely to produce different religious and moral profiles and it can be ex­ pected that religious and moral pluralism will be more widespread in these economically more advanced and prosperous countries. Also characteristic of modern society is that it is more and more part of a global community. Fuelled by technological developments and in­ novations in telephone and telecommunication, the spread and popular­ ity of computers, but also by the increased mobility of major companies and people, and the growing exposure to television, radio, video and movies et cetera, there is an intensification of worldwide social relations and flows of information. The modern world is regarded increasingly as a ‘global village’, where people encounter a great variety of alternative cultural habits and a broad range of lifestyles and modes of conduct. Globalization makes people aware of an expanding range of beliefs and moral convictions and thus with a plurality of choices. Because it has been argued that secularized people are liberated from the constraints imposed by the church and institutional religion, globalization implies that people can pick and choose what they want from a global religious and cultural marketplace. Globalization, thus, may be considered condu­ cive to religious and moral pluralism and differences in the degree to

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

375

which societies are exposed to the forces of globalization are likely to eventuate in different degrees of religious and moral pluralism. Of course, differences in religious and moral profiles may also be ex­ plained simply from differences in population parameters. For example, younger people are often found to be more secular compared with older people and lower educated people to be more religious than higher edu­ cated people. Countries may not only differ in terms of affluence and prosperity or in religious traditions, but also in levels of education and age structure, et cetera. Thus, the explanation of differences in religious and moral pluralism might be attributed to such differences in composi­ tion of the populations. Furthermore, previous studies yielded evidence that value orien­ tations change as a result of generation replacement. Because the opin­ ions, preferences, values and convictions of younger generations differ from the opinions et cetera, of older people, the gradual replacement of older generations by younger generations ultimately changes society. Young generations appear often more secular, and thus generation repla­ cement will imply a gradual shift towards a more secular society. We will investigate if generations also differ in levels of religious and moral pluralism. Finally, we tentatively investigate the longitudinal developments with regard to religious and moral pluralism by comparing the degrees of plu­ ralism in the three waves of the EVS project. These analyses can only be done on a limited number of countries, because the first EVS wave of 1981 was confined to Western European countries only, while the sec­ ond wave of 1990 did not include all societies included in the 1999 sur­ vey either. 2 Theoretical notions The decline of institutional religion and churches is often described in terms of secularization. This process has been subject of a large number of studies and publications. One important study is David Martin’s (1978a) attempt to develop a general theory of secularization. In his study he concluded that the European societies follow very different tra­ jectories of religious change that he attributed to the peculiarities and particularities of the various cultural heritages of societies. His interpre­ tation of Europe’s religious landscape was based on two dimensions: 1) the interaction between religion and national, regional, and sub-cultural awareness (centre/periphery); and 2) power and countervailing power. ‘The relationship of religion to power arises because it is not only the bearer of identities but a source of legitimacy and of philosophies sup­

376

CHAPTER THIRTEEN

porting legitimacy’ (Martin, 1978b: 235). The core of his analysis is constructed by attitudes towards unity or ideological monopoly versus pluralism. He distinguishes 5 types or clusters of countries in Europe. A first cluster includes countries with a dominant Protestant tradition, dominant in the sense that Catholic minorities are either not present or have entered at a moment that Protestant tradition had settled already. Scandinavia, England, and USA are examples. Countries with a Protes­ tant majority and substantial Catholic minorities are in a second cluster. Martin refers to the Netherlands, Germany, and Switzerland. A third cluster contains countries in which Catholicism historically dominates (e.g., Portugal, Italy, Belgium, and Austria). A fourth cluster is marked by polarization of people belonging to one denomination that resulted in a return to ‘integralistic’ authoritarian society, often sustained by the army (Spain, Greece). Finally, ‘authoritarian societies of the left’ (e.g., Russia) are the fifth category. According to Martin, there are, however, different forms of religious pluralism. ‘Complete pluralism’ results from competition among diverse denominations, which are not connected with authorities or a specific cultural elite. From such a situation, a notion as ‘civic religion’ can ap­ pear (e.g., USA). This ‘complete pluralism’ has to be distinguished from ‘qualified pluralism’, which is characterized by competition between a ‘particular religious body’ that is connected with a specific cultural elite and social authority. This is typical, for example, in England. ‘Qualified pluralism’ does also appear in another form. In Scandinavia for example, there is a competition between the cores of the state churches them­ selves. Finally, there is what Martin calls ‘segmented pluralism’. This denotes the situation of rival groups of Protestants and Catholics living in territorially separated areas. In such a situation, as well as when a meaningful but a small minority of deprived Catholics, who organized themselves in a specific socio-political subculture, live together with a Protestant majority, tolerance will be higher, as will the political will­ ingness to compromise. Here Martin refers to the Netherlands and Swit­ zerland (Martin, 1978a: 56-57, 111–113). In his analysis of the country-clusters, Martin draws attention to the question as to how theological and organization-sociological dimensions of the analyzed denominations influence the attitudes towards tolerance, pluralism, and peoples’ strive to uniqueness. He also demonstrates the importance of geographical factors (e.g., center-periphery-relations, the geographical distance to the sea (such as Malta)) for explaining identity consciousness. Other dimensions of religious identity consciousness are, according to Martin, migration, industrialization and modernization.

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

377

What seems clear is that a conglomeration of factors affects the tense pluralism-monopoly-positions. They are dependent upon e.g., historical, geographical, political, ideological, and theological aspects. His cluster­ ing clarifies that some constellations facilitate religious pluralism and personal tolerance, while other constellations hinder these processes. Martin’s analysis of the European denominational landscape demon­ strates that the degree of competition between the denominations in a society is an important factor. The ultimate consequences of this are still heavily debated (Beckford, 2000: 491). Some argue that the more com­ petition there is, the more likely churches and denominations are to adapt their supply to the religious demands and consequently religious participation will remain high or will increase (Stark & Finke, 2000). But the opposite is defended as well: more competition is inversely re­ lated to the degree of religious regulation. Yet others have argued that the impact of religious pluralism is moderated by the degree of diversity in religious demands. The more religiously diverse a country is, the greater the probability that people have found a religious and moral voice to their liking (Hamberg & Pettersson, 1994, 2002). However, since Martin’s publication in 1978, the European political and religious landscape changed dramatically. The far-reaching political and ideological transformations in Central and Eastern Europe had a pro­ found impact on the position of churches and denominations in the coun­ tries involved. Furthermore, an increasing part of Europeans do not con­ sider themselves as belonging to a denomination anymore. For a variety of reasons people are no longer attracted by the churches or religion(s) and likely these people will not only display lower levels of (traditional) religious beliefs, it can also be expected that they as a group are more heterogeneous in their beliefs and moral convictions. Thus, in countries where a majority is unchurched, heterogeneity can be expected to be strongest. Inspired by the theory of Martin, it seems likely that in countries where the majority is churched, heterogeneity will depend upon which denominations are present in a society (Catholic, Protestant, Orthodox people) and the degree to which denominations are present (monopoly position or not). The traditional doctrine of the Catholic church is as­ sumed to impose a stronger collective identity upon its members, while Protestant culture is assumed to be more conducive to religious indi­ vidualism, as Protestants are personally responsible before God in reli­ gious matters, and the clergy has a lesser role as mediator between the believer and God (Jagodzinski & Dobbelaere, 1995a: 81). Therefore, re­ ligious and moral pluralism will be highest in Protestant countries, and to a lesser degree in countries characterized by a mix of denominations

378

CHAPTER THIRTEEN

and countries with a mixed population of (diverse) denominations and unchurched people. It remains to be seen if we can find empirical evi­ dence for such claims. Further, due to the developments in the recent past, we expect that countries with a dominant Orthodox tradition can also be situated in this second group: the former USSR-ideological in­ fluence is still moderating the significance of Orthodox churches. Final­ ly, countries with a strong Catholic majority or monopoly will be least religiously and morally pluralistic. As we have argued in the introduction, the increase of religious and moral pluralism is regarded as an important consequence of moderniza­ tion. Inglehart argues that religion provided people with a sense of secu­ rity. In modern advanced wealthy welfare states, existential security is more or less obvious and for most people assured. In that respect, relig­ ion is no longer necessary. However, it does not imply that modern peo­ ple living in advanced wealthy societies with highly developed welfare states have become disbelievers or unbelievers. On the contrary, he ar­ gues, some core aspects of traditional religion are on a decline, namely providing people with a sense of security (Inglehart, 1997: 80). Similar arguments can be forwarded with respect to moral convictions: it is up to the individual, not only what to believe (in), but also what to consider right or wrong. Thus, one may assume that as society advances or be­ comes more modern in general and wealthier in particular, religious be­ liefs and moral convictions will become less homogeneous and thus in­ creasingly pluralistic. We also argued that the modern world is increasingly a global village in the sense that knowledge and information from all over the world are available to all people on the globe. It means that peoples’ scope is no longer limited to their own neighborhood, region or nation, but has grown to be worldwide. People living in contemporary European society are not only believed to be more autonomous and free to decide for themselves, they are also assumed to experience a wide variety of influ­ ences from other parts of the world. Increasingly they are faced with an extending range of alternative worldviews and value systems, provided by the mass media, modern communication means and mobility of peo­ ple. If people are freer to choose from an enlarged pool of religious and moral options, the homogeneity in their religious and moral value sys­ tems will almost by statistical necessity decrease. Thus, according to this theory, it seems likely to expect that the more globalized a country, in the sense of providing people possibilities of being informed about worldwide developments thanks to communication means, the more plu­ ralistic religiously and morally it will be.

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

379

Although the causal links are heavily debated, let us assume that in traditional societies moral beliefs were strongly under the influence of religion. As argued before, the religious system has been reduced to a subsystem of society alongside other subsystems and its overarching claims are of shrinking relevance in modern society (Dobbelaere, 1995: 1). Because fewer and fewer people adhere to the moral teachings of the churches, uncertainty about morals has grown and people’s moral sources have become quite diverse (Taylor, 1989: 401). In highly differ­ entiated societies, even to the core constituencies of the churches, the re­ ligious commitment might have become less relevant to various ‘secu­ lar’ life spheres as e.g., sexual relations and bio-ethical issues. Therefore, it seems likely that the more secularized a society is, the more pluralistic morally the population will be. 3 Measurements and analytical strategy

3.1 The dependent variables Religious and moral pluralism will be investigated by the degree of het­ erogeneity within each of these orientations. This kind of differentiation is referred to as internal pluralism and is measured by the size of the variance/standard deviation of religious and moral beliefs (see also Jagodzinski & Dobbelaere, 1995b). The more people score differently on a given dimension, the greater the variance and the higher the degree of internal differentiation1. Religious pluralism In the European Values Study scores of indicators are available, tapping different dimensions of religiosity. Analyses revealed that these dimen­ sions are strongly interrelated. Therefore we confine our analyses2 in this chapter to one general dimension of religiosity based on a combination of most of the ‘religious’ indicators: a) personal religiosity (importance of religion, belief in a personal God, being a religious person, impor­ tance of God, religion provides comfort and strength), b) traditional be­ liefs (belief in God, life after death, hell, heaven, sin), c) rites of passage (importance of religious service at birth, marriage and death), and confi­ dence in the church. Factor scores were calculated and the countries’ ______________ 1

There are of course other measures of heterogeneity or homogeneity available. We have also calculated inter-quartile distances, and correlated these with the measures of homogeneity based on standard deviations. The correlations were, however, .90 or higher. 2 On the merged data-file. Countries weighted equally.

380

CHAPTER THIRTEEN

standard deviations of these factor scores will be used to indicate the de­ gree of religious pluralism. Moral pluralism The EVS survey included a range of 22, quite different, behaviors. The respondents were asked to indicate on a scale from 1 to 10 if these be­ haviors could be never justified (1) or always justified (10). A factor analysis on these items yielded three dimensions3. One dimension can be interpreted as concerned with the debate about self-determination in relation to so-called ‘micro-ethical’ questions, or questions strongly associated with the private, personal life sphere, in­ cluding corporeality. Items included are justification of abortion, di­ vorce, homosexuality, euthanasia and suicide. A second dimension combines behaviors like cheating on tax, paying cash to avoid taxes, claiming state benefits illegally, lying in one’s own interest. Common theme of the second factor seems to be that the indi­ vidual gains personal benefits from such behaviors. We labeled this di­ mension personal interest morality. A third dimension includes behaviors such as driving under the influ­ ence of alcohol, throwing away litter, joyriding, speeding over the limit. These items refer to illegal activities, which do not provide immediate personal benefits when one breaks through the rules. As respect for laws and rules is involved, we have called this dimension civic sense-versuspermissiveness. For each of these dimensions scores were calculated from the factor analyses on the respective items and the countries’ standard deviations will be used as indicators of the degrees of moral pluralism. 3.2 Independent variables Composition variables The following composition variables will be included in analyses: age, level of education, degree of urbanization, and gender. Age is measured by year of birth and recoded into age in years. The level of education is measured by age when education was completed. The degree of urbani­ zation is indicated by 8 categories of size of the town: 1) under 2.000 in­ ______________ 3 Some items appear multi-dimensional. As we wanted clear one-dimensional items, the following items were excluded: taking the drug marijuana or hashish, married men/women having an affair, someone accepting a bribe in the course of their duties, ha­ ving casual sex, smoking in public buildings. Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization.

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

381

habitants, 2) 2-5.000, 3) 5-10.000, 4) 10-20.000, 5) 20-50.000, 6) 50100.000, 7) 100-500.000, 8) 500.000 and more. For these variables, country means were calculated and for gender we have included the per­ centages of men in a country. Religious profile We argued before that the analysis of Martin does not do justice to the current European situation, as it has developed in recent years. In par­ ticular, the growing group of unchurched people in Europe, and the Cen­ tral and Eastern European countries that were under USSR-influence, were underexposed in these analyses. Therefore we consulted the ‘World Christian Encyclopedia’ providing detailed information about denomina­ tions and religious adherence for all countries worldwide (Barrett et al., 2001). Not only church-membership is registered, also non-affiliation or ‘no longer-affiliation’ is indicated, as well as the number of agnostics and non-believers. This classification was more or less repeated by us, using the EVS data (see Figure 13.1).

100%

80%

60%

40%

20%

0% d k f i s e is

a t i e i t h r lt m t p l e s p t b e f r l u s i s k g r r o

% C a t h o li c

% P ro te sta n t

b g c z h u ru b yu a n l lv w g n i

% O rth o d o x

u k ee eg

% u n c h u rc h e d

Figure 13.1 Percentages of Catholics, Protestants, Orthodox, and unchurched in European countries

382

CHAPTER THIRTEEN

Our classification distinguishes 11 categories4: Protestant countries (more than 70% of the population is Protestant): Denmark, Finland, Sweden, Iceland; Catholic countries (more than 70% is Catholic): Aus­ tria, Croatia, Ireland, Italy, Lithuania, Malta, Poland, Portugal, Spain; Orthodox countries (more than 70% is Orthodox): Greece, Romania; Mixed Protestant/Catholics (Protestants and Catholics are almost equally sized): West Germany, Northern Ireland; Mixed Catholics/Unchurched (large parts of the people are Catholic, but a significant part of the people is unchurched): Belgium, France, Luxembourg, Slo­ venia, Slovakia; Mixed Anglican/Unchurched (large part is Anglican and large part is unchurched): UK; Mixed Orthodox/Unchurched (most peo­ ple are Orthodox and a significant part is unchurched): Bulgaria; Mixed Unchurched/Catholics (many people are unchurched and many people are Catholic): the Czech Republic, Hungaria; Mixed Unchurched/Protestant (most people are unchurched, a large minority is Protes­ tant): Estonia; Mixed Unchurched/Orthodox (many people are unchurched; but a large part is Orthodox): Russia, Belarus, Ukraine; Mixed Unchurched/Mixed (most people are unchurched, Catholics and Protestants are almost equally sized): the Netherlands, Latvia ‘Objective’ religious pluralism: the Herfindahl index Apart from the typology based on the kind and number of denominations in a country, the religious profiles can be obtained from what is known as the Herfindahl concentration index. This measure is based on the combination of 1) the number of churches/denominations, and 2) the ad­ herence to these churches. The more churches in a country and the more evenly distributed their market shares, the more religiously diversified the country is. The Herfindahl index is calculated by the formula si2 where si is the proportion with value i. The maximum value of the index is 1.0, and can be easily transformed into a pluralism index by subtract­ ing the Herfindahl index from 1. Information on the number of people affiliated with any religious organization in a country is simply taken from the questions on membership in the EVS surveys. The larger the number of churches and the more even the distribution among them, the lower the Herfindahl index, and the more pluralistic (less concentrated) the religious market.

______________ 4

In the figures presenting the results of our data-analyses, countries will be grouped based on this classification.

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

383

Economic situation: GPD A country’s welfare is measured by GDP per capita, which is commonly regarded as a good indicator (Van Snippenburg, 1986). The data comes from the Human Development Report 2000 and refers to the year 1999. Globalization Globalization refers to the increased exposure to modern communication means and flows of information. The spread of such modern communi­ cation technology devices can be used as an indicator. Both the World Development Report and Human Development Report provide compa­ rable measures for the national numbers of newspapers, telephones and mobile phones, and television- and radio receivers but also for the spread of IT-market issues such as number of pc’s, internet use and internet hosts. Unfortunately not all indicators are available for all countries and thus we have confined our measure to the number of televisions, tele­ phones, mobile phones, and Internet hosts in a country. The results from a confirmatory factor analysis of the various structural globalization in­ dicators justify that the statistics for TV-sets, telephones, mobile phones, and Internet-hosts can be combined into one scale of IT-market. The spread of communication means refers to structural globalization that can be distinguished from a cultural globalization that refers to the experiences of a global identity and cosmopolitanism. A person having a global identity is not only more likely to be more open to the rest of the world, but is also more often confronted with an extended range of alter­ native behaviors, habits, and attitudes. In the EVS-questionnaire this was measured by two questions revealing the geographical groups one sub­ jectively belongs to in the first and second place, respectively. Degree of secularization As we have argued, morality and religiosity are often regarded closely intertwined and the suggestion has been made that in the more secular societies moral orientations are more divers than in less secular, that is, more religious contexts. Thus, in case of the three moral orientations we also included the degree of secularization. The degree of secularization of a society is measured by a combination of religious beliefs (see our measure of religious pluralism) and practices (church attendance). We ran a factor analysis including all indicators of religious beliefs and con­ victions plus the variable on church attendance. Factor scores were cal­ culated and by reversing the scale the degree of secularization has been calculated for each country.

384

CHAPTER THIRTEEN

3.3 Analytical strategy The dependent variables are standard deviations and thus our analyses are aggregate level analyses. The number of countries is 33. In order to investigate the impact of the compositional variables we have regressed the 4 pluralism measures on the country’s mean age, percentage of men, the country’s mean scores on level of education and degree of urbaniza­ tion. In order to control for these effects, these variables have been in­ cluded in all subsequent analyses. In a next step we entered the religious profiles of the countries in the regression analysis. In a third step, ‘objec­ tive’ pluralism (the Herfindahl index) was entered in the model instead of religious profiles. In a next step we entered, instead of degree of plu­ ralism, the indicator of economic development or welfare in the regres­ sion analysis, and after that structural and cultural globalization were en­ tered. Finally, the degree of secularization was entered in a last step to explore the impact of this phenomenon on moral pluralism. At each step the R2 change indicates if the fit of the model was substantially improved or not compared with the (basic) model that included composition vari­ ables. The idea that pluralism has increased is explored first of all by com­ paring age categories in 1999. We have distinguished four categories or generations: a pre-war generation (those who are born between 1901 and 1940) and three post-war generations (those born between 1941 and 1955, a category born between 1956 and 1970, and those who are born after 1970). Each category is about one fourth of the population. For each generation the standard deviations on religious and moral orienta­ tions have been calculated and compared using Levene’s test of equal variances. Further, the changes over time are tentatively explored by comparing the degrees of religious and moral pluralism in 1981, 1990 and 1999 in these countries for which we have data available. For each year, coun­ tries’ standard deviations have been calculated on religious and moral orientations. Again Levene’s test of equal variances gives a formal test of the significance of the differences between years. 4 Results

4.1 Europe’s pattern of religious and moral pluralism Figure 13.2 displays an overview of Europe’s diversity with respect to religious and moral pluralism. Europe is far from a unity when it comes to the degree of religious and moral pluralism and it is not easy to find a

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

385

clear pattern. There does not appear an East-West or a North-South di­ vide and the clustering of countries is different for the four kinds of plu­ ralism we controlled, namely religious pluralism and the three distinct dimensions of morality. A religiously highly pluralistic country not nec­ essarily belongs to the most morally pluralistic societies (e.g., Hungary, Bulgaria, and the Netherlands) and vice versa. Portugal seems to be a country with a rather low degree of religious pluralism on the one hand and diverse acceptance of pluralism for the three dimensions of morality on the other hand. 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0 dk fi se is

at ie it hr lt mt pl es pt

rel plural

be fr lu si sk

pers interest

gr ro

bg cz hu ru by ua nl lv

self determination

wg ni

uk ee eg

civic permissiveness

Figure 13.2 Religious and moral pluralism in Europe (standard deviations)

There is no clear pattern that offers immediate convincing interpreta­ tions. Is pluralism highest in Protestant dominant countries? Based on our research results, the answer would be: no. The Protestant dominant countries (Denmark, Finland, Sweden, and Iceland) appear, in general, to be less pluralistic on all four dimensions compared to Catholic and mixed Catholic/unchurched countries (Belgium, France, Luxembourg, Slovenia, Slovakia). Further, there are sometimes obvious differences between countries that, as suggested by Martin (1978a) and Barrett (2001), belong to a same cluster. The neighbor countries Spain and Por­ tugal are explicit examples in this respect. Denmark and Iceland have a different profile compared to Sweden and Finland. Also countries that were under former USSR-influence do not yield a similar profile with

386

CHAPTER THIRTEEN

regard to religious and moral pluralism. Some similarities, e.g., between Bulgaria and the United Kingdom, are difficult to explain. But in spite of these remarks, the clustering of countries based on the theory of Martin and refined with Barrett-information seems to make sense, certainly as far as religious pluralism is concerned, and to a lesser degree as far as the dimensions of morality are concerned. 4.2 Understanding religious and moral pluralism in Europe We investigated the influence of composition effects. It appears that they hardly have any significant effects, neither on religious pluralism, nor on moral pluralism. Thus, it seems that differences in age structure, levels of education, degrees of urbanization, and differences in gender distribu­ tions, are not relevant issues that help explain country differences in re­ ligious and moral pluralism (see Table 13.1). Table 13.1 Results regression analysis religious pluralism Model 1 Model 21 Religiosity –.794 *** –.435 Age –.053 –.140 Education –.024 –.051 .070 .343 Urbanization % Men –.114 –.072 PROT –.025 ORTH –.176 .161 MCP .446 * MCU .076 MPUC .228 MOU .172 MUO .200 MUC .116 MUP .097 MUM Pluralism GDP It-market World .647 .828 R2 .581 .677 Adj R2 2 .138 Significance R change Note: 1. Catholic countries are reference group. * p < .05, ** p < .01, *** p < .001.

Model 3 –.597 *** –.195 –.114 .130 –.048

Model 4 –.799 *** –.060 –.030 .075 –.104

Model 5 –.800 *** –.025 –.061 –.014 –.077

.426 *** –.019

.768 .715 .001

.647 .566 .888

–.023 –.244 .696 .610 .156

Controlling for such social-demographic background characteristics, re­ ligious profiles were entered in the analysis of religious pluralism. Catholic countries were the reference category. It appeared that only countries with a mixed Catholic/unchurched profile are more religiously

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

387

pluralistic than these Catholic societies. The fact that denominational mixed societies are not significantly more pluralistic than Catholic socie­ ties seems a rejection of the idea that competition is necessary to have higher levels of pluralism. Also the idea that Protestant societies will be more pluralistic cannot be confirmed. The degree of objective pluralism measured by the Herfindahl index does not add much when entered in the analysis after the religious pro­ files. But when entered separately it is almost as powerful an explanation as religious profiles. Both the degree of urbanization and the degree of objective religious pluralism appear conducive to religious diversity. As such the suggestion that objective pluralism, in the sense of the number of denominations in a society and adherence to them, correlates with more diversity in the personal belief systems of individuals, cannot be rejected. The other structural characteristics have no effect on religious pluralism. Thus, neither welfare nor structural and cultural globalization contribute much to the understanding of the differences and similarities in Europe with regard to religious pluralism. Table 13.2 Results regression analysis pluralism with regard to personal interest morality Model 1 Model 21 Personal interest morality –.786 *** –.727 *** .087 Age .052 .158 Education .008 Urbanization –.232 –.264 % Men –.287 * –.265 PROT –.180 ORTH –.060 .050 MCP MCU –.055 .152 MPUC .031 MOU .053 MUO .021 MUC .169 MUP MUM –.151 Pluralism GDP It-market World Secularization .842 R2 .736 .703 Adj R2 .688 .391 Significance R2 change Note: 1. Catholic countries are reference group. * p < .05, ** p < .01, *** p < .001.

Model 3 –.698 *** –.058 –.051 –.211 –.290 **

Model 4 –.791 *** .034 –.013 –.212 –.252

Model 5 –.770 *** .052 .004 –.232 –.273 *

Model 6 –.762 *** .023 –.017 –.276 * –.316 *

.235 * –.062 –.032 .035 .776 .724 .043

.739 .679 .595

.738 .665 .912

.135 .749 .691 .257

388

CHAPTER THIRTEEN

The religious profiles do not add much for the explanation of the diver­ sity in personal interest morality (see Table 13.2). Objective pluralism as measured by the Herfindahl index does have a significant effect, how­ ever: the more religious pluralistic, the more heterogeneous with regard to personal interest morality. Countries where the percentage of men is lower are also more heterogeneous with regard to personal interest mo­ rality (e.g., cheating on tax, lying in one’s own interest). The analyses further reveal that the diversity of self-determination morality (questions associated with private, personal life sphere) cannot be explained strongly by any of the suggested phenomena. Dominant Protestant societies are, however, less pluralistic than Catholic societies (as are the other clusters of countries with other religious profiles). This seems contrary to our expectation that pluralism is highest in Protestant societies (see Table 13.3). Table 13.3 Results regression analysis pluralism with regard to self-determination morality Model 1 .807 *** –.147 –.429 * –.088 –.133

Model 21 .999 *** –.143 –.169 –.066 –.061 –.575 * –.004 .030 –.112 –.025 –.011 –.259 –.152 –.035 –.154

Self-determination morality Age Education Urbanization % Men PROT ORTH MCP MCU MPUC MOU MUO MUC MUP MUM Pluralism GDP It-market World Secularization .680 R2 .514 .398 Adj R2 .424 .564 Significance R2 change Note: 1. Catholic countries are reference group. * p < .05, ** p < .01, *** p < .001.

Model 3 .780 *** –.202 –.462 * –.076 –.111

Model 4 .818 *** –.150 –.436 –.087 * –.127

Model 5 .943 *** –.146 –.430 * –.030 –.108

Model 6 .829 *** –.135 –.425 * –.089 –.136

.113 –.018 –.236 –.136 .524 .414 .463

.514 .402 .922

.550 .424 .378

–.031 .514 .402 .883

Finally, diversity in civic permissiveness (e.g., driving under influence of alcohol) can be only to a very limited extent explained from differences

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

389

in religious profiles. Bulgaria, which represents the mixed Orthodoxunchurched profile, appears to be more pluralistic with regard to this kind of morality than dominant Catholic societies (see Table 13.4). Table 13.4 Results regression analysis pluralism with regard to civic permissiveness Model 1 .942 *** .144 –.054 –.157 –.177

Model 21 .902 *** .060 –.059 –.126 –.152 –.059 .074 –.082 .073 .063 .205 * –.017 –.054 .048 –.024

Civic permissiveness Age Education Urbanization % Men PROT ORTH MCP MCU MPUC MOU MUO MUC MUP MUM Pluralism GDP It-market World Secularization .843 R2 .914 .814 Adj R2 .837 2 Significance R change .267 Note: 1. Catholic countries are reference group. * p < .05, ** p < .01, *** p < .001.

Model 3 .962 *** .174 –.037 –.159 –.184

Model 4 .942 *** .139 –.059 –.152 –.169

Model 5 .927 *** .152 –.078 –.195 –.148

Model 6 .924 *** .121 –.072 –.184 –.189 *

–.049 –.016 –.029 .133 .845 .809 .604

.843 .807 .857

.859 .819 .266

.082 .848 .813 .379

In general, most of the suggestions with regard to moral pluralism can­ not be corroborated. This is clearly the case for the idea that the more economically advanced countries will also be the more pluralistic in their moral convictions. Also the statement that globalization has an im­ pact has to be rejected. More globalized countries are not more plural­ istic as far as religion and morality are concerned. Also statements about the impact of degree of secularization on moral pluralism are not clearly affirmative or negative for all three dimensions of morality. Referring to the ongoing debate about the relationship between reli­ gion and morality, we can at least state that there is not one single, dominant pattern to explain the influence of religion (or secularization) on moral opinions.

390

CHAPTER THIRTEEN

4.3 Generational differences Religious pluralism In most countries it is clear that younger generations are more heteroge­ neous in their religious beliefs than the older generation, but these dif­ ferences are not very big in a majority of the countries investigated here (19 out of 33). In West Germany, Lithuania, and Spain, the generational divide is between those born before 1955 and those born after 1956. In Slovakia, Poland, Ireland, Finland, and the Czech Republic the genera­ tional divide seems to be between pre-war and post-war generations. But this divide has, compared with the other countries, a different meaning for the Czech Republic: contrary to Poland, Finland, and Ireland, in the Czech Republic the oldest generation seems to be the one that demon­ strates most religious heterogeneity. In Malta, Greece, and Northern Ire­ land there is a more or less linear increase between generations towards more heterogeneity, while the patterns of East Germany, Hungary, and Portugal deviate from these patterns. In East Germany, the oldest and the youngest generations are almost as heterogeneous and as such different from the two other generational groups. In Hungary, those born between 1941 and 1955 appear as most heterogeneous, most homogenous is the pre-war generation with the generation born between 1956 and 1985 are in between these two extremes. Few generational differences concerning religious pluralism can be observed in e.g., Sweden, Austria, Belgium, Bulgaria, and UK (see Figure 13.3). We can conclude that younger generations are not always more het­ erogeneous as far as their religious opinions are concerned. Also, reli­ gious heterogeneity does not seem to be systematically spread over the various generations in all countries, and finally, countries that have been under the influence of USSR do not yield a similar pattern, indicating that the oldest generations are the most religious heterogeneous ones. Personal interest morality In general, pre-war generations are most homogeneous with regard to personal interest morality. However, in Belgium, France, Luxembourg, and the Netherlands, the differences between these generations are only modest, and in Slovenia, Portugal, Northern Ireland, Latvia, Iceland, and Spain, the divide is between those born before 1955 and those born after 1955. The generational differences between the pre-war and post-war generation for the dimension ‘personal interest morality’ are most im­ portant in the UK (see Figure 13.4).

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

391

1.2

1.0

0.8

0.6

0.4

0.2

.0 dk fi se is

at ie it hr lt mt pl es pt 1901-1940

be fr lu si sk

gr ro

1941-1955

bg

cz hu ruby ua nl lv

1956-1970

wg ni

uk ee eg

1971-1983

Figure 13.3 Degree of religious pluralism in four generations

1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0

dk fi se is

at ie it hr lt mtpl es pt

1901-1940

be fr lu si sk

1941-1955

gr ro

bg

cz hu ru by ua nl lv wg ni

1956-1970

uk ee eg

1971-1983

Figure 13.4 Degree of pluralism with regard to personal interest morality in four generations

392

CHAPTER THIRTEEN

In most countries, the intergenerational differences in homogeneity or heterogeneity with regard to self-determination morality are not very big. In Greece, Croatia, Hungary, Ireland, Lithuania, Malta, Northern Ireland, Poland, Romania, West Germany, and UK, the main distinction is between pre-war and post-war generations with the pre-war generation being most homogeneous. In Denmark, Sweden, and Iceland a reversed pattern is found (see Figure 13.5). Here the pre-war generation appears most pluralistic, and the youngest generation less! 1.2

1.0

0.8

0.6

0.4

0.2

0

dk fi se is

at ie it hr lt mtpl es pt

1901-1940

be fr lu si sk

1941-1955

gr ro

bg

cz hu ru by ua nl lv wg ni

1956-1970

uk ee eg

1971-1983

Figure 13.5 Degree of pluralism with regard to self-determination morality in four generations

Civic permissiveness In most countries, intergenerational differences in degree of pluralism with regard to civic permissiveness are clear, although the dividing line is not between pre-war and post-war generations in East Germany, Ice­ land, Ukraine, Italy, UK, Latvia, Northern Ireland, Romania, West Ger­ many, and Russia. In these countries the largest differences exist be­ tween those born before 1941 and those born after 1970. In Portugal, the generation of those born between 1955 and 1970 appears most heterogeneous, while in Denmark it is the group of people born between 1940 and 1955 (see figure 13.6).

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

393

2.0 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0

dk fi se is

at ie it hr lt mt pl es pt

1901-1940

be fr lu si sk

1941-1955

gr ro

bg

czhu ru by ua

1956-1970

nl lv

wg ni

uk ee eg

1971-1983

Figure 13.6 Degree of pluralism with regard to civic permissiveness in four generations

Although there is some evidence for a correlation between generations and the degree of pluralism, once again, we have to conclude that the patterns found do not indicate North-South, Protestant-Catholic, EastWest profiles. The differences between generations for all four dimen­ sions measured cannot be explained from one general theory. 4.4 Trends The suggestion that pluralism is increasing in Europe is explored by comparing the degrees of pluralism in the three years of surveys in 1981, 1990, and 1999. Two issues affect these analyses: 1) the number of countries in the previous waves was limited; and 2) some questions and items were not asked in the previous waves. Many of the Central and Eastern European countries participated for the first time in the project in 1999. The Baltic States, Poland, East Germany, Hungary, the Czech Re­ public, Slovakia, Hungary, Bulgaria, and Romania also participated in the 1990 wave, but not all data was comparable. Estonia, Lithuania, and Romania are therefore excluded from the analyses. Most of the indica­ tors of religious pluralism are available in the three waves, but due to a variety of reasons most of the items for self-determination morality are lacking. Therefore we have to confine our analyses of moral pluralism to

394

CHAPTER THIRTEEN

only two dimensions: personal self interest morality and civic permis­ siveness. In the following figures we have subtracted the standard deviation for 1981 from the one found in 1990, and the standard deviation of 1990 was subtracted from the standard deviation in 1999. Figures below zero indicate a decline, while figures above zero indicate increases. From 1981 to 1990 half of the countries show an increasing level of religious pluralism, while the other half of the countries had declining levels of religious pluralism. From 1990 to 1999 pluralism grew in Lat­ via, Northern Ireland, Bulgaria, and Ireland, while it declined in Portugal and Austria (see Figure 13.7). 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0

–0.05 –0.10 –0.15 dk fi se ic

at ir it pl es pt

be fr sk

1990 -1981

bg

cz hu

nl lv

wg ni

uk

eg

1999 -1990

Figure 13.7 Shifts in religious pluralism

Also with regard to personal interest morality there is no uniform change in the direction of more pluralism (see Figure 13.8). Denmark experienced a relative sharp decline from 1981 to 1990, while in most other countries the degree of pluralism hardly changed during these years. The differences between 1999 and 1990 appear bigger. In Hun­ gary, pluralism with regard to personal interest morality dropped, as was the case, although to a lesser extent in the Netherlands, Latvia, Sweden, and Finland. An increase in pluralism is found in most other countries. Most important increases can be found in Poland, Slovakia, and North­ ern Ireland and, to a little lesser degree, Austria, Ireland, Spain, UK, and former East Germany. From 1981 to 1990, people in Italy and Sweden became more hetero­ geneous with regard to civic permissiveness, while Italians became less

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

395

heterogeneous in the 1990-1999 period. From 1990 to 1999 increases in degrees of pluralism were found in East Germany, Lithuania, and North­ ern Ireland, while it declined in Hungary (see Figure 13.9). 0.3 0.2 0.1 0 –0.1 –0.2 –0.3 dk fi se is

at ir it lt pl es pt

be fr sk

1990-1981

bg

cz hu

nl lv

wg ni uk

ee eg

1999-1990

Figure 13.8 Shifts in pluralism with regard to personal interest morality

0.6 0.4 0.2 0 –0.2 –0.4 –0.6 –0.8 dk fi se ic

at ir it lt pl es pt

be fr sk

1990-1981

bg

cz hu

nl lv

wg ni

uk

ee eg

1999-1990

Figure 13.9 Shifts in degree of pluralism with regard to civic permissiveness

Taken into account the restrictions that affect this kind of analysis, we have to conclude that there is no uniform change in the direction of in­

396

CHAPTER THIRTEEN

creasing religious and more moral pluralism. Countries differ from each other, and developments in one country can even change in diverse di­ rections through the years. 5 Conclusions In this chapter we focused on the intriguing question if contemporary European society is indeed religiously and morally pluralistic. It is often argued that as a result of modernization pluralism has grown, but em­ pirical evidence is still lacking (Jagodzinski & Dobbelaere, 1996). We argued that the degrees of religious and moral pluralism likely vary across Europe. Apart from country differences in economic develop­ ment, globalization, and, as far as moral pluralism is concerned, degrees of secularization, differences in age distributions, gender distribution, levels of education and the degree of urbanization may be important at­ tributes to explain these differences and similarities. Further, we ex­ plored whether generations differ in degree of pluralism and we tenta­ tively tested the idea that pluralism has increased. Most sociological theories explain the gradual decline of institutional religion in (Western) Europe by the processes of social differentiation, specialization and individualization. As a result of these mutual influenc­ ing processes, theory suggests that religious beliefs and moral convic­ tions have become largely a matter of personal preference and options. The personal judgment concerning morality and religion may lead to the acceptance of and adherence to traditional religions, beliefs and prac­ tices, as well as to experimenting with new forms or rejecting all notions of religion. The same is assumed to apply to moral convictions and be­ liefs. Inglehart and Baker (2000) have argued that not only economic development is an important explanatory factor, but also a country’s cul­ tural heritage. Martin (1978a) drew similar conclusions in his work on secularization. A number of factors, ranging from historical, geographi­ cal, political, ideological to theological aspects, shape, according to Mar­ tin, the tensions that exist between religious pluralism and religious monopoly-positions. Based on such cultural dimensions he provided a clus­ tering of countries that enables some predictions on the degrees of reli­ gious and moral pluralism. In general, we do not find much evidence for such ideas. Reality ap­ pears more diverse and nuanced. As far as religious and moral pluralism is concerned, there is no clear pattern in Europe that offers convincing interpretations, e.g., referring to an East-West or North-South divide. Religiously highly pluralistic countries can be rather homogeneous as far as moral orientations are concerned and vice versa. Also it cannot be

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

397

concluded that protestant dominant countries are more pluralistic than other countries, or that country differences in age, level of education or gender distributions are relevant influencing factors. A similar picture with nuances arises with respect to moral pluralism. There is no straight­ forward answer to the question whether the degree of secularization af­ fects the degree of moral pluralism. Also the simple presupposition that the younger generations will be more heterogeneous as far as their reli­ gious and moral opinions are concerned compared with older genera­ tions is not supported, nor do changes over time provide much evidence that support the idea of increasing levels of pluralism. The clustering of countries based on the theory of Martin and refined with Barrett information and EVS data proves to be not sufficient to ex­ plain and understand the varieties in religious and moral pluralism in Europe. As such, other aspects of the cultural heritages have to be taken into account. An interdisciplinary approach, which pays attention to e.g., historical, cultural, and anthropological findings, may provide more ac­ ceptable and more convincing theories explaining the ‘patch work’ pat­ tern for religious and moral pluralism we observed in contemporary Europe in this chapter. Our findings show that generalizations cannot easily be made. We are aware of the fact that our analyses are just a first attempt to empirically picture religious and moral pluralism in contemporary Europe. This study needs to be repeated and refined. As far as theory is concerned, one has to reconsider the ideas about the impact of religious pluralism, as formulated in the well-known ‘religious market theory’ (Stark & Finke, 2000). The usefulness of this theory for the European context can be seriously doubted. It seems that the European religious market is not sufficiently pluralist to clarify the importance of religious diversity. Our analyses demonstrate that these theories cannot be applied in a European context. Also the theory that religious pluralism will be highest in Protestant dominant countries needs more reflection. Our re­ search results do not support this theory. Perhaps former Catholic domi­ nated countries have experienced an acceleration of the secularization process during the last decades. This acceleration may be understood as a consequence of ‘the strength of the past’ that has become ‘today’s weakness’, namely a strong hierarchy, including social control over people’s lives. It is possible that in a period of strong social control, people’s personal spiritual commitment to faith was weaker compared to protestant dominant countries, which accentuated a strong individual spiritual life. Further, the general category of ‘unchurched’ people that has recently grown in most countries deserves more attention. It might be clear that

398

CHAPTER THIRTEEN

‘unchurched’ is not a synonym for ‘not-religious’. In many countries, the number of non-religious people and the number of atheists is rather lim­ ited. Barrett (2001) registered the highest numbers of atheists in Estonia (10.9%) and Sweden (11.9%), while the highest percentages of ‘non­ religious’ people were found in Belarus (24.0%), the Czech Republic (31.9%), Estonia (25.1%), Latvia (26.0%), and Russia (27.5%). Barrett also predicts the future: he expects in these countries the number of non­ religious people to decline and the number of church-affiliated or church-members to increase. The unexpected rapid changes in former USSR will also have an impact on the secularization picture of Europe. Another point of attention for future research is to be found in what Bar­ rett indicates as ‘trans-megabloc groupings, such as Evangelicals, Pente­ costals, Charismatics.’ His figures reveal that in some countries the power of attraction of these ‘megabloc groupings’ is quite important, es­ pecially in the Nordic countries (Denmark: 12.9%, Finland: 27.2%, Sweden: 17.0%, Iceland: 10.5%), but also UK and Northern Ireland (29.6%), Ireland (16.4%), Malta (25%). Given the growing influence of these kinds of ‘megabloc groupings’ (see e.g., Van Harskamp, 2001), the future development of these groupings in Europe deserves more atten­ tion. And last but not least, our analyses seem to suggest that the current debate on moral developments is too narrowly focused on the idea of moral decline and growing moral diversity. It is often assumed that the great diversity in contemporary, individualized society generates con­ flicts instead of a consensus on values, moral behaviors and convictions (MacIntyre, 1981). Regularly, people warn for the ultimate consequen­ ces of so-called radical individualism and ethical relativism, caused by the loss of strong orientation schemes and an erosion of moral cons­ ciousness, social commitment, collective solidarity, and citizenship (Avi­ neri & De-Shalit, 1992). The only way to solve the problem of indivi­ dualistic, modern society is, according to proponents of the communita­ rian theories, the re-establishment of a firm moral order in society by (re-) creating a strong ‘we’ feeling and the (re-) establishment of a ‘spirit of community’ (Etzioni, 1993). The results of our analyses do not pro­ vide strong arguments in favor of such ideas about a moral decline and the necessity to invest more energy in communities built upon a strong ‘we-feeling’, as a dam against (perceived hyper-) individualization. Fu­ ture research can clarify to which degree this discourse is driven by a sense of reality or by fear for societal dynamics. The empirical findings presented in this chapter are not very suppor­ tive for most of the generally accepted theories on the developments in and current state of religion and morality in contemporary Europe. There

RELIGIOUS AND MORAL PLURALISM IN CONTEMPORARY EUROPE

399

is no convincing evidence of increasing moral and religious pluralism as is a popular view in modern discourses. As such, the grand narratives and theories need revision or at least reconsideration. There is still much to do in order to really understand what is happening with regard to re­ ligion and morality in European societies. References Avineri, S. & A. De-Shalit 1992. Introduction. Pp. 1–11 in S. Avineri & A. De-Shalit (eds.), Communitarianism and Individualism. Oxford: Oxford University Press. Barrett, D.B., G.T. Kurian & T.M. Johnson 2001. World Christian Encyclopedia. A Compara­ tive Survey of Churches and Religions in the Modern World. Vol. I. Oxford: Oxford Uni­ versity Press. Beck, U. 1992. Risk Society. Towards a New Modernity. London: Sage Publications. Beckford, J.A. 2000. ’Start Together and Finish Together’: Shifts in the Premises and Para­ digms Underlying the Scientific Study of Religion. Journal for the Scientific Study of Re­ ligion 39: 481–496. Davie, G. 2002. Europe: The Exceptional Case. London: Darton, Longman & Todd . Dobbelaere, K. 1993. Church Involvement and Secularization: Making Sense of the European Case. Pp. 19–36 in E. Barker, J. Beckford & K. Dobbelaere (eds.), Secularization, Ra­ tionalism and Sectarianism. Oxford: Clarendon Press. —— 1995. Religion in Europe and North America. Pp. 1–29 in R. de Moor (ed.), Values in Western Societies. Tilburg: Tilburg University Press. Ester, P., L. Halman & R. de Moor (eds.) 1994. The Individualizing Society. Value Change in Europe and North America. Tilburg: Tilburg University Press. Etzioni, A. (1993). The Spirit of Community. Rights, Responsibilities, and the Communitarian Agenda. New York: Crown. Halman, L. & R. de Moor 1994. Religion, Churches and Moral Values. Pp. 37–66 in P. Ester, L. Halman & R. de Moor (eds.), The Individualizing Society. Tilburg: Tilburg University Press. Hamberg, E. & T. Pettersson 1994. The Religious Market: Denominational Competition and Religious Participation in Contemporary Sweden. Journal for the Scientific Study of Re­ ligion 33: 205–216. —— 2002. Religious Markets: Supply, Demand, and Rational Choices. Pp. 91–114 in T. Jelen (ed.), Sacred Markets, Sacred Canopies: Essays on Religious Markets and Religious Plu­ ralism. Lanham: Rowman-Litttlefeld. Inglehart, R. 1997. Modernization and Postmodernization. Princeton: Princeton University Press. —— & W.E. Baker 2000. Modernization, Cultural Change and the Persistance of Traditional Values. American Sociological Review 65: 19–51. Jagodzinski, W. & K. Dobbelaere 1995a. Secularization and Church Religiosity. Pp. 76–119 in J.W. van Deth & E. Scarbrough (eds.), The Impact of Values. Oxford: Oxford University Press. —— 1995b. Religious and Ethical Pluralism. Pp. 218–249 in J.W. van Deth & E. Scarbrough (eds.), The Impact of Values. Oxford: Oxford University Press. —— 1996. Religious and Moral Pluralism. Application to the European Science Foundation, Brussels. MacIntyre, A. 1967. Secularization and Moral Change. London-New York-Toronto: Oxford University Press. —— 1981. After Virtue. A Study in Moral Theory. London: Duckworth. Martin, D. 1978a. A General Theory of Secularization. Oxford: Basil Blackwell.

400

CHAPTER THIRTEEN

—— 1978b. The Religious Condition of Europe. Pp. 228–287 in S. Giner & M. Scotford Archer (eds.), Contemporary Europe. Social Structures and Cultural Patterns. London: Routledge & P. Kegan. —— 1996. Religion, Secularisation, and Post-Modernity: Lessons from the Latin American case. Pp. 35–44 in P. Repstad (ed.), Religion and Modernity. Modes of Co-existence. Oslo: Scandinavian University Press. Stark, R. 1997. Bringing Theory Back in. Pp. 3–24 in L. Young (ed.), Rational Choice Theories of Religion. London: Routledge. —— 1999. Secularization RIP. Sociology of Religion 60: 249–73. —— 2001. God, Rituals and the Moral Order. Journal for the Scientific Study of Religion 40: 619–636. —— & R. Finke 2000. Acts of Faith: Explaining the Human Side of Religion. Berkeley: Uni­ versity of California Press. —— & L. Iannacone 1994. A Supply-Side Reinterpretation of the ‘Secularization’ of Europe. Journal for the Scientific Study of Religion 33: 230–252. Taylor, C. 1989. Sources of the Self. Cambridge: Harvard University Press. Therborn, G. 1995. European Modernity and Beyond. London: Sage Van Harskamp, A. 2000. Het nieuw-religieuze verlangen. Kampen. Kok Van Snippenburg, L. 1986. Modernisering en sociaal beleid. Nijmegen. Wilson, B. 1982. Religion in Sociological Perspective. Oxford: Oxford University Press. —— 1998. The Secularization Theses: Criticisms and Rebuttals. Pp. 45–65 in R. Laermans, B. Wilson & J. Billiet (eds.), Secularization and Social Integration. Leuven: Leuven Univer­ sity Press.

PART THREE

EUROPEAN VALUES IN REFLECTION

This page intentionally left blank

CHAPTER FOURTEEN

UNITY PRODUCES DIVERSITY: THE ECONOMICS OF EUROPE’S SOCIAL CAPITAL LANS BOVENBERG 1 Introduction The European Values Study (EVS) documents substantial cultural diver­ sity within Europe. European economic integration is sometimes seen as the enemy of this rich cultural diversity. It is argued that the internal market can flourish only if European countries increasingly resemble each other in terms of cultural and institutional features. This chapter1 investigates whether cultural homogeneity is indeed a necessary condi­ tion for economic integration. In so doing, it explores how social capital (which includes shared beliefs, agreements and values as well as infor­ mation networks) facilitates international trade and is affected by Euro­ pean economic integration. This chapter is structured as follows. Section 2 discusses the role of social capital in economic analysis. In particular, sub-section 2.1 ex­ plores why economists have recently become more interested in social capital. Sub-section 2.2 investigates how economists view and have modeled social capital. It first outlines the traditional approach based on extrinsic incentives originating in norms and concerns about reputation. Subsequently, it surveys more recent approaches, which allow for in­ trinsic motivations and values. Section 3 applies the insights derived in section 2 to Europe: it analyses the cultural base for European economic integration (in section 3.1) and the feedback of intra-European com­ merce on norms and values (in section 3.2). This section focuses on the relationship between European integration based on common European values and the cultural diversity documented by EVS. Section 4 makes a number of methodological remarks on the role of survey questionnaires (such as EVS) in measuring social capital (in section 4.1) and in shed­ ding light on the relationship between social capital and economic deve­ lopments (in section 4.2). Finally, section 5 contains some conclusions. _____________ 1

The author thanks Johan Graafland for helpful comments on an earlier draft.

404

CHAPTER FOURTEEN

2 Economics and social capital

2.1 The relevance of social capital For several reasons, economists have become increasingly interested in social capital on both the aggregate, business, and individual levels. Macro-economic level Social capital includes the norms and values that facilitate coordination if individual and collective goals are not perfectly aligned because nei­ ther the market nor the government confronts individuals with the social costs and benefits of their actions2. In these situations, social capital may substitute for explicit contracts in addressing the resulting externalities, which can be defined as missing markets (or incomplete contracts). Ex­ ternalities imply that transaction costs block mutually advantageous trade. The resulting incomplete contracts give rise to various coordina­ tion problems including hold up as a result of imbalanced trade, moral hazard due to asymmetric information on actions, adverse selection on account of asymmetric information on the transacted commodity, and public goods resulting from lack of well-defined property rights. Ex­ plicit, legal contracting is increasingly costly in modern economies. Ex­ plicit contracts are too rigid to do justice to the complex, dynamic world in which we live. In particular, in a dynamic economy relying on innova­ tion, transaction partners often cannot write down ex ante what they ex­ actly expect from each other in each contingency. Hence, courts cannot verify whether transaction partners have actually kept their commit­ ments. In any case, courts and governments often lack both the expertise and information to do so. As a result, externalities are pervasive in a tur­ bulent, knowledge-intensive society based on increased specialization and innovation. Principal-agent relationships abound in which principals are vulnerable to opportunistic behavior of their agents. As a substitute for legal contracts and government intervention based on explicit regulations, social capital is increasingly seen as an instru­ ment to internalize externalities, thereby benefiting both the adaptability and the growth performance of economies. Indeed, in addition to more familiar factor endowments like human and physical capital, social capi­ tal is an important factor affecting the economy-wide production func­ tion. Rather than a shortage of physical capital, a lack of trust due to a _____________ 2

As coordination mechanisms, public laws, contracts, norms, and values can be dis­ tinguished. For a formalization in the context of the well-known prisoner’s dilemma, see Bovenberg (2002).

UNITY PRODUCES DIVERSITY

405

dearth of social capital may in fact be at the root of the development trap experienced by several less developed countries (LDCs). Lack of social capital inhibits specialization, decentralization and risk sharing based on financial markets and the separation of ownership and control in modern corporations. Business level At the business level, a certain amount of social capital yielding intrafirm trust is an important determinant of competitiveness. Indeed, in a knowledge economy, the value of a firm is determined less by its physi­ cal assets and more by its organizational architecture and identity (i.e. a business culture that allows the firm to innovate and respond flexibly to changing circumstances and that motivates professional employees to utilize their discretion to act in the interests of the firm as a whole) and reputation (built on open communication with stakeholders so that con­ sumers, employees, debt- and shareholders are willing to entrust them­ selves to the firm). In contrast to explicit contracts within the firm, these intangible assets are difficult to imitate by competitors and thus consti­ tute an important source of enduring competitive advantage (see Kay, 1993). Micro level Social capital can be defined also on a micro-economic level, namely as a person’s social skills, charisma, and access to social networks. Social capital can be viewed as part of an individual’s human capital: human capital includes not only cognitive abilities and physical strength but also social skills. In modern knowledge-intensive economies, one’s so­ cial capital is in fact an increasingly important determinant of one’s earning power. Indeed, in a service-oriented economy, work is more about relating to other people than to nature (as it was in an industrial society). 2.2 Social capital: Extrinsic or intrinsic motivation? Social capital can be modeled in various ways. In the standard economic approach, social capital provides extrinsic incentives to cooperate. More recently, economists have started to explore social capital as social pref­ erences yielding an intrinsic motivation to cooperate3. _____________ 3

For the distinction between intrinsic and extrinsic motivations in the context of so­ cially responsible businesses, see Graafland (2002).

406

CHAPTER FOURTEEN

The standard economic approach The standard approach in economics is to model trust on the basis of a repeated prisoner’s dilemma. In this setting, people can be trustworthy because it is in their own pecuniary interest to comply with an agree­ ment. In particular, although cheating provides a short-term benefit, it generates a long-term loss because potential trading partners respond to cheating by no longer engaging in future trade (or at terms that are less favorable). Several conditions must be met for a repeated prisoner’s dilemma with selfish players to result in a cooperative equilibrium. First of all, the transactions must be repeated infinitely many times or should be con­ cluded at an uncertain date. Second, the players must be sufficiently pa­ tient so that they attach a large enough weight to the future. Third, the players must believe that the strategies of their trading partners depend on their own decisions. In other words, the players are believed to play so-called trigger strategies. In particular, a cooperative equilibrium re­ tit-for-tat'strategies (i.e. if they punish non-coopesults if agents play ' rative behavior by playing non-cooperatively in the future and reward cooperative behavior by continuing to play cooperatively). Although punishment will not be observed in a cooperative equilibrium, the beliefs about this out-of-equilibrium behavior are crucial because these beliefs encourage players to behave cooperatively. Fourth, past behavior of agents should be known to the market as a whole so that the market can sanction deviant behavior. Indeed, the literature on economic history has documented how institutional innovations in information transmission, such as medieval merchant guilds, facilitated efficient trade (see, e.g., Greif, Milgrom & Weingast, 1994). Modern institutions that transmit in­ formation about the actions of others (such as credit bureaus, trading or­ ganizations, public auditors, consumer watch groups, craft associations and networks) and that coordinate responses by setting standards of be­ havior (such as trade unions, industrial associations, professional organi­ zations that accredit members) are helpful in this respect. As shared implicit agreements and common beliefs about others parti­ cipating in the punishment of non-compliant behavior, norms can ensure that reputational forces help select the cooperative equilibrium (see, e.g., Binmore, 1998). In this way, norms as beliefs about out-of-equilibrium behavior of other players act as a device to select the cooperative equi­ librium from the multiple equilibria in a repeated prisoner’s dilemma. In the traditional economist’s view, social capital thus consists of three elements. The first element is the belief that trading partners re­ ward cooperative behavior and punish non-cooperative behavior. Also part of social capital are mutual agreements about what is considered co­

UNITY PRODUCES DIVERSITY

407

operative behavior (and thus should be rewarded) and what is non-cooperative behavior (i.e. behavior that deserves to be punished). The networks that facilitate monitoring and the transmission of the ac­ quired information constitute the third component of social capital. By helping to enforce implicit agreements through pecuniary sanctions, so­ cial capital results in a calculating type of trust. This trust depends on particular institutions (such as an information network) rather than the inherent features of the individual personalities involved. Social capital as preferences Whereas economists traditionally model trust as a type of calculativeness, sociologists typically view trust as a belief in someone’s personal attributes (or constitution): agents comply with norms of fairness and morality even if they do not derive pecuniary benefits from doing so. Hence, they are intrinsically rather than extrinsically motivated. In other words, agents derive non-pecuniary benefits from acting in accordance with particular norms. The intrinsic values of people are thus part of so­ cial capital: these values can be viewed as individual preferences that value particular behavior. People earn self-respect and feel virtuous (i.e. enjoy non-pecuniary benefits) by behaving in accordance with their con­ science, whereas they may feel guilty and lose self-respect if they do not behave this way. Economists are rather hesitant to accept the richer view of human mo­ tives adopted by sociologists, even though the founder of modern eco­ nomics, Adam Smith (1759), described various forms of non-pecuniary preferences4. Allowing for more general, individual-specific preferences helps to explain individual actions but comes at the expense of the pre­ dictive power of economics in predicting aggregate responses to chan­ ging incentives (see Tirole, 2002). Indeed, explaining cooperation on the basis of preferences for cooperation risks a tautology: virtually any be­ havior can be explained by simply positing a taste for it. Economists pre­ fer a simple, parsimonious5 theoretical framework, which relies on a small number of simple foundations and has predictive power and thus can be falsified. _____________ 4

To illustrate, Smith (1759) argued that man’s desire for the approval of the ‘impartial spectator’ constitutes the moral basis for society. By restraining self interest, these non­ pecuniary preferences facilitate cooperation and the division of labor. Smith thus stressed the reflective nature of man, which could offset to some extent man’s emotional and in­ stinctive urges to engage in the pursuit of immediate pleasure. 5 This means that a minimum number of causes should be assumed to model beha­ vioral regularities.

408

CHAPTER FOURTEEN

Several economists have started to model simple preferences in order to provide a more realistic but still rigorous and unified account of hu­ man behavior in social situations and its consequences for aggregate economic behavior. These models yield predictions that can in principle be falsified. Individuals exhibit social preferences if they care not only 6 about their own pecuniary payoffs but also about the payoffs of others . Hence, individual utility depends also on non-pecuniary elements. In these models, trust is not purely calculative but is also based on the per­ sonal attributes of the person who is being trusted. Francois (2002), for example, argues that reputational considerations alone are not sufficient to yield the trust that is required to get entrepre­ neurial relationships off the ground in dynamic, rapidly changing envi­ ronments in which relationships tend to be unstable and short lived. Francois stresses trustworthiness as the economically relevant part of culture: trustworthy individuals refrain from cheating their trading part­ ners even though they may reap pecuniary benefits from doing so. These individuals thus enjoy a non-pecuniary benefit from keeping their com­ mitments rather than acting in a purely opportunistic manner. Fehr and Falk (2002) argue that a substantial fraction of people fea­ ture so-called reciprocal preferences: they exhibit stable preferences for fairness in the form of reciprocity. They claim that this parsimonious framework can explain data from many different experimental games. Reciprocity implies that people are willing to reward friendly actions and punish hostile actions. Hence, people engage in non-selfish behavior conditional on the actions of others. In fact, just as they have tastes for food and thus want to spend money on it, they have a taste for fairness and are willing to spend money on being reciprocal. Reciprocal prefer­ ences transform social situations with the pecuniary pay-off structure of the prisoner’s dilemma in a coordination game. This implies that even a one-shot game of this type may yield a cooperative equilibrium if agents are allowed to communicate. Indeed, in coordination games, creating the belief that other players act cooperatively becomes essential. Fehr and Fishbacher (2002) stress that agents’ preferences are hetero­ geneous: some agents may feature reciprocal preferences, while others are completely selfish. Indeed, even in carefully controlled situations, experiments show substantial heterogeneity in social behavior. Repeated interaction is important in the presence of heterogeneous individuals: it _____________ 6

In addition to the assumption of the self-interested nature of preferences, economists have started to relax other assumptions underlying the self-interested, rational homo economicus. In particular, various models of bounded rationality have been developed that incorporate imperfections in man’s cognitive abilities (such as imperfect selfknowledge and recall) and willpower (see Tirole, 2002).

UNITY PRODUCES DIVERSITY

409

allows players to learn more about the intrinsic integrity of their trading partners. At the same time, selfish agents may want to gain a credible reputation for being cooperative by imitating the behavior of agents fea­ turing reciprocal preferences. Fehr and Falk (2002) maintain that agents feature preferences for so­ cial approval, which may be closely related to preferences for a good self image. Such preferences were already discussed by Adam Smith. He ar­ gued that individuals want to be admired by both themselves and others. By putting themselves in the shoes of others, people constantly imagine how others must see them. An individual’s utility thus depends on the approval of others. The desire for social approval may result in confor­ mity effects if the strength of social approval associated with a particular behavior increases with the number of other people within a reference group engaging in similar behavior. In that case, multiple equilibria may occur (see also Durlauf, 1996). Rabin (2002) argues that people have a taste for fairness. He main­ tains also that new theories that depart from the traditional model of the homo economicus will ultimately yield not only more realistic but also simpler, more tractable theories of human behavior than existing models based on self-interested, rational actors. Endogenous social preferences Without explaining how social preferences are formed, economics would not have much predictive power about cooperation facilitating economic growth. It could only give a tautological account: growth on the basis of successful cooperation occurs in the presence of the proper social preferences. In order to be able to explain growth, some authors have started to look for theories that explain where social preferences come from. What determines, for example, social capital defined as in­ herent trustworthiness, and how do economic incentives affect the ac­ cumulation of that capital? Exploring endogenous preference formation provides an interdisciplinary research program for economists and sociologists. Most economic studies on endogenous preference formation employ evolutionary game theory. They explore whether players who play co­ operative strategies survive in a population with players who play non-cooperatively. Just as the literature on the reputation mechanism, this literature points to the crucial importance of information transmis­ sion (in this case, about the player’s type). Being cooperative pays off if the other players know whether one is cooperative or not. Only under these strong informational assumptions are cooperative strategies evolu­ tionarily stable.

410

CHAPTER FOURTEEN

Even though an individual’s type is not revealed to other traders, al­ truistic preferences are still evolutionarily stable in Bisan and Verdier (2001) and Francois and Zabojnik (2001). The reason is that the replica­ tor dynamic in these models depends not only on pecuniary benefits (i.e. ‘fitness’, as in traditional models), but also on the parents’ evaluation of the children’s life-time returns. Social capital in the form of social prefe­ rences is shaped by rational, forward-looking parents who inculcate pre­ ferences in their children on the basis of their own social preferences 7 (see also Becker, 1993) . Indeed, parents foresee that morally committed persons obtain emotional benefits from acting in accordance with their values, which more than offset the additional pecuniary benefits reaped by those living in a moral vacuum. The evolution of mankind is thus not completely determined by the anonymous forces of nature but is also af­ fected by the conscious decisions of parents, who exhibit the ability to reflect on themselves and their children. The distinction between the fitness and non-fitness replicators paral­ lels the distinction between extrinsic and intrinsic motivations of indi­ viduals. In particular, the fitness replicator typically employed in evolu­ tionary game theory cares about only pecuniary benefits, while a nonfitness replicator based on the parental social preferences includes also non-pecuniary payoffs. With a non-fitness replicator, therefore, cultural evolution is not determined by materialistic considerations alone. Also non-pecuniary factors, such as the cultural and spiritual heritage, may play a role. This richer model of cultural evolution makes cultural deve­ lopment less predictable, deterministic, and materialistic. Indeed, cul­ tural mutations may generate enduring effects as cultural change be­ comes path dependent. The predictions of a richer model of cultural evolution based on non­ pecuniary returns is consistent with the empirical evidence provided by Inglehart and Baker (2000)8. They find that, although economic develop­ ment results in some predictable changes in values, the broad cultural heritage of a region yields an enduring effect on the distinctive value systems. While values change, these values continue to reflect a region’s cultural heritage. Indeed, Inglehart and Baker (2000) find that religious

_____________ 7

Families play a crucial role in the intergenerational transfer of social capital. The bonding of parents and children in the family may thus constitute the basis for the accumu­ lation of not only bonding but also bridging social capital. Bonding in informal social networks may play a similar role. Bonding and bridging social capital can thus be comple­ mentary. 8 For a critical assessment of the methodology adopted by Inglehart and Baker (2000), see Van Schaik and Beugelsdijk (2000).

UNITY PRODUCES DIVERSITY

411

beliefs persist9. Also this finding suggests that cultural change is driven by more than material factors alone. Economics and cultural values in­ teract: economic interaction shapes cultural values, but is also shaped by these values (see also Inglehart, 1990). 3 European economics and values This section applies the main insight from the previous section—namely that economics and cultural values affect one another—to European in­ tegration. In particular, section 3.1 investigates the cultural base for eco­ nomic integration in Europe. Subsequently, section 3.2 considers the feedback of intra-European commerce on norms and values. 3.1 Cultural base for European economic integration Shared values European economic integration is based on a number of shared values. Indeed, an integrated European market can survive only if it is firmly based on a number of core values. To illustrate, the European monetary union, which undergirds the internal market by removing internal ex­ change rate volatility, requires fiscal and monetary discipline. The Maas­ tricht Treaty guaranteeing an independent status to the European Central Bank as well as the stability and growth pact constraining fiscal policies are explicit contracts enforcing the required financial discipline. The Northern European countries, in general, and Germany, in particular, in­ sisted on these contracts because they did not trust the Southern Euro­ pean countries’ commitment to this discipline (see Bovenberg & De Jong, 1996). Indeed, in contrast to post-war Germany, these latter coun­ tries cannot boast of a tradition of financial discipline. Ultimately, how­ ever, these contracts must be enforced by the values of the European citizens; commitment to these contracts can endure only if it is based on the internalized preferences of European citizens, and in particular on the values of patience and trustworthiness. Also respect for the rule of law, property rights and the rights of mi­ norities are important values supporting an efficient common European _____________ 9

In fact, Inglehart and Baker document that spiritual concerns become more wide­ spread in advanced industrial societies. Indeed, emotional needs associated with non­ pecuniary preferences may gain in importance compared to material needs as material growth increasingly satisfies these latter wants. Especially in affluent societies, man lives by more than bread alone. Hence, parents may attach great importance to teaching their children the language of moral reasoning so that these children can reason about and ar­ ticulate their moral values.

412

CHAPTER FOURTEEN

market associated with a level playing field. These values help alleviate hold-up problems that potentially hinder intra-European commerce (see, e.g., Bovenberg & Teulings, 1999). Again, various institutions, such as legal safeguards (including the separation of church and state as well as the separation between legislative, executive and judicial powers) protect these values but can endure only if they are based on the values of the European citizenry. Especially important for modern production based on international division of labor is general morality that extends the duty to respect the rights of others even if they do not belong to one’s own family, personal acquaintances, or ethnic group. In other words, this part of social capital should be ‘bridging’10. This general morality is the basis also for the modern corporation in which decentralized, specialized professionals are given ample room for personal discretion. Indeed, in­ dividual freedom, lack of coercion, pluralism and diversity go together with individual responsibility and restraint based on commitment to ba­ sic, common values. Also the reputation mechanism facilitates an efficient market in which consumers can trust producers to supply high quality products. As noted in sub-section 2.1, the proper functioning of the reputation mecha­ nism relies on shared beliefs and views within the market. Moreover, agents should feature low discount rates so that they attach substantial importance to the future. Diversity in values Whereas European integration requires convergence to a set of common values, it may actually benefit from diversity in values in other dimen­ sions. EVS documents the cultural diversity within Europe. This cultural pluralism does not need to be an obstacle to European integration. In­ stead, it increases the potential for intra-European trade and allows Europe to benefit from the specific strengths of each of the cultural tradi­ tions11. The internal market encourages each region to increasingly spe­ cialize in its comparative advantages, which are in part shaped by its specific social capital. In this way, European integration produces more heterogeneous production structures drawing on intra-European diversity in cultural heritages. Indeed, internationalization in general and Euro­ _____________ 10

In this connection, one can also distinguish between open and closed social capital, depending on how members of a group value the welfare of those outside the group. 11 Indeed, each region occupies different positions on fundamental trade-offs such as flexibility versus commitment, scale versus diversity, incentives versus risk sharing, and experimentation versus certainty. For an overview of these fundamental dilemmas that bedevil human interaction, see CPB (1997). The benefits from diversity combined with the need for some shared values yield a trade-off between widening and deepening the European Union.

UNITY PRODUCES DIVERSITY

413

pean integration in particular go together with an increased emphasis of regions on their own identity. Just as specific social capital producing difficult to imitate intra-firm trust is an important determinant of interna­ tional competitiveness of each specific firm, region-specific social capital constitutes an important source of comparative advantage for a region. The insight that diversity is produced by and raises the benefits from European integration strengthens the arguments in favor of subsidiarity. Europe should leave room for differentiation based on cultural differ­ ences. Whereas there should be convergence on the essential values (such as respect and tolerance for other values that are not intolerant of others), diversity and differentiation is called for on non-essential values. 3.2 European integration as driver of cultural change Shared values The founders of the European Union perceived economic integration as a way to foster a political union. They hoped that increased intra-Europe trade would eliminate violent European conflicts not only by increasing mutual dependency and by fostering transparency that would facilitate the reputational mechanism (i.e. fostering extrinsic incentives to coope­ rate), but also by enhancing mutual understanding and respect (i.e. fos­ tering intrinsic incentives to cooperate). In other words, European eco­ nomic integration was meant to result in the accumulation of open social capital. Economics was thus viewed as an instrument to reach political and cultural ends. In addition, various European institutions (such as the European Court of Justice) were instituted to help transform closed so­ cial capital, which serves to exclude other people, into open social capi­ tal (or ‘bridging’ social capital). The view that economics should drive cultural change is not uncon­ troversial. The debate on the proper sequencing of political and mone­ tary union illustrates this. German central bankers typically stressed that a monetary union cannot work properly without a political union founded on common values. Politicians, in contrast, viewed monetary union as a instrument to help create such a political union. Indeed, the speed with which monetary union should be implemented poses difficult dilemmas. On the one hand, economic adjustments typically occur faster than changes in cultural values; this argument supports the political view that economic changes should precede cultural changes. On the other hand, moving too fast on economic integration may prove counterpro­ ductive if this integration backfires because of lack of popular support. In that case, it may give rise to serious intra-European conflicts, thereby eroding instead of accumulating open social capital. Indeed, whereas

414

CHAPTER FOURTEEN

successful repeated interactions can strengthen the emotional and affec­ tive ties between parties, unsuccessful interaction tends to have the op­ posite effect12. In view of these dilemmas, a symbiotic relationship be­ tween economic and cultural changes requires that these changes move in tandem and thus occur only gradually. Another important value affected by European integration is solidar­ ity. Several observers contend that European integration reduces the scope for European governments to redistribute by increasing the mobil­ ity of production factors (Sinn, 1990). This may harm insurance pro­ vided by European welfare states. However, increased intra-European mobility may also help to gradually extend social preferences that value solidarity towards residents of other countries. This may eventually build popular support for more extensive redistribution on a European level, thereby protecting valuable insurance13. Diversity of values European integration encourages diversity in production structures (in­ volving not only the sector structure but also the internal organization of firms) through increased specialization (see sub-section 3.1). In this way, European integration may actually nurture rather than eliminate cultural differences between regions. Hence, economic and cultural developments in European regions do not converge to a unique, uniform equilibrium. Instead, these developments depend importantly on region-specific cultu­ ral heritages given by history and are thus path dependent. The bridging social capital and the uniform values undergirding the internal market on the European level may in this way help to strengthen dense regional clusters with bonding social capital on a more decentralized level. Brid­ ging may thus form the basis for bonding. The prediction that European integration may stimulate cultural di­ versity is consistent with the empirical evidence in Inglehart and Baker (2000) described in sub-section 3.1: the broad cultural heritage of a re­ gion exerts an enduring effect on the distinctive value systems14. In the same way, EVS reveals persistent cultural differences instead of a univer­ _____________ 12

Policymakers face similar dilemmas in determining the speed with which the Euro­ pean Union should welcome new entrants. Whereas early entry of these countries may help lock in core values in these countries, premature entry of countries lacking the proper institutional and moral setting may backfire and thus erode mutual understanding. These dilemmas are relevant also in determining the optimal size and speed of inward migration of ethnic groups. 13 European integration may facilitate insurance also through more efficient capital markets. 14 Van Schaik (2000) documents substantial heterogeneity of European countries in terms of culture and institutions.

UNITY PRODUCES DIVERSITY

415

sal convergence of European values in all dimensions (see also Hofste­ de, 2001). In fact, European integration may even cause regional cultures to diverge on certain dimensions as a result of increased specialization. 4 Methodology

4.1 Measuring social capital The multi-dimensional concept of social capital is hard to measure em­ pirically. EVS attempts to measure social capital using a survey ques­ tionnaire. To assess the level of trust in a society, for example, one can consider the question ‘Generally speaking, would you say that most people can be trusted, or that you cannot be too careful in dealing with people?’ Economists are generally skeptical of such attitudinal survey questions, which typically receive only socially acceptable answers. In­ deed, numerous studies have documented that people often do not do what they say. Moreover, questions like the one above are rather abstract and vague. The responses are therefore difficult to interpret. If one says that one trusts most people, is that because of one’s beliefs or because of one’s interpretation of what trusting someone involves or what ‘most people’ means? In fact, these interpretations may vary across countries. Hence, it is not always clear what the (international differences in) re­ sponses to these questions actually measure15. Economists therefore prefer revealed instead of stated preferences: they attach more importance to studying actual behavior of people. In this connection, the experimental method has become popular in eco­ nomics. In experimental games, one observes actual decisions of people in carefully controlled situations in a laboratory. Economists tend to be after stylized, parsimonious theories capturing the fundamentals of hu­ man behavior. To test these theories, they have to be able to carefully control the external environment, which can be done in experiments. So­ ciologists, however, tend to be skeptical about this method because they believe experiments are too artificial to provide useful information about actual behavior of people in complex, socially embedded situations out­ side the laboratory. Moreover, the elements that they believe are impor­ tant in determining behavior, such as the social context implied by the structure of society, cannot be easily manipulated within a laboratory. Finally, whereas sociologists tend to focus on realistic descriptions of actual behavior of groups in specific situations, economists are typically _____________ 15

Methods have been developed to better identify what the answers to questions actu­ ally measure (see, e.g., Scherpenzeel & Saris, 1997).

416

CHAPTER FOURTEEN

more interested in developing a unified view of human behavior with predictive power based on individual behavior. Sociologists and econo­ mists thus tend to choose different locations on the fundamental tradeoff between realism and control. Surveys and experimental games can be combined, thereby combi­ ning the preferred methodologies of economists and sociologists. By correlating answers to survey questions with actual behavior in experi­ ments, one can obtain information about which dimension of social capi­ tal the survey responses actually measure. Glaeser et al. (2000), for ex­ ample, find that responses to concrete survey questions about actual trusting behavior predict actual trusting behavior in experiments better than survey responses to general, non-specific attitudinal questions. At the same time, these latter responses appear to provide some information on the trustworthiness of individuals16. Self-reported trustworthiness in the questionnaire, in contrast, is negatively (but insignificantly) related to actual trustworthiness. These results suggest that specific questions about past trusting behavior are the best measure of trust. This may ex­ plain why Beugelsdijk and Van Schaik (2003) find that regional growth performances are positively correlated with active group membership (i.e. actual past behavior)17 rather than the responses to attitudinal survey questions about trust. 4.2 Measuring the impact of social capital Empirical studies show correlations between trust as measured by survey questions and economic growth across countries (see Beugelsdijk & Van Schaik (2003) for an overview)18. The correlations between growth and survey questions are not always robust. Another problem with these studies is that they find it difficult to control for non-cultural variation. Indeed, trust variables may pick up not only cultural factors (such as so­ cial preferences and general morality) but also institutions, such as the legal and market systems, and reputation mechanisms. Hence, it is not clear whether the correlation between these trust variables and growth _____________ 16

This is consistent with the Dutch saying: ‘Zoals de waard is, vertrouwt hij zijn gasten.’ An interpretation of this result is that people look at themselves when estimating whether others can be trusted. Indeed, people tend to overestimate the degree to which others are similar to themselves. 17 Still, it is not exactly clear what active membership of civic groups actually mea­ sures. To illustrate, it can measure information networks but also the inculcation of con­ cern for others in or even outside the group. Some of these functions of civic groups may have been taken over by other institutions in some countries (such as internet chatgroups, businesses, schools). 18 Similarly, at an individual level, high-income individuals report that they are more trusting. These individuals report also higher membership of non-professional organiza­ tions (see, e.g., Glaeser et al., 2000).

UNITY PRODUCES DIVERSITY

417

implies that the capacity to grow is based on deeply rooted cultural heri­ tages, which change slowly, or the result of legal and market institutions, which can be manipulated more quickly and easily. These empirical studies also do not make clear the direction of causality between economic and cultural variables. What is the cause (i.e. the exogenous variable) and what is the effect (i.e. the endogenous variable)? As noted in sections 2 and 3, economics and culture are mutually dependent and exert feedback effects on each other. Hence, many empirical studies linking growth to social capital suffer from simultaneity bias. In order to inform the debate about the interaction between economic and cultural changes, theoretical models should be developed that define precisely the mechanisms through which various endogenous and exogenous variables interact. Subsequently, the predictions of these models can be falsified19. Theoretical and empirical work should thus proceed in tandem20. 5 Conclusions This chapter has stressed several symbiotic relationships, such as the re­ lationship between economic and cultural changes. On a European level, the internal market is based on the synergy between high levels of gen­ eral morality, an effective legal system, and well-functioning reputation mechanisms. Convergence in essential core public values (such as fiscal and monetary discipline, and respect for property rights, the rule of law, and the rights of heterogeneous minorities) stimulates and is stimulated by European economic integration. At the same time, homogeneity in these core values can produce divergence in non-essential, heterogene­ ous private values. Bridging social capital on the European level can thus be the basis for bonding social capital on the regional level so that European integration goes together with regionalization. Bridging and bonding social capital can be complementary, also because bonding children may provide the basis for bridging adults. Pluralism in values enhances the benefits from intra-European trade. European integration should thus leave substantial scope for diversity. Indeed, modern know­ ledge economies are based on extensive specialization and decentraliza­ _____________ 19

A nice example of such a procedure is Beugelsdijk and Smulders (2003), who em­ ploy EVS to empirically test the implications of their theoretical model. See also Zak and Knack (2002). 20 Also in this connection, experiments can play a useful role because an experimenter can carefully control the situation. This facilitates the testing of particular behavioral theo­ ries. Indeed, experimental work has contributed to the development of new parsimonious theories about how humans behave in social situations.

418

CHAPTER FOURTEEN

tion as well as differentiated, heterogeneous tastes. As Adam Smith al­ ready pointed out, freedom, pluralism, diversity and specialization go together with individual responsibility and restraint based on moral sen­ timents and commitment to basic, common values. On methodology, we stressed the potential benefits from joining the complementary ap­ proaches of sociologists and economists, namely survey questionnaires and experimental games. Moreover, theoretical and empirical work should stimulate each other in uncovering robust relationships between economic and cultural developments. The key question that arises from this chapter is which common core public values are required for successful European integration and in which areas the various European countries and regions can retain and even strengthen their own specific cultural heritage. Together with ex­ periments and theoretical work, EVS can play a useful role in shedding light on this and other issues. References Becker, G.S. 1993. Nobel Lecture: the economic way of looking at behavior. Journal of Politi­ cal Economy 101: 385–409. Beugelsdijk, S. & A.B.T.M. van Schaik 2003. Participation in civil society and European re­ gional economic growth. Chapter five in this volume. Beugelsdijk, S. & S. Smulders 2003. Bridging and bonding social capital: Which type is good for economic growth? Chapter six in this volume. Binmore, K. 1998. Playing Fair. Cambridge: MIT Press. Bisin, A. & T. Verdier 2001. The economics of cultural transmission and the dynamics of pref­ erences. Journal of Economic Theory 97: 298–319. Bovenberg, A.L. 2002. Norms, values, and technological change. De Economist 50: 521–553. —— & A. de Jong 1996. De EMU: Hoe en waarom? Economisch Statistische Berichten. 81: 420–424. —— & C.N. Teulings 1999. Concurrentie als alternatief voor rechtsprincipes. Economisch Sta­ tistische Berichten 84: 364–367. CPB Netherlands Bureau for Economic Policy Analysis 1997. Challenging Neighbours: Re­ thinking German and Dutch Economic Institutions. Berlin: Springer Verlag. Durlauf, S. 1996. A theory of persistent income inequality. Journal of Economic Growth 1: 75– 93. Fehr, E. & A. Falk 2002. Psychological foundations of incentives. European Economic Review 46: 687–724. —— & U. Fischbacher 2002. Why social preferences matter – The impact of non-selfish mo­ tives on competition, cooperation, and incentives. Economic Journal 112: C1–C33. Francois, P. 2002. Social Capital and Economic Development. London: Routledge. —— & J. Zibojnik 2001. Culture and development: an analytical framework. CentER discus­ sion paper. No. 25. Tilburg: Tilburg University. Glaeser, E.L., D.I. Laibson, J.A. Scheinkman & C.L. Soutter 2000. Measuring trust. Quarterly Journal of Economics 115: 811–846. Graafland, J. 2002. Modelling the trade-off between profits and principles. De Economist 150: 129–154. Greif, A., P.R. Milgrom & B.R. Weingast 1994. Coordination, commitment, and enforcement: The case of the merchant guild. Journal of Political Economy 102: 745–776.

UNITY PRODUCES DIVERSITY

419

Hofstede, G. 2001. Culture’s Consequences: Comparing Values, Behaviors, Institutions, and Organizations Across Nations. Thousands Oaks: Sage. Inglehart, R. 1990. Culture Shift in Advanced Industrial Society. Princeton: Princeton Univer­ sity Press. —— & W.E. Baker 2000. Modernization, cultural change, and the persistence of traditional values. American Sociological Review 65: 19–51. Kay, J. 1993. Foundations of Corporate Success. Oxford: Oxford University Press. Rabin, M. 2002. A perspective on psychology and economics. European Economic Review 46: 657–685. Scherpenzeel, A.C. & W.E. Saris 1997. The validity and reliability of research questions: a meta Analysis of MTMM. Sociological Methods and Research 25: 341–383. Sinn, H.W. 1990. Tax harmonization and tax competition in Europe. European Economic Re­ view 34: 489–504. Smith, A. 1759. The Theory of Moral Sentiments. D.D. Raphael & A.L. Macfie (eds.) 1976. Oxford: Clarendon Press. Tirole, J. 2002. Rational irrationality: Some economics of self management. European Eco­ nomic Review 46: 633–655. Van Schaik, A.B.T.M. 2000. Americanization? Tilburg: Tilburg University (mimeo). —— & S. Beugelsdijk 2000. Cultural change and economic development, a comment. Tilburg: Tilburg University (mimeo). Zak, P.J. & S. Knack 2001. Trust and growth. Economic Journal 111: 295–321.

This page intentionally left blank

CHAPTER FIFTEEN

EUROPEAN IDENTITY AND INTERRELIGIOUS DIALOGUE ERNST HIRSCH BALLIN 1 Introduction The classic European treatises on tolerance—by Erasmus, Spinoza, and Locke—mainly concerned religious tolerance. Tolerance of people with a different belief is still one of the values on which Europeans pride themselves. This is clear from the European Values Study surveys, in which an attempt is made to ascertain the values of Europeans. To ran­ dom samples of the populations of European countries, all kinds of ques­ tions were presented about their opinions on a variety of fields. The 1999/2000 EVS (Halman et al., 2002) shows that the European popula­ tions are relatively tolerant. The majority do not object to living next door to people of a different race, immigrants, or adherents of a different faith. People have most problems, however, with Muslims, and this situation has not improved since September 11. Therefore, popular sup­ port among the populations of the European countries for a government policy aiming at religious tolerance, especially with respect to Islam, is limited. The danger of sociological survey research, as that of the EVS, is that its findings will function as the most significant guideline for government policy. This would make public opinion the highest ground of legitimacy. Public opinion, however, is not a static entity. ‘(C)alls to find a common ground and to deliberate in a civil manner’ (Etzioni, 2001: 108) lead to the values and convictions in a society interacting with one another, resulting in insights that cannot as yet be deduced from surveys. This more demanding form of tolerance raises the ques­ tion as to how far the variety of convictions can go, if one wants to bring them together in the unity of one society. Consequently, the question concerning the measure of religious resilience, in post-September 11 Europe, is a most pressing one.

422

CHAPTER FIFTEEN

In the European societies, the question as to religious resilience is ever more sharply coming to the fore. Christians, Jews, and humanists have managed to get along for quite some time now, and until recently a level-headed approach to the integration of other religions, such as Is­ lam, seemed to belong to the possibilities. Mostly, however, contacts re­ mained superficial. Religious variety was ignored rather than looked up­ on as a possible problem. After September 11, doubts arose about the tenability of that indifferent view of tolerance. Some argue for more complete assimilation, others for more control-based screening off. Those who are aware, however, of the tensions in the Muslim world pre­ fer to intensify the dialogue with Islam (Cf. Lewis, 2001). The motive for this cannot be the hope of appeasement of Osama bin Laden, alQuaeda and related movements. The extreme crimes they committed do not lend themselves to being covered up by cowardly conciliatoriness, since true peace can only arise from justice. In addition, it would be fun­ damentally wrong to identify Islam with these terrorist movements, no matter how much these movements would like us to. The virtue of toler­ ance must not be an alibi to allow injustice to grow rampantly. The considerations below about tolerance and interreligious dialogue were written in the awareness that precisely the relationship with Islam requires special attention. Often, tolerance is lightly praised as a charac­ teristic of ‘our’ society. One should realize at this point that this tole­ rance is the result of the unanimous view, characteristic of Western Eu­ rope, that the holy and the profane, the spiritual and the temporal, the re­ ligious and the secular, are to be distinguished from each other. ‘The two may be joined, or in modern times, separated’ (Lewis, 2001: 98). What remains, however, is ‘the idea that religion and political authority, church and state are different.’ This thought ‘is, in a profound sense, Christian’ (Lewis, 2001: 96). In present-day Islam, however, this very principle is anything but generally accepted. In states with a predomi­ nant Muslim population, Turkey is an exception, since it has constitu­ tionally guaranteed the separation between the state and the religious in­ stitutions (Lewis, 2001: 108). Therefore, the acceptance of ideological variety, if ‘the’ Islam be­ longs to it, requires more than a prolongation of the existing tolerant cli­ mate. This climate could be taken for granted in the relations between communities that—no matter how strenuously they have fought one another—have much in common, precisely as regards their views on the political and social order. Christianity, Judaism, and humanism are historically and ideologi­ cally more closely related than is usually realized. The EVS research brings to light assessment patterns that to a large extent are the product

EUROPEAN IDENTITY AND INTERRELIGIOUS DIALOGUE

423

of shared experiences in the common history of Western Europe. That in the New Age, Islam has been outside it for centuries—apart from rela­ tions on the Balkans with the multifaceted Osman Empire—does not mean, however, that it is possible or desirable to continue to see Islam as a ‘foreign’ element. The already existing mutual incomprehension is reinforced when en­ emy images get the upper hand. To speak in terms of a clash of civiliza­ tions is to characterize incorrectly the present conflict with a movement that appeals to a fundamentalist and violent interpretation of Islam. Nor is the clash of civilizations described by Samuel Huntington an inevi­ table future development. What is happening is the abuse of religion to set the people at one another’s throats. Humankind has seen this before. Anyone who experiences faith as a duty to work for peace must deeply regret that history has been sullied with wars and violent acts for which people were mobilized with an appeal to their faith. This also goes for many centuries of Christianity, with respect both to dissenters and to Christians amongst themselves. This history is not counterbalanced ei­ ther by the fact that countless believers have become victims of perse­ cution by atheist fundamentalists – for example, in the former Soviet Union and in China. It was these desecrations of faith—by violence in the name of faith— which prompted the Pope’s great Mea Culpa, at the beginning of Lent in the year 2000. It does not seem superfluous to recall this before— again—casting the first stone in a false battle of faith. But a solemn in­ tention on the part of the leaders of the world religions to devote them­ selves to peace, no matter how valuable, does not suffice to prevent such a catastrophe. Mostly, however, the real religious leaders do not play a role in such conflicts, or at best a marginal one. The infamous struggle between Catholics and Protestants in Northern Ireland is not the work of bishops or synods, but of political leaders who in an essentially social conflict use faith to mobilize their followers1. In Bosnia, people often did not know—until the moment when the Yugoslav political system collapsed—which of their friends were Serbs or Croats, and which were Muslims. Hence, interreligious dialogue must not be misunderstood as a recognition that the violence is a consequence of the intolerance of relig­ ions. What interreligious dialogue can be is an attempt to prevent relig­ ion from being abused in this manner. _____________ 1

Since fundamentalism also transforms religious doctrines to a political programme and has totalitarian aspirations, fundamentalist convictions have such a mobilizing poten­ tial par excellence. This does not mean that fundamentalism and terrorism are necessarily linked (Cf. Tibi, 2001: 15, 24 and 154).

424

CHAPTER FIFTEEN

2 A complicated relation As was said before, the branches of Islam whose actions dominate the headlines represent different attitudes towards law, politics, and culture than is deemed acceptable in Western European value patterns. For pre­ cisely this reason, as argued in the manifesto Diversity Within Unity, published in 2001 by an international group of (ex-)politicians and aca­ demics (CN, 2002) established on the initiative of Amitai Etzioni, nei­ ther getting rid of the differences imposed by assimilation policy, nor ‘unbounded multiculturalism’ is the right answer. What should be the guiding principle is a combination of respect for the whole and respect for each individual. It is obvious that a predominant view of religion is seen as a threat by those who do not share that view, including Muslims. Only by in-depth exploration, involving the idealistic and religious va­ riety in the (further) development of communality, can the one exclusion be prevented from evoking the other. This is precisely why interreligious dialogue must be more than a noncommittal exchange of views. At the beginning of the 21st century, much in Western Europe that in the past had led to discrimination and worse has been replaced with an emphasis on shared values. For the re­ lationship between Christians and Jews—respectfully referred to by the Pope as ‘our older brothers in faith’—new, wholesome perspectives have opened up in recent years, after a journey through a valley of tears. Humanism is manifesting itself both through Christianity and apart from it. The relationship between Christianity and Islam, however, is an ex­ tremely complicated one in all kinds of respects, both for reasons of faith and for historical reasons. It must be added that the idea of the state un­ der rule of law and democracy has hardly taken root in the Arab world (it has, albeit with varying degrees of success, in other Muslim countries, such as Indonesia and Turkey; and in some Arab countries there are also interesting initiatives2). The dependence of the West on oil from the Middle East has set in motion a very substantial flow of money in the di­ rection of the Arab world. Strangely, this flow of money has not resulted in a general cultural and economic revival of these countries – but it has led to growing frustration of the majority of the population. This is even exacerbated by the material, not to say materialistic, parading of wealth by Westerners.

_____________ 2

See, especially about Egypt, e.g., Cotran and Sherif (1999).

EUROPEAN IDENTITY AND INTERRELIGIOUS DIALOGUE

425

3 The contribution of Islam to European culture The current lags in the cultural and economic development cannot be blamed on the Islam or Arab culture. It is precisely in the Arab world from the 8th to at least the 13th century that Islam—as a religion no less varied in its range of thought and its manifestations than Christianity— saw an enormous flourishing of science. The Muslim contribution to Western civilization—this was the title of the inaugural address of F. Sierksma at Leyden University in 1974—concerned many branches of science. It also encompassed the conception of the university. The ‘uni­ versity as a cultural institution, in the context of which lay people em­ pirically and experimentally investigate sensory reality in different facul­ ties,’ says Sierksma, ‘is, if not entirely, then at least largely ‘(...) an Arab invention’: the great university of Bagdad was founded in 1065; in Cór­ doba this example was followed, and the crusader emperor Frederick Barbarossa was the person who in Bologna founded the first university in Christian Europe (Sierksma, 1974: 17). Consequently, the crusaders where the ones who through their con­ tacts with the Muslims came home with more culture than what they had set out with. ‘For the Muslims’—I once again quote Sierksma (1974: 7)—‘the Franks were simply barbarians, who only as a result of enough contact with Muslim culture started to look a little like humans.’ Fortu­ nately, these contacts not only took the form of armed combat but also of trade and, after the settlement of the crusader kingdoms, of fairly pro­ longed communal living. Another historian, G.E. van Grunebaum (1986: 178), claims that the crusades, despite their failure, have thus only bene­ fited the Occident. The crusades were really a second front besides the battle for the Ibe­ rian peninsula that had been raging for much longer. According to histo­ rians, the idea for the crusade has itself been directly borrowed from Islam’s jihâd, turning it into its opposite. Spain was the other territory where Muslims, Christians, and Jews ran into each other. Here, despite the military battle over Andalusia, el An­ dalus, Muslims, Christians, and Jews lived peacefully together for centu­ ries, too—from the 8th to well into the 11th century—in accordance with the duty to protect them that the Muslim rulers had accepted on the grounds of the Koran (9, 29), (Hagemann, 1999: 15). In the caliphate of Cordoba, there was an unprecedented flourishing of science. Libraries disclosed knowledge from Greek antiquity, and doctors and practitioners of the natural sciences were held in high regard. In 1492, after the fall of Granada, cardinal Cisneros ordered the public burning of thousands of Muslim books in the Plaza Bib Rambla because it was feared that they

426

CHAPTER FIFTEEN

contained heresies. In a manuscript found at the time by one Jaime Fer­ rán after the Moors had fled into the mountains, he wrote in Catalan that he was unable to find anybody who could decipher this Arab writing; hence he was afraid it might be Mohammed’s Koran. However, this book, which is now kept in the university library of Valencia, was a grammar (Molina, 1994: 169). In the 12th and 13th centuries, Europe, in spite of all the fighting, was the stage of a great intellectual—philosophical and theological—debate associated, among others, with the names of the Muslim Averroes (1126–1198), the Jewish philosopher Maimonides—born in Córdoba in 1135—and Thomas Aquinas (1225–1274). They shared a deep respect for, and intellectual interest in, the pre-Christian Greek philosopher Aris­ totle. The Summa contra Gentiles, one of Thomas’ most important works, is a theological criticism of Islam – and precisely because of that a document of their contentious proximity. The academic life in Paris, which in those days was no less tumultuous, played an important role in it within Christianity. A recent culture-historical study of medieval Spain, which was often governed with remarkable leniency by Muslim rulers, calls this multicultural community The Ornament of the World (Menocal, 2002). This history deserves recalling for two related reasons. In the first place, I want to underline that Islam—like Christianity, for that matter— has manifested itself in very divergent ways. The other reason is that it would be a historical fallacy to consider the culture of Islam to be infe­ rior to that of Christian Europe. If an interreligious dialogue is to bear fruit, what must be found, besides mental willingness, is a basis of communality. In the past, this existed in the desire to account reasonably for the possibility of believing. For Maimonides, Averroes, and Thomas, therefore, the philosopher Aristotle’s thinking was (almost) the norm. Thus Aristotle was for them the embodiment of thinking as a generally human possibility. This does not diminish the significance of God’s mer­ cy: it does indicate, however, that it is man for whom Revelation is knowable. 4 Respect as a condition for interreligious dialogue Interreligious dialogue can only bear fruit if one is prepared to respect the other as a human being. Therefore, fundamentalism—no matter of which denomination—is an absolute obstacle to such a dialogue. For the fundamentalist, the ‘other’ religious conviction does not count, and communality only begins after the other has ‘surrendered’ to the true faith. Nor can there be a real interreligious dialogue if the interlocutors

EUROPEAN IDENTITY AND INTERRELIGIOUS DIALOGUE

427

see their own religious convictions as no more than the product of their adventitious upbringing and background. This relativism is an inverted kind of fundamentalism: nothing is of such a value that it would be worth transferring to somebody else. On superficial observation, interreligious dialogue should be no prob­ lem in the Netherlands. In my view, however, we are facing a consider­ able problem in that we are forced to do the splits between relativism and fundamentalism. People with a ‘strong’ religious conviction often give it an almost fundamentalist interpretation. In their eyes, it is not really possible to talk to those who do not embrace the same faith. It will be clear that such a view is not in keeping with Catholic doctrine, with its appreciation for the reasonableness of every human being. Neverthe­ less, believers are often indiscriminately blamed for it, alas: in the Dutch media, after September 11, 2001, commentators have every now and then come up with the message that religion again and again turns out to be a source of intolerance and violence. On the other hand, there is—not only among unbelievers but even among the ‘moderately’ religious—a strong tendency towards relativization, causing the willingness to propa­ gate faith to wane. It seems possible quickly to find a practical way out: let’s meet in a good atmosphere and let’s not make things too difficult – so what if we do not have an interreligious dialogue? Is it not enough for a multicul­ tural society not to bother each other? I think this is not enough, not enough to develop real relations, and not enough to come to grips with the border-crossing issues of fundamental misunderstanding. This is the model of ‘multicolouredness’, a patchwork society: lovely to see, but a society of loose pieces put together. Thus, it can happen in the Nether­ lands that Muslim parents who have come to ask if their children will not experience problems with their faith at the Catholic or Protestant school in their neighborhood are given a reassuring answer by the head­ master along the following lines: ‘Don’t you worry, Sir, Madam, they will hardly notice that this is a Christian school. We don’t talk about it a lot, so you really won’t be bothered.’ That the Muslim children are wel­ come is fine, but in the seemingly tolerant answer something is hidden that is far less tolerant, namely: ‘please keep your faith private, too, that’s the way we like it here.’ Multiculturalism with the covert message that religion is a private matter that we had better keep silent about in public is therefore not all that tolerant. It excludes the deepest motives of the encounter between the various communities that make up a society. In contrast, inter­ religious dialogue requires respectful openheartedness. It is therefore ap­ propriate that the call of the German bishops for a respectful dialogue

428

CHAPTER FIFTEEN

between Christians and Muslims also contained the warning that dif­ ferences in perception should not be glossed over. ‘Wir müssen uns ehr­ lich und in gegenseitigem Respekt sagen, wie wir zueinander stehen. Kritische Fragen dürfen nicht ausgeklammert werden. Wir setzen uns dafür ein, dass Muslime in unserem Land ihrem Glauben entsprechend leben können. Gleiche Bedingungen müssen auch für Christen in islami­ tischen Ländern gelten.’ In an interview on September 21, 2001, Car­ dinal Meisner (Cologne) said: ‘Nun muss man sicherlich unterscheiden zwischen den islamitischen Terroristen und dem Islam als Religion. Aus dem Ansinnen einiger weniger Terroristen einen Krieg der Kulturen zu folgern ist sicherlich zu weit gegriffen. Wohl aber sollten wir den Unter­ schied unter den Religionen ernst nehmen! Das Christentum unter­ scheidet sich vom Islam durch ein anderes Gottes- und Menschenbild.’ It is therefore important to go more deeply into this portrayal of man­ kind. In the second half of the 20th century—undoubtedly also as a reac­ tion to the holocaust—Christianity’s portrayal of mankind was theo­ logically and ethically rethought, especially by thinkers like Jacques Maritain (2000). Since the encyclical Pacem in terris (1963), the rights of man have become part of the church’s teaching, and were linked, es­ pecially by Pope John Paul II, with the doctrine of the transcendental dignity of the human person, which goes back to being created as a visi­ ble portrayal of the invisible God (Dulles, 1999: 147). This culminated in the Pope’s plea for a new humanism at the dawn of the Third Mil­ lennium. Thus, Christianity has—with ups and down—accepted the personal dignity of every human being and with it the human rights man as a ba­ sis for the ordering of society. In any case, this fact puts us at a great dis­ tance from a not unimportant number of Muslim countries. It also ex­ plains that in the Western European context many tend to lay down a kind of precondition for the dialogue with Muslims, saying something like, ‘all this talk is fine, but the first thing to ascertain is your loyalty to the principles of the democratic state under rule of law.’ I fully agree that the state under rule of law and democracy must never be sacrificed to a policy of appeasement. It would, however, be a false start to begin the conversation by demanding a kind of declaration of loyalty. A dia­ logue always means accepting the other as well. This implicitly lays the foundation for the protection of basic rights, namely the mutual accep­ tance and thus the mutual respect for every human being. Hence, the ac­ ceptance of the state under rule of law and democracy must be seen as an expected outcome of a successful interreligious dialogue rather than as a precondition.

EUROPEAN IDENTITY AND INTERRELIGIOUS DIALOGUE

429

Here lies the essential distinction between a real interreligious dia­ logue and the much-vaunted multicultural tolerance of the Netherlands. This self-proclaimed tolerant, multicultural, and multi-religious society does not have a very impressive aura, however. Marcel ten Hooven re­ ferred to it as ‘empty tolerance’, quoting the historian Kossmann, who claimed that tolerance—as the uninterested acceptance of a number of ‘oddities’—even has ‘something discriminatory’ (Ten Hooven, 2001: 27). This certainly goes for the relativization of religiously anchored values, and the banning of religion to private life of which I have just given an example (‘Why don’t you leave your religion at home?’). A climate in which the different religious communities get to know one another better, and inspire each other in mutual respect, will not arise in such a way. Theo de Wit (2001: 95) warns against a tendency to look upon values, truths, and ideals as subjective preferences of the indivi­ dual, which must be treated with consideration by the government. Tolerance that is based on respect for everyone’s deepest convictions is therefore in principle distinct from what Marcel ten Hooven described as empty tolerance and what Anton Zijderveld (2001: 185) called ‘nega­ tive tolerance’. Doing justice to each other means, to begin with, respec­ ting each other, with every human being enjoying the inalienable right to be respected. Where convictions fall demonstrably short in this funda­ mental respect for every human being—I am here thinking of pheno­ mena like the sexual mutilation of girls and the exclusion of women from public life, leaving aside whether these are rightly related to reli­ gious convictions—nobody is legitimized to keep silent about it – on the contrary! Respect in interreligious relations requires a certain frankness, but even more, a willingness to listen and to study. 5 An in-depth dialogue These remarks about the conditions for interreligious dialogue have, I hope, clarified two things, which I will now briefly summarize in con­ clusion of this contribution. The first is that such a dialogue is possible, and the second that it will not be easy to realize in practice. There is no way back to the Córdoba of a thousand years ago, and if there were, we would have to observe, besides a flourishing culture, many violations of human rights as well. In our 21st century conditions of life, we do find Muslim organizations in the countries of the European Union, but only few initiatives for an in-depth exchange between Islam and modern Western culture. So far, the existing Muslim organizations have been given a one-sided emphasis in the sphere of social work and of commu­ nity building, also as a result of the subsidizing policy of the govern­

430

CHAPTER FIFTEEN

ment. An unintended effect of this is that for theological ‘nourishment’, people have continued to depend on links with the countries of origin of the first generation. In this part of the world, hardly any Muslimtheological institutes of a(n international) scientific level have got off the ground. This is counterproductive not only from the perspective of inte­ gration, but also from the perspective of interreligious communication. From an international perspective, a unique opportunity is being missed to build a bridge towards the Muslim world in a Western European con­ text. If this opportunity is seized, however, the interlocutors in the interre­ ligious dialogue will, on either side, be facing questions that so far have hardly been addressed. Not only in the theological field, this dialogue will have to extend further than the usual noncommittal remarks, such as the assurance that we are all Abraham’s children (Leggewie, 2001): ‘Wir sollten also nicht allein über die Präliminarien reden (...), sondern mehr zu den Sachen selbst kommen. Und das wären so lebenspraktische Fragen wie Feierstagarbeit, Chancen und Risiken der Gentechnologie und Lebenswissenschaften, Marktversagen, Informationsfreiheit – und dergleichen mehr.’ To give substance and quality to the tolerance that we so eagerly claim for the European societies, it would be a good thing if also on the part of Islam a vital theological and philosophical link was established between tradition and current questions of communal life3. To achieve this, initia­ tives can be observed in a number of scientific institutes in Muslim countries, but in practice, crash courses for imams have often been felt to be sufficient. Interreligious dialogue presupposes a well-considered and thoroughly felt relationship with whatever belongs to one’s own faith. But to the call implicit in this, the question must be linked—in all modesty—whether Christianity itself lives up to its standards in this respect and whether it does not all too often limit itself to self-indulgent acqui­ escence4. This—and perhaps only this—can help to prevent a reversal of the reversed jihâd.

_____________ 3

The grandmufti of Marseille, Soheib Bencheikh, feels left in the lurch in his own country by government policy, because this in fact legitimizes fundamentalist groups pre­ cisely by giving them a place in a French Muslim council (International Herald Tribune, 30 November 2001). 4 ‘Die Christen haben zu viel Kleinglauben, eine zu große geistig-spirituelle Trägheit und Feigheit. Sie brauchen mehr Mut zum Bekenntnis und mehr Freude am Evangelium. Dann brauchen sie auch keine Angst mehr zu haben vor der gegenwärtigen Her­ ausforderung.’ (Cf. Lehmann, 1997: 4-16).

EUROPEAN IDENTITY AND INTERRELIGIOUS DIALOGUE

431

6 Diversity requires unity Communities have symbols, shared memories, and institutions in which their coherence finds expression. The idealistic and cultural identity of people is determined by the experiences and memories they share with others. A mutual interest in the experiences and memories—and a com­ mon language, in which one can make oneself comprehensible for one another—is necessary if, in a society with many cultures, unity is also to be made recognizable. These cultures, in their turn, will then find their expression in the story and the symbols of society as a whole. The mani­ festo Diversity Within Unity says about this: ‘Only as these deeper is­ sues are addressed might societies be able to work out satisfactory reso­ lutions of the symbolic issues.’ It seems that people are insufficiently aware that ‘diversity’ is only safe in relation with ‘unity’. If diversity were the one and only issue, it would soon turn out to be a dead loss. Without unity being in place at such a fundamental level that it is capable of sustaining diversity, the in­ tolerant mind would have no incentive to realize unity without this being at the expense of others. A peaceful dialogue can only be established in mutual recognition of diversity on the basis of unity. References CN 2002. Diversity Within Unity. A Communitarian Network Position Paper. Washington: The Communitarian Network. http: //www.gwu.edu/~ccps/dwu_positionpaper.html. Cotran, E. & A.O. Sherif (eds.) 1999. Democracy, the Rule of Law and Islam. The Hague /London/Boston: Kluwer Law International. Dulles, A. 1999. The Splendor of Faith, The Theological Vision of Pope John Paul II. New York: Herder & Herder. Etzioni, A. 2001. The Monochrome Society. Princeton/Oxford: Princeton University Press. Grunebaum, G.E. 1986. Der Islam. Pp. 121–179 in G. Mann & A. Nitschke (ed.), Propyläen Weltgeschichte. Part 5. Berlin/Frankfurt a. M.: Propyläen. Hagemann, L. 1999. Christentum contra Islam, Eine Geschichte gescheiteter Beziehungen. Darmstadt: Wissenschaftliche Buchgesellschaft. Halman, L. & O. Riis (eds.) 2002. Religion in Secularizing Society. The Europeans’ Religion at the End of the 20th Century. Leiden-Boston: Brill. Huntington, S.P. 1997. The Clash of Civilizations and the Remaking of World Order. London: Simon & Schuster. International Herald Tribune 2001. A Progressive Muslim Struggles to be Heard. 30 November. Leggewie, C. 2001. Römische Minarett und Deutscher Islam, Wie weit geht der religiöse Pluralismus. Bad Homburg v.d. Höhe: Herbert Quandt-Stiftung. Lehmann, K. 1997. Das Neue Europa. Konrad-Adenauer-Stiftung Auslandsinformationen 13: 4-16. Lewis, B. 2001. What Went Wrong: Western Impact and Middle Eastern Response. Oxford University Press. Menocal, M.R. 2002. The Ornament of the World, How Muslims, Jews, and Christians Created a Culture of Tolerance in Medieval Spain. Boston/New York /London: Little, Brown & Company.

432

CHAPTER FIFTEEN

Molina, A.M. 1994. Die Stadt der Kalifen, Historische Streifzüge durch Córdoba (tr. K. Jetz). Reinbeck bei Hamburg: Rowolt. Maritain, J. 2000. Humanisme intégral, Problèmes temporels et spirituels d’une nouvelle chré­ tienté. Paris: Aubier. Sierksma, F. 1974. Een en ander over de Moslimse bijdrage aan de Westerse beschaving. Lei­ den: Universitaire Pers Leiden. Tibi, B. 2001. Fundamentalismus im Islam, Ein Gefahr für den Weltfrieden. 2nd edition. Darm­ stadt: Wissenschaftliche Buchgesellschaft. Ten Hooven, M. (ed.) 2001. De lege tolerantie, Over vrijheid en vrijblijvendheid in Nederland. Amsterdam: Boom. Wit, Th.W.A. de 2001. De trivialisering van de tolerantie: Zowel ‘te veel’ als ‘te weinig’ be­ dreigt de verdraagzaamheid. Pp. 85–108 in M. ten Hooven (ed.), De lege tolerantie, Over vrijheid en vrijblijvendheid in Nederland. Amsterdam: Boom. Zijderveld, A.C. 2001. Negatieve tolerantie: Verdraagzaamheid wordt autoritair als weder­ kerigheid ontbreekt. Pp. 185–190 in M. ten Hooven (ed.), De lege tolerantie, Over vrij­ heid en vrijblijvendheid in Nederland. Amsterdam: Boom.

CHAPTER SIXTEEN

ONE EUROPEAN WORLD OF WELFARE OR MANY: THE ROLE OF VALUES, BEHAVIOR AND INSTITUTIONS RUUD MUFFELS 1 Introduction A number of comparative studies including the European Values Study have documented widespread social, economic and cultural diversity in Europe. This chapter deals with values, institutions and behavior con­ cerning the welfare state and the question addressed here is whether this diversity as documented by these studies might be a barrier to attaining a more integrated Europe in social terms1. The endeavor to attain a ‘Social Europe’ as it is coined in the European debate presumes that there should be more emphasis on ‘social’ in ‘European’ policy and that ‘Europe’ should be more prominent in the ‘social’ policies of member states and, indeed, sub-national tiers of governance. There is ample po­ litical debate on the viability and popular support of more coordination and even integration in the social domain. Representatives of the labor unions and political parties at the left side of the political spectrum sometimes argue that the underlying institutional and ideological dispari­ ties across the nation-states are so large that it is not viable to put more social into Europe and more Europe into the social domain. Others argue that attempts to weaken the aim to attain a ‘Social Europe’ are futile while the social and economic forces at the international level require more integration and convergence of the various national systems. Even if one would adopt such a view on the future of Europe it is still unclear whether the integration process would imply a spiraling downward (a _____________ 1

In this chapter we focus on welfare state values by which we mean values associated with the principles embedded in welfare systems such as distributive justice, solidarity, fairness and support for the role of the state. However, these values are affected by more general values like social-cultural values, religious values and political values. Although we focus on the relationship between welfare state values, welfare state institutions and welfare state behavior the information required includes also information on the broader set of value structures while they affect welfare state values.

434

CHAPTER SIXTEEN

‘social dumping’ process) of the level of welfare state benefits to the least generous ones in Europe (Portugal) or an upward leveling process (‘California effect’) to the most generous ones (Sweden). Much of what is likely to happen depends on the role the Commission wants to play and in reality can play given the restraints of the coordination process on social policy. In the social domain the Commission’s legislative power is governed by the subsidiarity principle implying that the role of the Commission to manage the coordination process is restricted because the member states are considered primary responsible for the design and im­ plementation of welfare state policies. Some argue that the coordination process already entered an irreversible stage with the launch of major social initiatives at the Lisbon and Nice European Councils in March and December 2000. The approach adopted at these summits to attain coordination on social issues came to be known as the ‘open method of coordination’ (OMC) which to date has been applied in the domain of employment and social inclusion policies. It represents a new form of policy integration in which the role of the European bodies is to set the framework and objectives, and to orchestrate the monitoring and review of the action plans on employment and social inclusion, while leaving member states free to decide on detailed policies and their implementa­ tion. During the Belgium presidency the Belgium Minister Vanden­ broucke (Esping-Andersen, 2002) expressed high expectations about the process of open coordination in the social field. It is in his view ‘the most promising way to give concrete shape to social Europe’ and ‘to firmly anchor it into the European co-ordination process as a common good’. The challenges Europe faces are due to emerging new risks that are a consequence of more fragile and fragmented work patterns over the life course and rising numbers of unstable families as a result of marriage break-ups. These challenges stress the need to adjust the European social model and even to formulate a new paradigm for social policy. A recent co-edited book by Esping-Andersen (2002) got the illustrative title ‘Why We Need a New Welfare State’ to express that Europe needs a new so­ cial architecture to create an ‘active and dynamic welfare state’—as it is coined at the Lisbon summit—to cope with these emerging new risks. Also Esping-Andersen sees the model of open coordination as a new promising tool to arrive at this ambitious goal. The question to be answered yet is whether there is cultural support among the populations of the existing and candidate member-states for a more integrated social Europe. The European Values Study is the unique data source for answering this question that is at the heart of the current debate on the ‘European Social Model’. But value research is not only

ONE EUROPEAN WORLD OF WELFARE OR MANY

435

praised for its scope and endeavor to document value structures and value change but also criticized by others, especially by economists, who raised doubts about the reliability of responses to attitudinal questions in large surveys. Their critique boils down to two arguments, first that ‘stated preferences’ might not represent the true motivations and beliefs of people but merely the existing social norms in society and second that values are very unstable over time. They believe that ‘revealed prefer­ ences’ as they can be deducted from real behavior render more accurate information about the true preferences of people. This approach to look at behavior instead of values and motivation is implicitly adopted in socio-economic research where the focus is on looking from a crossnational perspective to the real outcomes of the social and economic in­ teraction processes as a result of actual behavior. But whether we focus on values or on behavior the basic question to be answered is whether Europe in terms of values, institutions and be­ havior is currently moving to one European social model or to a diver­ sity of clearly distinct social models. Social scientists dealing with the evolution of the welfare state have argued that countries are not unique but that they cluster around some particular configurations or welfare state regimes. These attempts were conducted in the domain of value studies (Inglehart & Baker, 2000; Arts & Gelissen, 2001, 2002; Ter Meulen et al., 2001; Gelissen, 2002) but also in the domain of socio­ economic research (Esping-Andersen, 1990, 1999; Goodin, Headey & Muffels, 1999). The results show that these attempts seem to make sense: there is ‘Unity in Diversity’ as well as ‘Diversity in Unity’. How­ ever, in value research the diversity found is very much dependent on the type of values studied (work values, social security values, religious values, political values et cetera) and the subject of study (pensions, health, welfare et cetera) and very much different from the diversity found in socio-economic research. It remains therefore an open question how the two modes of classifications are related. To solve this issue an integrated design is needed in which values and institutional differences in the socio-economic domain are jointly exami­ ned (see footnote 1). Due to lack of information in the questionnaires of value research no systematic account is given of the relationship be­ tween value structures on the one hand and institutional and socio­ economic differences on the other whereas in socio-economic research for the same reason differences in value structures are seldom captured. This chapter focuses on the relationship between cultural and socio­ economic factors in explaining the documented diversity in the social domain in Europe and the challenge it offers to the values study frame­ work. In section one we highlight the questions addressed in compara­

436

CHAPTER SIXTEEN

tive value research and in comparative socio-economic research. Then in section two we sketch a heuristic framework for examining jointly the role of values and institutions in the performance of welfare systems. In section three we will discuss some results in which we look at the role of institutions explaining variations in outcomes across countries and regi­ mes. In the last section we draw some conclusions with respect to the role of values, behavior and institutions in explaining diversity across countries. 2 comparative research into the welfare state

2.1 Value research One of the key issues in the value literature is the relationship between cultural values and economic development. In the first macro analyses of the data from the European Values Study and the World Value Surveys the diversity across countries in cultural values was expressed by assign­ ing country labels. It was however rather unclear what the country label represented. To solve this problem reference was made to modernization theory (Bell, 1973), which states that cultural values are shaped by eco­ nomic development. The solution was to assign a figure to each country label indicating the extent of modernization. In later research also other factors were involved (language, family, religion, political ideology) to draw rather impressionistic lines in country plots of value structures (cultural zones) with levels of economic welfare and growth and to as­ sign equally impressionistic labels to these country groupings (see In­ glehart, 1997; Inglehart & Baker, 2000). Progress was made in using cluster analysis to define particular groupings of countries, but these clustering techniques were data driven and not theoretically under­ pinned. A theoretical advancement was attained in using EspingAndersen’s typology of welfare regimes to explain the diversity in value structures. The findings suggest that regime type effects are significant but small. It also emerged that the clustering was very dependent on the sort of values studied and the subject of study involved. An accompany­ ing issue is that regimes are not stable entities; they might change over time due to economic and social forces. The issue raised then is how re­ gimes change and how values change and whether they converge or not. Modernization theory has been used to explain how values alter over time. The expectation was, that the driving political and economic forces behind modernization lead to convergence of values. In the process of modernization traditional values such as economic and physical security (survival values) would be replaced by postmodern values of self­

ONE EUROPEAN WORLD OF WELFARE OR MANY

437

expression, subjective well-being and quality of life (Inglehart & Baker, 2000). Other authors among which Weber ((1904) 1958), argue that cul­ tural values have an enduring and autonomous impact on the institutions of a society. Cultural traits endure over time and shape a society’s politi­ cal and economic performance. Societies with high trust and much social capital are better equipped to deal with the challenges posed by global­ ization and competitive pressures from the international economy. Therefore the convergence thesis is very unlikely to be met; on the con­ trary, a parallel evolution of societies in different cultural zones might be expected. The conclusions drawn by Inglehart and Baker (2000) using the European Values Study and World Value Surveys over a longer pe­ riod of time, are that economic development brings about change in cul­ tural values but at the same time the cultural heritage of a country leaves an imprint on values that endures despite modernization. For that reason different societies follow different paths of development even when they are confronted with the same economic forces. Cultural change appears to be path dependent. As Inglehart and Baker put it: ‘Economic devel­ opment tends to push societies in a common direction but rather than converging they seem to move on parallel trajectories shaped by their cultural heritages’. The underlying reason is that value structures also shape institutions and hence, affect social and economic interaction. Do values, as Weber argues, leave an imprint on institutions or do institu­ tions shape people’s values because people adopt themselves to the pre­ vailing institutions and consequently change their value structures in or­ der to cope with the institutions’ requirements? The empirical evidence presented by Inglehart and Baker tends to suggest that the two opposing views on whether changes in value structures are economically or cul­ turally determined are both true because values affect institutions and therewith social and economic interaction (behavior). However, the final answer to his question cannot be given due to caveats in the design and content of most value studies. To disentangle the complex relationship between values, institutions and behavior richer data are needed than the European Values Study and World Value Surveys provide. This implies that we need to enrich the values surveys with proper information on in­ stitutions and behavior and we need a longitudinal design to assess changes in values over time and to disentangle the causal chain between values and behavior more properly at the micro level. 2.2 Socio-economic research Comparative socio-economic research into the role of welfare regimes and how they change over time is largely based on national panel sur­ veys that are made comparable. The ones available to date are running

438

CHAPTER SIXTEEN

already for more than a decade2. Recently from the mid 1990s on there also exists a European Community Household Panel covering 15 Euro­ pean countries. It started in 1994 and to date a six wave dataset has been released for academic purposes. This section focuses on the classifica­ tions that have been used in these socio-economic studies to identify welfare regimes. Different welfare regimes represent different ‘worlds of welfare capi­ talism’. As such they represent different ways of organizing not only the transfer sector, represented by social welfare policy, but also the produc­ tive sector of the economy. Welfare regimes bunch particular values to­ gether with particular institutions carrying out policy programmes and concrete policy tools related to these sectors. In democratic societies the welfare policy values represent the commonly shared social values in society. Therefore in examining the role of institutions and regimes we implicitly assume that the prevailing social values are embedded in insti­ tutions and in concrete policies. Hence, values exert their impact on so­ cial outcomes through the ruling institutions and policy programmes. The researcher’s concern, however, is also moral to discover how the values traditionally associated with welfare regimes are best served (Goodin et al., 1999). In other words the object of research is to examine how and to what extent these different institutional designs of the wel­ fare state try to attain their policy goals. Researchers try to achieve an improved understanding of the interplay between the driving economic, social and political forces and how they affect the performance of Euro­ pean welfare states in terms of safeguarding a high level of economic welfare, maintaining an efficiently operating labor market and promoting social integration and social cohesion (Muffels et al., 2002). Therefore, the focus of much research is on the extent to which welfare/employment regimes or a country’s institutional set-up in terms of socio-economic policy matter in explaining the differences in performance across welfare states. Researchers make use of comparable panel surveys to examine this question. However, the evidence in these panel studies is limited and not all welfare state values are readily couched in terms of socio-economic outcomes tracked in panel-studies. Also, the measures taken often repre­ sent only imperfect or indirect indicators of the matters that concern us. The particular countries scrutinized may or may not be perfect represen­ _____________ 2

Among others these are the panels for the United States (the Michigan panel), Ger­ many (the German Socio-Economic Panel), the Netherlands (the Dutch Socio-Economic Panel) and the United Kingdom (the British Household Panel Study). Except for the Dutch panel these datasets have been made comparable by Syracuse University and re­ leased at hardly any cost to the academic community.

ONE EUROPEAN WORLD OF WELFARE OR MANY

439

tative for the type of welfare regime we examine. The particular period covered may or may not be peculiar in ways that might confound our findings. Despite these caveats and fallacies these panel studies gave us the best and indeed real evidence available for answering the sort of questions the research is occupied with. 3 A conceptual framework To examine the role of values and institutions jointly in explaining the documented diversity across countries we stated earlier that an integrated framework is needed. Such a framework might be found in Sen’s capa­ bility approach (1999) combined with the notion of the ‘inclusive soci­ ety’ (see Figure 16.1). In recent work, Sen (1999) pays tribute to the notion of ‘freedoms’ to get hold of what ‘produces’ welfare and contributes to well-being. These freedoms are people’s opportunities and capabilities to live the kind of life they have reason to value. For Sen, these freedoms are ends in them­ selves, not requiring any further justification on their instrumental ef­ fects on other outcomes such as economic growth. Sen’s focus is neither on outcomes nor on people as passive recipients of these outcomes, but rather on individuals’ acting and bringing about change, where the achievement can be judged in terms of their own values and objectives. The capabilities approach entails that social welfare policy goals have to be reformulated in terms of ‘freedom to act’ instead of ‘freedom from want’. The challenge to policies is that instead of compensating the dis­ integrated for the consequences of lack of integration, it should aim at providing the individual with capabilities in a way that allows him/her to increase his/her future prospects (Muffels et al., 2002). This implies a shift from compensating income risks to providing opportunities to peo­ ple. The aim of such policies is to enable people to achieve the function­ ing they have reason to value and to invest in their ‘capabilities’ to main­ tain or raise their human and social capital and to prevent their capabili­ ties from becoming obsolete or redundant. It is exactly for this reason why according to Esping-Andersen et al. (2002) we need a new welfare state in Europe. Instead of using the term ‘inclusive society’ he pays credit to Giddens’ notion of a social investment society which meaning however is very similar. The ultimate goal of any modern democratic society is to enhance people’s well-being. To arrive at this goal economic and social objec­ tives have to align in order to be able to promote economic welfare and social integration. The inclusive or social investment society (in Gid­ dens’ words) demands a reciprocal relationship between the three basic

440

CHAPTER SIXTEEN

components of what produces welfare in a society: productive capabili­ ties (inputs from labor, capital and technology), social capabilities (ari­ sing from norms, values, trust, institutions and social and human capital) and well-being (arising from consumption of goods and services, family ties, health, job security, community, freedoms and opportunities). There are several linkages between these components. For example, social capability influences economic performance through cultural val­ ues, trust, interpersonal skills, social networks and institutions and their impact on risk-taking, incentives, innovation and economic participation. Economic performance, in turn, being part of productive capability, af­ fects the social capability through its improvement of social cohesion by rendering opportunities for participation and learning and through in­ creased public and private spending. There are various positive and negative feedback loops between the concepts of productive and social capability, which tend to affect the outcomes in terms of well-being. So­ cial capital, in the form of strong social ties within the community where one lives, fosters trust that may stimulate voluntary exchange and also reduces transaction costs (see the contribution of Bovenberg, chapter fourteen in this book). In turn, these developments smooth broader social interactions and, hence, increase the amount of human and social capital.

Productive capability arising from inputs of labour, capital and technology

Social capability arising from norms, values, trust, institutions, social and human capital

Well-being arising from consumption of goods and ser­ vices, family ties, health, job security, commu­ nity, freedoms and opportunities Modified source: The Treasury (2001).

Figure 16.1 Capabilities and the ‘inclusive economy’

There might also be some potential trade-offs; for example, a policy that boosts economic growth might be at the expense of some form of social capability. It may raise the tensions between winners and losers. A high level of social spending seems favorable from a social viewpoint be­ cause taxes and social revenues might be used to raise collective welfare.

ONE EUROPEAN WORLD OF WELFARE OR MANY

441

At the same time, however, high levels of social spending might reduce people’s incentives that in turn, hinder economic growth. The basic idea is that the alleged trade-off between efficiency and eq­ uity is a simplification of the complex relationship between the eco­ nomic and social performance of an economy and the wide range of in­ terrelated factors that affect this relationship such as social and human capital and norms and values. Since there is a positive correlation of risk and return due to the market forces that tend to select activities with highest expected returns, the welfare state may enhance efficiency when it allows people to bear more risk. Values and institutions enter the model in several ways. First they form part of the social capability domain affecting the economic per­ formance of a society as well as affecting the overall outcomes in terms of the extent to which well-being is attained for all citizens in society. But values also shape the sort of freedoms people have reason to value for their own lives in terms of subjective well-being and quality of life considerations. These values at the individual level may also shape insti­ tutions and the amount of social and human capital that allows people to improve their current and future prospects. Hence, from the viewpoint of values there is also a reciprocal or feedback relationship between indi­ vidual well-being and the social capability of a society. 4 Results from socio-economic research The most influential study on welfare regimes is undoubtedly the study of Esping-Andersen (1990) who drew on earlier studies by Titmuss (1958, 1974), Wilensky and Lebeaux (1958), Korpi (1983) and Therborn (1986). In his socio-political account of the development of the welfare state he distinguished three types of welfare regimes, which are readily known and heavily criticized for various reasons. However, his typology is still the starting point for much comparative research in the social do­ main. The liberal welfare-capitalist state is viewed as giving priority to economic growth and efficiency. Work disincentives and welfare de­ pendency are to be avoided and welfare benefits should be targeted to the truly needy. The corporatist state also seeks strong economic per­ formance but gives priority to social stability particular household in­ come stability, and social integration. The social-democratic welfare state values all these goals but claims high priority for reducing inequal­ ity and poverty and particularly unemployment. His typology was criti­ cized by authors like Leibfried (1992), Ferrera (1996) and Bonoli (1997) for his neglect of what they called a Southern or ‘Latin-Rim’ model of the welfare state. They argued that the Southern, Mediterranean coun­

442

CHAPTER SIXTEEN

tries belong to a different welfare regime type with its familial characte­ ristics and its immature welfare system. Much of the comparative re­ search dealing with welfare regimes start from Esping-Andersen’s clas­ sification testing subsequently whether the real outcomes support the contended ones. Exemplary for this approach is the work of Goodin et al. (1999). They concentrate on what they call the ‘best cases’ of the va­ rious types of welfare regimes and try to assess their economic and wel­ fare policy performance over a decade. They selected three best country practices: the Netherlands for the social-democratic type, the United Kingdom for the liberal type and former West Germany for the cor­ poratist type. The question Goodin et al. (1999) poses is whether it is possible in practice to maintain a generous welfare state without incurring major ef­ ficiency costs. This question pertains to the ‘inclusive society’ frame­ work discussed in section 2 where we assumed that there is a complex reciprocal relationship between the productive and social capability sec­ tor of the economy and not necessarily a one-way relationship between economic and social performance. The panel analyses over a decade seem to suggest that all those familiar trade-offs are not strictly neces­ sary. The Netherlands, as ‘best case’ for the social-democratic regime and having the most generous welfare state, performs best across the three countries during this decade (1985-1994) in most instances either dealing with per capita GDP growth, reduction in inequality, prevention of persistent poverty and labor market performance (keeping unem­ ployment rates low). Only when it comes to welfare dependency it is true that the social-democratic welfare state bears more risk of capturing people in a ‘welfare trap’. It seems that there is some evidence for a small efficiency loss due to keeping people longer in welfare than strictly necessary but without endangering a high growth in economic welfare. Conversely, the liberal welfare regime of the United States not only comes bottom of the league in terms of welfare-related goals; it also does fare worse at promoting economic well-being for average citizens. Whether the situation has changed drastically after the mid 1990s might be doubted. Although economic growth rates of the U.S. and employ­ ment growth rates exceed those of many European countries the for population growth corrected figures of per capita GDP growth were again—as OECD figures show—about the same across the three welfare states. Another approach in the literature is to start from country differences and to derive empirically a classification of countries that explains nearly as much as the unique country variables. Using data from the European Community Household Panel, Muffels et al. (2002) have

ONE EUROPEAN WORLD OF WELFARE OR MANY

443

tested what type of classification would best fit to the data. They use data for 12 European countries over three years during the mid 1990s. They tried to explain the diversity in social exclusion and labor market integration. They concluded that an amended version of EspingAndersen’s typology including a Southern welfare regime type makes sense in empirical terms. They found large dissimilarities across Europe in the performance and functioning of welfare systems. The success of welfare states to deal with problems of misfortune for particular groups seem to be dependent on the way the welfare state is able to maintain a highly efficient labor market with much flexibility and at the same time guaranteeing sufficient levels of social protection and work and income security3. The regime analyses suggest that the social-democratic Scan­ dinavian regimes perform best in attaining a high level of flexibility and providing much security. The liberal regime is weak in providing in­ come security and the continental systems are weak as regards flexibil­ ity. Although the southern regime cluster is very heterogeneous the find­ ings suggest that they generally perform worse with respect to labor market flexibility, creating opportunities for upward mobility and pro­ viding work and income security. Contrary to general beliefs the evi­ dence suggest that work incentives are not highest in the liberal regime but in the corporatist and particularly the southern regimes. This should not come as a surprise though because this is exactly what creates the insider-outsider structure of their employment and welfare systems. The excluded facing multiple barriers to return to the labor market once they have dropped out, are seemingly unlikely to benefit from traditional supply side or even traditional activation policies. Flexibilization alone that is not supplemented with targeted social policies seems also not a sufficient strategy for social integration of these groups. The findings suggest that is why employment and education strategies to increase the productive capabilities of a society need to be integrated with social policies aimed at raising the social capabilities of their citizens through investments in social and human capital formation. Simultaneous treat­ ment of both is likely to increase the effectiveness and performance of welfare states in guaranteeing high levels of economic welfare and at the same time tackling social misfortune. The evidence from this European account of welfare state performance supports the view that there is no simple one-way relationship between the economic and social perform­ ance of a society. On the contrary, to believe in linear and simple rela­ tionships between the productive and social sector of the economy de­ _____________ 3

In a recent publication of Esping-Andersen (2002) he made a plea for prevention strategies over the life course in which the focus of social policy should be on income maintenance and income security particularly for families with children.

444

CHAPTER SIXTEEN

nies the complexity of current economic and social welfare systems and the role institutions play in affecting the outcomes and performance of the various policy regimes. The findings align very much with the ones reported from the Goodin et al. (1999) study. They clearly suggest that economic growth is not a sufficient strategy to eradicate social disinte­ gration. Institutions matter and they explain together with socio-cultural barriers why the disintegrated do not profit equally from the fruits of economic growth like the rest of the population. Institutions matter not only at the national level but also at the inter­ national level, in explaining the diversity in social and economic per­ formance across countries. The findings support the view that institu­ tions play a role in redistributing welfare within countries if the aim of social policy is to create a society that is fair and to the benefit of all. In doing so they do not need to jeopardize efficiency and economic growth dependent on how the welfare state institutions are designed. Across countries they show why some regimes are performing ‘better’ than oth­ ers. The general conclusion is that welfare states greatly reduce poverty and inequality and work effectively to stabilize household incomes when adversity strikes. Welfare states are therefore great feats of social engi­ neering; achievements of the twentieth century to be proud of, not be slowly whittled away (Headey et al., 2000). The welfare state institutions responsible for these achievements are designed in the course of time and reflect commonly shared moral val­ ues that subsequent policy regimes transformed in specifically designed policy programmes reflecting these shared values. Value patterns might have an enduring impact on a country’s social and economic interactions but certainly also change over time autonomously or due to economic, social and political forces. Also policy regimes change over time due to these changes in values, attitudes and social and economic conditions shaping institutions and policies. But there is little reason to expect these regimes converging substantially over time provided that the basic tenets of the various welfare systems couched in system values and moral be­ liefs remain quite different and unique for any particular welfare state. Again, values come into the picture here while they are likely to shape the institutions as they are designed and how they operate as well as through the way they affect individual behavior of the social and eco­ nomic actors involved in the social and economic interaction process. Due to lack of information in the various surveys, value research as well as socio-economic research is not capable to disentangle these complex relationships between institutions, values and behavior. This intriguing issue has to be left for further scrutiny to future research. The road to fol­ low would be either to enrich the information in value studies with in­

ONE EUROPEAN WORLD OF WELFARE OR MANY

445

formation on institutions and behavior or to combine it with the informa­ tion on institutions and behavior from other surveys like the socio­ economic panel surveys. The latter approach has to date never been adopted in research practices but it might prove a promising road to en­ rich our knowledge on the relationship between values, institutions and behavior. Up to this point we did hardly mention the impact of time on our findings with respect to the role values and institutions play in explai­ ning the documented diversity in the social domain. Real lives are not just a jumble of unconnected episodes. Rather they are temporally ex­ tended, as is the impact that welfare regimes has on people’s lives. Any serious assessment of the impact of welfare regimes needs to take ac­ count of their impact upon people’s lives over time rather than just at one instant. Unlike the repeated cross-sectional snap-shots in value re­ search, in both socio-economic studies, shortly reviewed here, use was made of panel surveys in which the same people are repeatedly asked the same sort of questions about their lives. Panel studies provide the re­ searcher with information about the timing and sequence of events the respondents have experienced during their life times. This permits to ex­ amine whether and how people adjust their values and behavior to cope with these life events. Repeated cross-sectional studies such as the Euro­ pean Values Study that are used to examine how groups of people fare over time, tend to overestimate the amount of stability that is docu­ mented with these surveys. They cannot capture the underlying dynam­ ics and processes that longitudinal data can. At the group level it appears that things have not changed much, but the results are biased because we don’t take the flows into and out of these groups into account. At the in­ dividual level things might have changed rather drastically but since we cannot follow people over time with these cross-sectional snap-shots we do not notice these changes. The conclusion in most value studies that values seem to be rather stable and do not change much might therefore be a biased outcome due to the limitations of the data. If we would have proper information from panel surveys it might have been true that at the individual level values change quite rapidly due to the occurrence of critical life events, like the loss of work, a divorce or death of a child or the partner. This would argue for a panel design of value studies but its virtue is also its disadvantage, because apart from its costs it necessarily takes quite some time before we have enough information to assess changes in values.

446

CHAPTER SIXTEEN

5 Summary and conclusions In section one of this chapter we highlighted the way the role of values and institutions has been studied in value research as well as in socio­ economic research dealing with the welfare state. In value research no systematic account at the micro-level is given of how values relate to structural socio-economic differences across societies whereas in socio­ economic research the impact of value structures is largely neglected. These omissions are partly due to lack of information in the type of sur­ veys used and must partly be attributed to the different disciplinary an­ gles from which the issues at hand are studied. To study the intriguing relationship between values, institutions and individual behavior we made a plea for an integrated theoretical framework and research design. In section two we lend credit to the work of Giddens and Sen to define a heuristic model, a conceptual framework for studying in a comparative design the relationship between values, institutions and the performance of a welfare state in safeguarding an efficient and fair society for the benefit of all citizens. This framework in analogy to Giddens’ work has been grounded on the notions of the social investment society or the in­ clusive society in conjunction with Sen’s ‘capabilities’ approach. We have argued that such a framework is needed to disentangle the afore­ mentioned complex relationship between welfare state values, institu­ tions and individual behavior. In value research hardly any evidence is provided on this relationship. There is, however, need for richer dataanalyses at the micro-level particularly because the way values affect in­ stitutions and institutions affect values is mediated by subjective consid­ erations and individual behavior. The heuristic model presumes that there is not only a reciprocal feed-back relationship between the produc­ tive and social capability sector of the economy to which values and ‘so­ cial capital’ belongs, but also between the social capability sector and the outcome sector determining individual, subjective well-being and quality of life aspects. In section three we report on the results of socio-economic research with respect to the role institutions play in determining the performance of welfare states. The available information in socio-economic panel surveys permits only to look at the role of institutions without taking the role of values as object of study in its own right. The findings of this re­ search clearly show that welfare state institutions and the way they are designed matter in explaining how people fare over time in the various welfare state regimes. The results also indicate that there is little evi­ dence for expecting a converging of the various welfare state regimes. Regimes do change over time as a result of changing values, and chang­

ONE EUROPEAN WORLD OF WELFARE OR MANY

447

ing social and political forces but the typically way these regimes bunch values and institutions appear to alter not substantially over the decade. The basic tenets of the regimes couched in system values, moral beliefs and institutions remained rather different and unique. For that reason the question put into the title whether there will be one European world of welfare need to be answered in the negative, there is no one ‘European Social Model’ but there are many and it depends on the historical, social, economic and cultural roots of the various welfare systems which insti­ tutional constellation best fits the preferences of the community at large. This is not to say that over time systems might not tend to converge. On the contrary, we can distinguish common traits of the various European worlds of welfare but their national, sub-national and even local features are too imperative to deny them by forcing the various national welfare systems into ‘one-size-fits-all’ policy. That would contradict with the large and true disparities we found comparing the different worlds of welfare. People adhere to the own welfare system, which they like to sustain because they align with their historical, social and cultural roots. The enlargement of Europe with the 10 candidate member-states, which are profoundly different from the other members, would even more stress the need to give ample room to the nation states to diverge in their socio-economic policies. Eventually, we looked at the role of time with respect to research in this field of study. We stated that repeated cross-sections as the Euro­ pean Values Study surveys are unable to capture the underlying dynamic processes at the individual level. For the same reason we suspect that the found stability in value structures might be due to the cross-sectional de­ sign of these studies overestimating the amount of stability. A panel de­ sign would likely show that there is much more variation underneath the surface than these surveys incorrectly show. A panel design would also permit to disentangle the complex relationship between values, institu­ tions and individual behavior that repeated cross-sections cannot pro­ vide. In an ideal world we would have the sort of information academics like; information that picture the real lives of people over their entire life course. However, the ideal world only exists in our mind and it is ex­ actly because of the non-existence of an ideal world that social science can prove its legitimacy. References Arts, W., & J. Gelissen 2001. Welfare States, Solidarity and Justice Principles. Does the Type Really Matter? Acta Sociologica 44: 283–300.

448

CHAPTER SIXTEEN

—— 2002. Three Worlds of Welfare Capitalism or More? A state-of-the-art report. Journal of European Social Policy 12: 137–158. Bell, D. 1973. The Coming of Post-Industrial Society. New York: Basic Books. Bonoli, G. 1997. Classifying Welfare States: A Two Dimension Approach. Journal of Social Policy 26: 351–372. Esping-Andersen, G. 1990. The Three Worlds of Welfare Capitalism. Oxford: Polity Press. —— 1999, Social Foundations of Post-Industrial Societies Oxford: Oxford University Press. —— 2002. Why We Need a New Welfare State. Oxford: Oxford University Press. Ferrera, M. 1996. The ‘Southern Model’ of Welfare in Social Europe. Journal of European So­ cial Policy 6: 17–37. Gelissen, J. 2002. Worlds of Welfare, Worlds of Consent? Public Opinion on the Welfare State. Leiden/Boston: Brill. Goodin, R., B. Headey, R. Muffels & H.J. Dirven 1999. The Real Worlds of Welfare Capita­ lism. Cambridge: Cambridge University Press. Headey, B., R.E. Goodin, R. Muffels & H.J. Dirven 2000. Is there a Trade-off Between Eco­ nomic Efficiency and a Generous Welfare State? A Comparison of Best Cases of the ‘Three Worlds of Welfare Capitalism’. Journal of Social Indicators Research 50: 115– 157. Inglehart, R. 1997. Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies. Princeton: Princeton University Press. —— & W.E. Baker 2000. Modernization, Cultural Change, and the Persistence of Traditional Values. American Sociological Review 65: 19–51. Korpi, W. 1983. The Democratic Class Struggle. London: Routledge & Kegan Paul. Leibfried, S. 1992. Towards a European Welfare State? On Integrating Poverty Regimes into the European Community. Pp. 245–279 in Z. Ferge & J. Kolberg (eds.) Social Policy in a Changing Europe. Frankfurt am Main: Campus Verlag. Muffels, R., P. Tsakloglou & D. Mayes (eds.) 2002. Social Exclusion in European Welfare States. Cheltenham: Edward Elgar. Sen, A.K. 1999. Commodities and Capabilities. Oxford: Oxford University Press. Ter Meulen, R., W. Arts & R. Muffels 2001. Solidarity in Health and Social Care in Europe. Dordrecht/Boston/London: Kluwer Academic Publishers. The Treasury 2001. Towards an Inclusive Economy. Treasury paper 01/15. Wellington: New Zealand Government, The Treasury. Therborn, G. 1986. Why Some People are More Unemployed Than Others – The Strange Para­ dox of Growth and Unemployment. London: Verso. Titmuss, R.M. 1958. The social division of welfare. London: Allen & Unwin. —— 1974. Social Policy. London: Allen and Unwin. Weber, M. (1904) 1958. The Protestant Ethic and the Spirit of Capitalism. New York: Aldine de Gruyter. Wilensky, H.L. & C. Lebeaux 1958. Industrial Society and Social Welfare. New York: Russell Sage Foundation.

CHAPTER SEVENTEEN

EUROPEAN SURROGATE EGIDIUS BERNS Für Staaten kann es nach der Vernunft keine andere Art geben, als dass sie sich zu öffentlichen Zwangsgesetzen bequemen, und so einen (freilich immer wachsenden) Völkerstaat, der zuletzt alle Völker der Erde befassen würde, bilden. Da sie dieses aber durchaus nicht wollen, so kan an die Stelle der positiven Idee einer Weltrepublik nur das negative Surrogat 1 eines Bundes aufhalten. I. Kant, Zum ewigen Frieden .

1 Introduction Leafing through the sourcebook of the European Values Study: A Third Wave (Halman, 2001), one can only conclude ‘that Europe is all but united where fundamental views, attitudes and value orientations are concerned. Of course, there are fields in which Europeans agree, but the major conclusion seems to be that, when it comes to contemporaneous orientations as well as longitudinal developments, it is a unity in diver­ sity’ (Halman & Kerkhofs, 2002: 36). This diversity is expressed in the small number of Europeans who feel especially attached to Europe and in the restricted confidence the European Union is able to raise. Never­ theless, one has to take into account that Europeans feel attracted in the first place to their local and regional surroundings, and that the slight at­ tachment to Europe does not indicate a strong national involvement. On top of this, the European Union has its share in the aversion that over­ takes not only many institutions like parliament, civil service, unions, press or Nato, but also politics in general. However, these kinds of rela­ tivizations of the resistance against the political unification of Europe only add to the fundamental character of the European diversity and fragmentation. At the same time, the Europeans seem convinced that the problems they would like to be solved, for example, pollution, crime or _____________ 1

Kant (1982: 212-213). All quotations from Zum ewigen Frieden are taken from V. Ferraro’s translation which I sometimes slightly adapted; the page numbers in my text re­ fer to the original German text of the Werkausgabe.

450

CHAPTER SEVENTEEN

unemployment, do not call for a national approach but require interna­ tional concerted action2. Apparently there is a clash between, on the one hand, the more or less emotional attachments which are expressed by values and on which the EVS is focused, and, on the other hand, a realis­ tic understanding of things. Something analogous is to be found in the contradiction between the poignant lack of interest in politics and the overwhelming support for having a democratic political system. It makes one wonder whether the European does not suffer from silliness or be­ haves like a spoilt child. This crumbled Europe has been paving its way towards unity for the last fifty years. Is it possible to give a philosophical explanation of this process? Various and all but insignificant philosophers have made an at­ tempt at it: Jürgen Habermas (1999)3, Jean-Marie Ferry (2000) and Jacques Derrida (1990)4. In this contribution I put the question whether Kant’s 1795 treatise Zum ewigen Frieden is capable of helping us through. The fact is, peace is often looked upon as the key issue in the process of European unification. This is clear, for example, from Robert Schumann’s appeal, on May 9, 1950, to arrive at a ‘European Coal and Steel Community’, which is commonly considered to be the startingpoint of this process. In this appeal, which was written by Jean Monnet, various elements can be found—such as federalism as the result of a process of development and the priority of the economy in it—that we will also encounter in Kant. Most of all, however, Schumann’s Declara­ tion points to the aim of this process, namely peace: ‘The pooling of coal and steel production should immediately provide for the setting up of common foundations for economic development as a first step in the federation of Europe, and will change the destinies of those regions which have long been devoted to the manufacture of munitions of war, of which they have been the most constant victims. The solidarity in production thus established will make it plain that any war between France and Germany becomes not merely unthinkable, but materially impossible. The setting up of this powerful productive unit, open to all countries willing to take part and bound ultimately to provide all the member countries with the basic elements of industrial production on the _____________ 2

Halman (2001: 200-202). Unfortunately these data comprise only some ten-odd countries. It would moreover have been much more interesting to question cooperation/concerted action not only within the framework of ‘international organizations’ but also within the scope of the European Union. 3 This compilation of studies comprises an evaluative analysis of Kant’s treatise on perpetual peace: ‘Kants Idee des ewigen Friedens – aus dem historischen Abstand von 200 Jahren’. 4 In another work on Levinas (Derrida, 1997), he elaborates on Kant’s treatise on per­ petual peace.

EUROPEAN SURROGATE

451

same terms, will lay a true foundation for their economic unification. This production will be offered to the world as a whole without distinc­ tion or exception, with the aim of contributing to raising living standards and to promoting peaceful achievements’ (Salmon & Nicoll, 1997: 44– 45). The use of Kant’s text for the clarification of the nature of European unity does, however, require that, when reading the treatise, we should not take rash decisions whenever Kant threatens to get stuck in a split, but rather to use this as indications for a better understanding of the status of the political unification of Europe. Perhaps this can clarify the curious and never before exhibited mixture we call European Union – an alliance of states, which maintain their own sovereignty but share it with each other in the Union and also hand over part of it to institutions of the Union, which thus acquires traits of sovereignty without in its turn be­ coming a state. In his essay, Kant offers a kind of digest of the most important themes of his political philosophy and philosophy of law while exploring the idea of a cosmopolitan law. This constellation already indicates that in his view cosmopolitanism and the issue of international peace are not a secondary appendix of his political philosophy but indeed its ultimate goal. It is often said that Zum ewigen Frieden lies at the basis of the United Nations. This is certainly true insofar as Kant asserts—as we shall see—that peace can only reign on a world scale. By no means does he reject regional bonds—on the contrary, they constitute the historical way to world peace—but they can never become an end in themselves. The picture of the cosmopolitan union Kant draws, however, reminds one more of the European Union. The reason for this is that the aporiae which can be observed in Kant’s writing between, on the one hand, the plurality of sovereign peoples and a world republic on the other only al­ low for surrogate solutions, which bear a strong resemblance to the im­ age the daily routine in Brussels offers. In this chapter, I shall stress the fact that this surrogate is in fact desirable and that the daily routine in Brussels is not so bad after all. 2 Analogy As was often the case in the rich tradition of writings on peace in the seventeenth and eighteenth centuries Kant’s treatise is structured like a peace treaty, including preliminary articles, definitive articles, appendi­ ces and secret clauses. When things get serious, in the beginning of the section in which the three definitive articles for perpetual peace are dealt with, he opens as follows: ‘The state of peace among men living side by

452

CHAPTER SEVENTEEN

side is not the natural state (status naturalis); the natural state is one of war. This does not always mean open hostilities, but at least an unceas­ ing threat of war. A state of peace, therefore, must be established, for in order to be secured against hostility it is not sufficient that hostilities simply be not committed; and, unless this security is pledged to each by his neighbor (a thing that can occur only in a civil state), each may treat his neighbor, from whom he demands this security, as an enemy’ (203). The way in which Kant opposes a number of notions in this passage will structure his exposé to a large extent (Derrida, 1997: 154): the state of war is the natural state and peace must be established (gestiftet). On the basis of this, war and peace have nothing in common. War comes first and concerns man as a natural being, that is to say as a being which is determined by inclinations like self-preservation and preservation of the species and by social aspirations in which man finds his self-respect through the opinion of others. Often enough, Kant claims that this natu­ ral human being is made of crooked wood, which means that he is not oriented towards the heaven of his moral destination but bent towards his particular self. Peace, on the other hand, comes afterwards and is not naturally determined but established. This makes peace a political and a legal concept, that is to say a concept which has got to refer to another determining ground, different from purely natural grounds. Of course Kant is not blind to the fact that wars can be ended and peace made on the basis of natural determinations like self-preservation or social com­ munication. But these kinds of truces leave the threat of war untouched. For peace to be something quite distinguished from war, it must above all make an appeal to an Anlage, a human disposition, which is of a to­ tally different character than man’s natural dispositions. That is the case when man is no longer determined, but self-determining. That is why Kant says that peace has to be established. This way of cataloguing notions enables Kant to deal with the pro­ blem of establishing peace between states by analogy with that of esta­ blishing peace between human beings when they renounce the natural state in the so-called social or original contract and found a state. For a long time the theorem of the social contract played a crucial role in po­ litical philosophy. It is Rawls’ enormous merit to have actualized it for the present time. In such a contract, individuals (fictitiously) say to each other: ‘There ought to be no war among us, for we want to make our­ selves into a state; that is, we want to establish a supreme legislative, ex­ ecutive, and judiciary power which will settle our quarrels peaceably’ (212). In the natural state, that is to say in the state in which human be­ ings allow themselves to be determined by their natural inclinations, people are quarrelsome because they are mainly concerned with them­

EUROPEAN SURROGATE

453

selves. Only when living in this manner becomes intolerable and mutual fine-tuning inevitable, is there a change ‘through which the multitude become one people’ (206). In doing so, the multitude leave their illegal freedom behind and submit as a people to a system of law. Analogous to this process in which individuals give themselves a state, Kant develops an argumentation in the defense of a cosmopolitan order. After all, the multitude only submit to the law within the context of a people giving itself a state. But there are many of these states, and they in their turn have no system of law above them, since the system of law only applies within a state. That is why in international law they are called ‘sovereign’, and that is why this ‘right’5 is founded in the princi­ ple of non-interference in the internal affairs of an other state. At the same time, however, these states cannot guarantee the right for which they were created (Mertens, 1990: 12). For the other states cannot be but a threat, since states are not submitted to a system of law. As a matter of fact, between states exactly the same natural state rules which existed between individuals before the social contract was signed. ‘War, how­ ever, is only the sad recourse in the state of nature (where there is no tri­ bunal which could judge with the force of law) by which each state as­ serts its right by violence’ (200). In terms of analogy it sounds like this: ‘Peoples, as states, like individuals, may be judged to injure one another merely by their coexistence in the state of nature (i.e. while independent of external laws)’ (208–209). And Kant concludes that, just as in the so­ cial contract, lawless freedom is given up in favor of reasonable freedom within a system of law, ‘for states in their relation to each other, there cannot be any reasonable way out of the lawless condition which entails only war, except that they, like individual men, should give up their sav­ age (lawless) freedom, adjust themselves to the constraints of public law, and thus establish a continuously growing state consisting of various na­ tions (Völkerstaat) (civitas gentium), which will ultimately include all the nations of the world’. A few lines further, Kant calls this state of na­ tions ‘a world republic’ (212–213). Entry into this continuously growing state of nations requires the candidates to embrace a republican constitution. Although republicanism is inconsistent with an absolute kind of monarchy, it by no means rejects (a constitutional) monarchy, contrary to what is often claimed in the Netherlands6. As a matter of fact, republicanism does not pronounce upon who rules, but on how this is done. The principles of the republican _____________ 5

I place this right between brackets because ‘the concept of a law of nations as a right to make war does not really mean anything’ (212). 6 Nor has it much to do with what is called national republicanism in France, or with what the ‘republicans’ represent in the United States.

454

CHAPTER SEVENTEEN

constitution are liberty and equality on the one hand and the separation of powers and representation (instead of for instance direct democracy) on the other. Kant defines this juridical freedom to which each human being is entitled as a human being as ‘the privilege (Befugnis) to lend obedience to no external laws except those to which I could have given consent’ (204n). Besides, in this legal context equality means that man as a citizen is subjected to the same legislation as all other citizens, a legislation to the realization of which he had the opportunity to contri­ bute as a citoyen in exactly the same way as all other citizens (Mertens, 1990: 73ff). These principles hold a priori, that is to say, they ensue from the concept of law itself. This means that they are unassailable but abstract as well, and thus leave room for experience. If I am allowed at this stage to draw a conclusion regarding the Euro­ pean political unification, I would like to point out that this unification from such a Kantian perspective can never be an end in itself. Kant would indeed applaud the European Union as a step in a historical evolu­ tion. There is every indication for it in his reference to a ‘continuously growing Völkerstaat’ and elsewhere (211) he does not outline the por­ trait of a General Assembly of the United Nations, which by definition comprises all sovereign states of the world, but a historical crystalliza­ tion process in which some like-minded republican states move on to­ gether in order to win over an increasing number of states in favor of a cosmopolitan world order. Nevertheless, the exclusive goal and justifica­ tion of this process is the said world republic. The restriction, on princi­ ple, to Europe in a geographical sense—and thus the exclusion of Turkey—has no moral significance and cannot be deduced rationally from the concept of law. The sole criterion for deciding on a possible entry into a cosmopolitan order is an internal republican constitution. Empiri­ cal data do play an important role for Kant, and expansion can be tempo­ rarily discouraged on the basis of prudential considerations. This slow­ down ‘does not permit us to fail to do it, but allows a delay to prevent precipitation which might injure the goal striven for’ (201). 3 Gap and churchyard Kant’s analogous argumentation is founded on the unbridgeable gap be­ tween natural war and established peace which touches upon the very heart of his philosophy. Just how radically Kant disconnects war and peace appears from his definition of peace with which he opens his trea­ tise: ‘No treaty of peace shall be held valid in which there is tacitly re­ served matter for a future war. Otherwise a treaty would be only a truce, a suspension of hostilities but not peace, which means the end of all hos­

EUROPEAN SURROGATE

455

tilities – so much so that even to attach the word ‘perpetual’ to it is a du­ bious pleonasm’ (196). Genuine peace, therefore, can only be perpetual peace; peace, that is, from which no new war can ever evolve. The fun­ damental point at stake here is the irreducibility of the moral law or categorical imperative. Every contextualization of this law with which­ ever content, be it natural or historical, affects its categorical character and must therefore be barred, at least if we are to assume that there exists a morality in which the unconditional is advanced. Such a law cannot but be purely formal. It merely presupposes man’s capacity to act by fol­ lowing general rules, independent of the ends he aims at. For if man were to let himself be guided by ends, his acts would be determined by them. The Anlage, which I already mentioned in the beginning of the previous section, and which must be presupposed in case of morality, can only refer to a capability to define oneself as a rational being guided by general laws. That is why Kant sees this as freedom. Today one can find this rigid interpretation of morality in authors like Levinas and Der­ rida. Often enough they are labeled as ‘postmodern’. That postmodern­ ism can be identified with moral relativism, as is often claimed, is there­ fore a rash consideration. To me the problem is rather the other way around: How can man carry on, now that, for what compels him, he is only dependent on himself and no longer on a church, state or commu­ nity. Relativism can only be the case here in a derived sense: as a flight or diversion from a prior obligation. Of course, Kant’s rigorous formalism immediately raises the question of what the relation can still be then between morality and factual, more especially political action. Zum ewigen Frieden tries to answer that ques­ tion. Or rather: it tries to show that the radical absence of such a relation does not yet mean that all that remains is despair and cynicism. On the contrary, there are many good reasons to assume that the factual histori­ cal development, independent of the moral motives people can have to promote this development, evolves in the direction of more peace. ‘What has nature done to favour man’s moral purpose, and how has she guaran­ teed (by compulsion but without prejudice to his freedom) that he shall do that which he ought to but does not under the laws of freedom?’ (223). One can find in Kant three grounds to found such a development (Habermas, 1999: 199–208). In the first place republican constitutions are claimed to make states peaceable. ‘The reason is this: if the consent of the citizens is required in order to decide that war should be declared (and in this constitution it cannot but be the case), nothing is more natu­ ral than that they would be very cautious in commencing such a poor game, decreeing for themselves all the calamities of war’ (204–205).

456

CHAPTER SEVENTEEN

Nevertheless, Habermas rightly points out that in the twentieth century the most total wars were fought with the help of the people’s consent. Therefore, the people and its consent are not unambiguous categories. Secondly, Kant thinks that ‘the spirit of commerce, which is incom­ patible with war, sooner or later gains the upper hand in every state. As the power of money is perhaps the most dependable of all powers (means) included under the state power, states see themselves forced, without any moral urge, to promote honourable peace and by mediation to prevent war wherever it threatens to break out’ (226). Perhaps this proposition makes light of the contradictions which are generated by the economic system behind the spirit of commerce. Nevertheless, as we will see later on, Kant’s cosmopolitanism is at right angles to any form of colonialism and imperialism. The proposition also confirms the wis­ dom of Monnet’s policy to let the economy be the driving force of the unification of Europe. If this is true, globalization is a process that has caught up with Europe and outrun it, and in which the European Union has its place only insofar as it contributes to the goal of world peace in a world republic. This specifically means that Europe guarantees that re­ publican freedom and equality remain the regulating principles of law. Finally, Kant has great expectations of the free judgment of reason which the intellectual specifically will uphold (228). This will generate a public cosmopolitan consciousness in which ‘a violation of rights in one place of the earth is felt throughout the world’ (216) (Ferry, 2000). In this respect, too, Kant would be disappointed by the frequent trahison des clercs of the past two centuries. He especially could not foresee the major changes in the field of propaganda, communication and the mass media, and the possibilities of manipulation that go hand in hand with them, just as he overestimated the peacelovingness of the people. This does not, however, alter the fact that since the Vietnam war we do ex­ perience the awakening of a global consciousness. The current antiglobalization movement also points in this direction. The gap Kant postulates between morality and historical facticity or, if one wishes, between the transcendence of the law and the imperfect­ ness of its inscription (Evink, 2002) has far-reaching consequences. The heart of the matter lies in the insight that the gap constitutes the condi­ tion for morality. The moral law can hold unconditionally precisely be­ cause it withdraws from historicity. On the other hand, this leaves ample room for historical development, which in Kant’s view even implies progress. In this development, the shift from natural state to a state gov­ erned by law cannot but be violent. For naturaliter man is made of crooked wood. Only the stern hand of a ruler (231) or emergency is able to keep this in check. The social contract should therefore not be under­

EUROPEAN SURROGATE

457

stood as the origin of the state but as its end (Mertens, 1990: 47–56); it formulates the standard that the citizen has to approve of the order to which he must submit. This ‘republican’ condition is a priori included in the concept of law. It does not mean, however, that it is ‘established’ overnight. For Kant, it is the ‘most difficult’ (223) task for mankind, precisely because man is so crooked. It is the most time-consuming es­ tablishment there is, although, according to Kant, there is no reason for despair or cynicism. Hence the decisive, self-contained and to Kant un­ assailable position of law between the violent genesis of a political con­ tract and morality: it imposes on people made of crooked wood external, enforceable rules, the formal generality of which they moraliter ought to want spontaneously, but which they, because of their crookedness, in fact often do not want, so that only an enforceable law can bring relief. This ‘natural’ priority of violence due to the crookedness of mankind also explains why republicanism offers the only satisfactory solution for society. In every system of ruling in which individuals wield power, the demand for generality is under fire, ‘since the possession of power inevi­ tably corrupts the untrammeled judgment of reason’ (228). That is why law must rule alone. Crooked as man is, however, no human is able to establish this. Only history can produce this, given the three develop­ ments I have outlined before. This gap contradicts every dialectical relation between history (as the unfolding of natural processes) and morality in general and between war and peace in particular. In Kant’s view, peace is not pregnant with the next war, nor is war the necessary porch to peace. The latter interpreta­ tion of peace, often attributed to Hegel, would permit, according to Kant ‘perpetual peace only in the vast burial ground of the human race’ (200). ‘Ordinary’ peace, then, is a truce and political realism is the assertion of the permanent possibility of war and the ridiculing of the unconditiona­ lity of the law7. 4 Surrogate Kant’s view of the gap between morality and nature, however, is not un­ problematic. In very general terms, this already emerges from the aporetic tension between morality as self-determination on the one hand and the necessity to postulate an Anlage present within human nature on the other hand. Notwithstanding the gap, nature, therefore, returns in morality. Such motions are active in Zum ewigen Frieden as well. _____________ 7

I evidently refer here to the very Kantian phrases with which Levinas opens the Préface of his Totalité et infini (1961).

458

CHAPTER SEVENTEEN

The world republic Kant’s analogous argumentation resulted in, leads to the stipulation in the third and last article of his treatise that ‘the law of world citizenship shall be limited to conditions of universal hospital­ ity (…) Hospitality means the right of a stranger not to be treated as an enemy when he arrives in the land of another. (…) It is not the right to be a permanent visitor (Gastrecht) that one may demand, it is only a right of temporary sojourn (Besuchsrecht)’ (213). This right forbids the inhabitants of a country to treat strangers in advance with hostility as well as strangers to appropriate the land of others. Strangely enough, Kant founds this most perfected and pure form of right in nature: ‘A right of temporary sojourn, a right to associate, which all man have by virtue of their common possession of the surface of the earth, where, as a globe, they cannot infinitely disperse and hence must finally tolerate the presence of each other’ (214). Most strikingly, however, the acme of law appears to be very restrictive: it is restricted to hospitality, and this is in its turn restricted to a right of temporary sojourn. Thanks to these restrictions, appropriation, hence colonialism and imperialism are forbidden, but at the same time it does not exactly open up the world to cosmopolitans. Derrida sees in this a return of nature, which asserts its violence even within this law (Derrida, 1997: 156). The return of nature, however, goes much further than this. In Kant, man as a human being is the source of any law. We have already seen that the republican principle of freedom is man’s due as a human being. And, in the same way, we have seen that all people enjoy the cosmopoli­ tan right of temporary sojourn. This line of argumentation brings Kant to his world republic. In the above restrictions, however, the sovereignty of states in international law is taken into consideration; the very sover­ eignty, that is, to which Kant ascribed the violence he so much wanted to overcome. This return of nature dominates the second definitive article, in which Kant arrives at a world republic but at the same time argues in favor of a confederation. Much has been written on this matter. The commentators (Mertens, 1990: 99–106) have always considered it their task to come to a decision where Kant apparently had got stuck in a split. Some put down the am­ biguity to a development in Kant’s thought. Others, like Mertens, argue in favor of a confederation. Habermas concludes without much ado that Kant is ‘contradictory’ and ‘inconsistent’, interprets Kant’s expression of ‘permanent congress’ as a summary of this, because according to Kant a congress is terminable and termination is per definition excluded because of the permanence of the congress, and decides along the lines of a world republic for an ‘executive, which if necessary could ensure the observance of the Declaration of Human Rights by interfering in the

EUROPEAN SURROGATE

459

sovereignty of national states’ (Habermas, 1999: 212). To me, however, it seems much more interesting and truthful to hold on to the undecida­ bility of Kant’s split. Conceptually this preserves the gap, hence the claim to morality; empirically, it sheds much more light on the process of European unification in which federal and confederate, supranational and intergovernmental tendencies also vie for priority. Kant’s account in support of his second article is extremely tortuous. On the one hand, ‘according to reason (Vernunft) there can be only one way’, which is ‘in thesi right’ and ‘which reason must of necessity con­ nect with the notion of the law of nations’, namely the foundation of a ‘state of nations’, ‘world republic’ or ‘social alliance between citizens, viz. the free federalism’ of cosmopolitans. On the other hand, he always finds good reasons to say that it must be a ‘federation8 of free states’ which ‘would be a league of nations (Völkerbund), but it would not have to be a state consisting of nations (Völkerstaat)’ since ‘what is right in thesi, (is) in hypothesi (to be) rejected’. There are indeed good reasons, which are moreover very relevant to what is going on in Europe nowa­ days. They are twofold, depending on whether one starts from one side of the gap or the other to retrieve oneself on the opposite side. On the side of history and nature Kant often points out that a global state would be very undesirable from the point of view of morality, be­ cause antagonism in general and between states in particular is a driving force for humanity and unintentionally furthers the cause of peace. Even war, notwithstanding all its horrors and meanness, is an instigation for individuals or peoples because it incites discipline and raises man above his egoism and a community above its conflicts of interest. Furthermore, in a state of nations, ‘many nations in one state would then constitute only one nation’ (209), which is a contradiction. Such a merger—which pictures in a nutshell the problematic issue of integration versus multiculturalism, of globalization and the European extension— violates plurality. This plurality, then, looks for another way out, pro­ voking tensions which in their turn can only be checked by a despotic way of ruling. Besides to the fact that this is incompatible with the re­ publican goal of the world republic, this despotism leads to civil war and hence to the disintegration of the republic. Unity thus relapses into plu­ rality of states, hence into threat of war, which was exactly what we set out to overcome. This makes Kant extremely critical of what he calls a ‘universal monarchy’ which flows from the tendency of a state to con­ tinuously conquer more and more states for its own safety. ‘As laws al­ ways lose in vigour what government gains in extent, hence a soulless _____________ 8

In this context this notion refers to what is called a ‘confederation’ in the context of the European debate.

460

CHAPTER SEVENTEEN

despotism falls into anarchy’ (225). Is there a better and more convin­ cing plea for the principle of subsidiarity? Finally, Mertens points to the ineffectiveness of a world republic gi­ ven the end, namely peace, since apart from the fact that such a republic leads to anarchy, it is even redundant if it demands explicitly that the participating countries be organized as republics. ‘After all, the will to deal peacefully with other states is already guaranteed by the republican constitution, so it does not have to be enforced (by a world republic). A coordinating, international organization is enough to support the already existing will to settle conflicts peacefully’ (Mertens, 1990: 103–104). It is precisely on this point that Habermas, unhindered by any considera­ tion whatsoever for the plurality and sovereignty of the states, comes to the opposite conclusion: the Kantian project is inconsistent precisely be­ cause Kant recoiled before the supra-state institutionalization of the cos­ mopolitan law, without which nothing will come of this right due to the sovereignty of the states. On the other hand, morality returns in the historical evolution as well. As it is, Kant’s analogous argumentation falls short. To a certain extent, he himself formulates the objection: ‘The obligation which men in a lawless condition have under the natural law, and which requires them to abandon the state of nature, does not quite apply to states under the law of nations, for as states they already have an internal juridical constitu­ tion and have thus outgrown compulsion from others to submit to a more extended lawful constitution according to their ideas of right’ (211). The shift to a cosmopolitan law does not take place from an absolute absence of law but from communities that already live together under an existing legal order. The acceptance of the cosmopolitan law does not have the same degree of necessity as that which led to the national state governed by law. On the contrary, exactly insofar as the cosmopolitan law leads to a world republic and the national state governed by law must be given up, the question arises whether in a moral sense this does not signify a setback. It can therefore easily be understood that, also from a moral point of view, European citizens oppose a European regulation, because to them it could signify the destruction of their national system of law. It seems, then, that the gap cannot be maintained. Are the Real­ politiker and their philosophical yes-men right after all and does perpet­ ual peace exist only in the churchyard where morality is nature and na­ ture morality? Kant, however, resists: ‘Reason, from its throne of su­ preme moral legislating authority, nevertheless absolutely condemns war as a legal recourse and makes a state of peace a direct duty’ (211). He cannot choose, or rather he is not allowed to choose. The gap must be confirmed but at the same time it is a fact that one side is in perpetual

EUROPEAN SURROGATE

461

contact with the other, which necessitates permanent steersmanship. Hence his conclusion that ‘if all is not to be lost, there can be, then, in place of the positive idea of a world republic, only the negative surrogate of an ever growing alliance which averts war, endures, spreads and holds back the stream of those hostilities which fears the law, though such an alliance is in constant peril of their breaking loose again’ (213). Haber­ mas says that the surrogate solutions are no good and that one has to ac­ cept the ‘real’ solution: the world republic or mutatis mutandis a supra­ national Europe. Mertens says that the surrogate of the confederation is not an Erzatz9 but a genuine solution in itself. However, it is in the inter­ est of Europe and the world to keep the double blind undecidable and to be satisfied with what is ex post always only a surrogate, a compromise. If Europe has one merit, it is precisely that of having brought up the idea of perpetual peace. It should therefore not aim at becoming a ‘genuine’ state. It has long since passed this stage with all its national states. It should not re-open this matter, neither in the sense of the foundation of a European state after the image of the national states, only bigger, nor in the sense of the dissolution of those states. It must go on, it must learn to share sovereignty in a political order that demands perpetual peace and does not fear surrogate. References Salmon, T. & W. Nicoll (eds.) 1997. Building European Union: A Documentary History and Analysis. Manchester: Manchester University Press. Derrida, J. 1990. L’autre cap. Paris: Minuit. —— 1997. Adieu à Emmanuel Levinas, Paris: Galilée.

Evink, E. 2002. Transcendentie en Inscriptie. Delft: Eburon.

Ferry, J.M. 2000. La question de l’état européen. Paris: Gallimard.

Habermas, J. 1999. Die Einbeziehung des Anderen. Studien zur politischen Theorie. Frankfurt

am Main: Suhrkamp. Halman, L. 2001. European Values Study: A Third Wave. Tilburg: EVS, WORC, Tilburg Uni­ versity. —— & J. Kerkhofs 2002. Het Europese waardenonderzoek: enkele resultaten. Pp. 11–54 In H. van Veghel (ed.), Waarden onder de meetlat. Het europese waardenonderzoek in discus­ sie. Budel: Damon. _____________ 9

In the French Pleiade edition of Kant’s Oeuvres philosphiques, Paris: Gallimard (1986: 349) the German ‘Surrogat’ is translated as ‘supplément’. What is meant here is indeed what Derrida calls ‘supplément’: an addition to something which it substitutes, while at the same time this ‘something’ cannot do without this addition.

462

CHAPTER SEVENTEEN

Kant, I. (1795) 1982. Zum ewigen Frieden. Werkausgabe. Band XI. Frankfurt am Main: Suhrkamp. Levinas, E. 1961. Totalité et infini. La Haye: Nijhoff. Mertens, T. 1990. Kritische filosofie en politiek. Immanuel Kant over oorlog en vrede. Nij­ megen: Gerard Noodt Instituut.

CHAPTER EIGHTEEN

HE WHO TEARS DOWN ALSO BUILDS UP. LITERATURE AND THE DEBATE ON NORMS AND VALUES JAAP GOEDEGEBUURE Sociologists judge society by counting and measuring. Writers do the same, but in a different way. They don’t count and measure, but they at­ tempt to capture some of the social dynamics in the mirror of their imagination. And those testimonies are frequently the ones that stick the longest. Does not Renard the Fox tell us much about medieval classsociety? And didn’t marxist literary theorist Lukács owe a debt of grati­ tude to Balzac for a huge amount of knowledge about emerging capita­ lism? Who has captured more accurately the horrors of World War I than Ernst Jünger? The work of the best authors is not only of great aes­ thetic value, it also serves as a seismograph for the shocks and surges that occur over the course of the historical process, especially when norms and values are under debate. There was a time in Europe, approximately between 1750 and 1900, during which literature especially was credited with a formative and en­ nobling power. In the works of the great writers (Homer, Dante, Milton and so on) the Truth, Beauty and Goodness were merged in an alterna­ tive Trinity. When Enlightenment also brought the unstoppable emer­ gence of secularization, and the authority of institutionalized Christianity gradually crumbled away, numerous writers and critics started to believe that reading the classics automatically made you a better person, not merely because learning and scholarship would improve, but especially because the morality of soul and mind were thought to gain substan­ tially. The humanities owe their central position in nineteenth century Bildung to this conviction. Although Bildung, like the Enlightenment it sprang from, was closely linked with, if not supported by the liberal middle classes, later it was firmly embraced by the social-democratic movements which emerged from halfway through the nineteenth century onwards and which ended up in parliament in many a European country, and eventually also in office after 1900. In the Netherlands, for example, the Bildung ideal adapted by the socialists was realized in ‘de Maat­

464

CHAPTER EIGHTEEN

schappij voor Goede en Goedkope Lectuur’ (the Society for Good and Inexpensive Literature), also known as ‘De Wereldbiblioteek’ (the World Library) (established in 1905), and in the post-war cultural policy of renowned minister G. van der Leeuw. It is not my intention to turn this essay into an historical exposition on the rise and fall of Bildung. The observation that the middle-class liberal ideal of education has died a none too gentle death by suffocation in the age of digitization, computerization, mediatization, and last but not least the direct benefit principle, will suffice. This is also true of the sacro­ sanct status of Truth, Beauty and Goodness. These concepts are and were as fleeting as the Platonic world of Ideas in which they had always been integrated. In the materiality of literature itself, i.e. the written and printed texts, Truth, Beauty and Goodness were not always given a great deal of prominence. In literature, first and foremost in the literature of the great writers, murder, mayhem, and adultery are the order of the day. If any moral norms are postulated at all, then not in reference to living exemplary persons or situations which embody and propagate these norms, but predominantly in the form of an ideal that is not realized in ‘this, the best of all worlds’. Particularly the texts that were thought of most highly by the enligh­ tened civilizers, Homer’s Iliad and Odyssey, the tragedies by Aeschylus, Sophocles and Euripides, Dante’s Inferno and Shakespeare’s tragedies, are swarming with failing heroes, double-dealing, human weakness eve­ rywhere. In a sense Plato therefore saw more clearer than Herder, Les­ sing, Schiller, Matthew Arnold and all the other optimists: there is no room for the poets in the utopian state, as the bad examples they set for the young make it impossible to discern anything but potential moral corruption in their work. If there is a particular relationship between literature and norms and values, then especially in the capacity of Goethe’s Mephistopheles, ‘der Geist der stets verneint’. Particularly in the centuries following the civi­ lizing, educating and disciplining offensive of the Enlightenment, writers have revealed themselves first and foremost as culture critics, who, not by accident, joined forces against the belief in progress and the deeprooted faith in judiciousness and benefit that were cherished by the libe­ ral bourgeoisie as central dogmas. They do not display any optimism; on the contrary, they are convinced of the indestructibility of the evil which in their view is closely linked with the human mind. In that sense the work of the Marquis de Sade is paradigmatic for the way in which wri­ ters become the counterparts of the optimistic educators. Under the guise of issuing a licence for total amoralism and allowing all actions, regard­ less of how perverted or cruel, Sade actually places a heavy burden on

HE WHO TEARS DOWN ALSO BUILDS UP

465

the conscience. He violates Kant’s categorical imperative (loosely trans­ lated as ‘do unto others as you would have them do unto you’), thereby heavy-handedly opening his readers’ eyes to their most secret desires and deepest urges, and in this way provokes a moral debate, which is subsequently continued by Baudelaire, Dostojevski, Nietzsche, Céline, Pasolini and W.F. Hermans. In our time contrary writers like Frans Kel­ lendonk, Brett Easton Ellis and Michel Houellebecq have continued on this disruptive track. Their moral solicitude can be summarized in the adage that tender anarchist Louis Paul Boon once formulated as his motto: ‘He who tears down also builds up.’ The first thing that stands out in the works of the abovementioned au­ thors, is not so much their affinity with the tragic dimension of exis­ tence, but rather their fascination with evil. It is both the driving force behind their creativity and the inexhaustible reservoir they draw on for their stories. They know that beneath the film of reason and decency the emotions and instincts seethe and boil like the lava in a volcano about to erupt. For that reason some moralists talk about ‘the animal in man’, a designation that, all things considered, implies a more serious violation of animal rights than the torment of vivisection or bio-industry. If ani­ mals were subject to the urges that hold the human species in a kind of permanent stranglehold, we would have no objection to the use of terms like ‘perverted’ or ‘degenerate’. For exactly this reason Michel Houelle­ becq emphasizes in his novel Les particules elementaires—which is set partly in the future—that the hedonistic, individualistic type of person that has started to dominate the Western world since the second half of the twentieth century, to the detriment of his own sort and the entire planet, has burned himself up and will be replaced by a clone that has been freed of evil urges. Before he reveals this utopian perspective, Houellebecq has disgusted us extensively with the excrescences of indi­ vidualism: the separation of love and sex, and the succession of sex and unrestrained violence (pleasures that are no longer, as in Sade’s day, the privilege of the wealthy and the prominent, but have come, democratized and all, within the reach of everybody who has a video and can afford a trip to the sex club or a flight to Thailand every now and then). Whereas Houellebecq, following Aldous Huxley (Brave new world), uses the genre of the utopian, science fiction-like novel to enforce his culture critique, Swedish author Torgny Lindgren goes the exact oppo­ site way in his novel The light. He presents his view in the shape of a semi-historicizing, semi-mythological parable, which is said to be set in a diffuse medieval past. The light is set in the indefinable era of ‘once upon a time’. Location is an area in the northern part of Sweden, on the edge of the world. All is

466

CHAPTER EIGHTEEN

well in the village of Kadis when an epidemic breaks out one not so fine day. A man, who vainly went in search of a wife, brings back a rabbit with young, so he does not have to return totally empty-handed. The animal has been impregnated by a giant hare, and in addition jumped by a flea that is infected with the plague bacterium. After the disease has run its course, there are only seven survivors left in the village. They are left with unanswerable questions. Who are to be heirs to the estates of those who died? Who is supposed to ensure justice now that there is only one person in authority left? Who is to read mass, perform marriages, make sure the dead are buried? The situation in Kadis most closely resembles the appearance of a forest after a big fire. The new shoots and vines push their way straight through the charred tree-stumps, but the order they conform to is a dif­ ferent one than before. The need for rituals and rules among the illiter­ ate, but deeply religious survivors is such that they create new ones as they search and feel their way. But the old conventions can still be iden­ tified in the alternative mythology, religion and ethics, just like a forest remains a forest even after all the trees and plants have been replaced. A wonderful example of the (re)construction of norms and values is found in the latter days of the local priest, who divulged his name just before he died: he is called Blasius. When a monstrously large boar is born some time later, the name comes into fashion again. Blasius the boar has all the qualities of a mythical animal and is at the centre of a cult that celebrates intemperance and anarchy. It continues until this un­ stable balance is disturbed by an event that cannot be tolerated, even un­ der the new morality. The second Blasius, who regularly pulverizes a rabbit between his jaws, breaks out and crushes a child. The people have no choice but to make him pay, and so he is hanged following a real trial. And there is more. When the need to beautify the church with holy figures re-emerges, the devil is the first in line. His face has features in which every observer recognizes somebody. The existence of God is taken for granted, but He is sought first of all in the rare castoreum which is believed to cure the plague. The lesson we can learn from these and other examples is that nature will always be incorporated in culture, but in the process will also sneak in all the good and evil it carries inside. In that respect the impregnation of the tame rabbit by the wild hare is an allegorical image from which the whole story develops. I have addressed The light in more detail here because it demonstrates so beautifully, i.e. in an almost mythological form, the extent to which the development of religion, morality, and ethics is linked with the proc­ ess of the growth of a society. Regardless how barren and desolate the

HE WHO TEARS DOWN ALSO BUILDS UP

467

soil is, norms will always crop up, and the appropriate adage here is: a tree is known by its fruit. In that sense Lindgren’s novel has more than a little in common with William Golding’s Lord of the flies, the horrifying moral tale about a group of boys who are stranded on a desert island as survivors and start a micro-society in which the law of the jungle deter­ mines the norm. At the same time Houellebecq, Lindgren, Golding and kindred writers also make another point in their work. Happiness, har­ mony, pleasure, these are all goals worth striving for, for the same rea­ sons that evil is not. However, both edifying and hedonistic literature grow stale very quickly, as do rococo paintings of playful shepherds and shepherdesses. The writer who is unaware of evil, and who is therefore not hounded and tempted by hell, can never be a great writer.

This page intentionally left blank

CHAPTER NINETEEN

EUROPE: A QUESTION ONTO ITSELF? DONALD LOOSE Questio mihi factus sum. Saint Augustine, Confessions.

1 Introduction I have become a question onto myself. This is what Saint Augustine dis­ covered and reported in his Confessions. The tenth book is not only a pinnacle of the European literature that grew into the genre of autobiog­ raphy, but it is also the foundation of the European mind in search of its own identity. Augustine’s surprising discovery is that the answer to the question: ‘Who am I’ must be obtained through the question: ‘What am I’, which, in turn, must be answered through the question: ‘What is it that knows what I am and what I am looking for’ Augustine, who has become a question onto himself, gets trapped in a spiral of questions. The answer to the question: ‘What is it that I love when I love the god who knows me’ also turns out to be a new question. The answer is that you must keep asking questions and that only by doing so will you find an answer to the question: ‘Who am I’. I would now like to launch the rather bold hypothesis that a European Values Study survey is a distant relative of such questions, still hoping to find an answer to the question: ‘What is a European soul’. Simultane­ ously, however, it closes this question with a pseudo-answer. I would first have to show that this question is constitutive for the European soul, which I will do by referring to Paul Valéry and the Czech phenomenologist, Jan Patocka. The result of a survey of the values that the European soul can relate to would then have to produce the recogni­ tion that ‘we still have questions about our values.’ That would have to be the final value and the true answer. Could it be the case, however, that we desperately tend to smother that answer in objective tables and fig­ ures to pacify our agitation—which Augustine also felt—when the only real answer is that we have to keep asking questions. Valéry and Patocka

470

CHAPTER NINETEEN

are convinced that the European soul, which is in essence a selfquestioning and completely individual, Christian, and philosophical soul, has chiefly evaded itself and the agitation and uncertainty involved in all this. Modern individualism is increasingly being unmasked as collecti­ vism, and collectivism as this false individualism. The individual has be­ come a variable in a sociological survey and has the illusion that he or she is what the mirror of science reflects in fragments. 2 The European soul is a question onto itself In La crise de l’Esprit (1919), Paul Valéry addressed the question: ‘Who is European’. He considers as European anyone who is influenced by three cultural patterns. Firstly, the universal legal order of Rome, which turned race, territory, people, or folklore into issues that were irrelevant as determinants of identity. Anyone could be regarded as Roman (civis romanus) anywhere in the world if he or she recognized Rome’s legisla­ tion. Christianity was able to graft itself onto that, which is why it may have been one of the few religions Rome distrusted. Christianity was a formidable opponent because it appeared to be making an even higher bid for universalism than had ever been staged by Rome. Christianity, too, did away with the distinctions between Jews and heathens, slaves and free men, or men and women: ‘Thou art all children of God in the faith in Jesus Christ’ (Paul in his letter to the Galatians 3: 28). For Rome, the gods were no longer tied to tribe or place, or living on a mountain or in a temple, but could be worshipped anywhere; and Christianity like­ wise spread to the heathens. As for Rome, the race or the language, or even the very distinction between conqueror and conquered, no longer mattered because yesterday’s enemy could be granted Roman law today if he surrendered. All those who were baptized were likewise granted the dignity of the Christian, and the liberty of the children of God thus be­ came universal. However, the Roman conquest only aimed to incorpo­ rate its subjects as citizens of the state and to rule over them and their external conduct in a legal sense. The Christian conquest was oriented at depth of conscience and governance of the soul and thus introduced the phenomenon of subjective and completely universal anchorage for all morals and values. This subjective pool of the universally valuable uni­ fication of all moral values was a corrective of the objective Roman le­ gal unity. This new religion required everyone to search their own con­ science and to ask themselves what they needed to do and who they wanted to be. This religion appealed, educated, challenged, and incited the spirit to find an answer to an assignment that could never be met: be perfect.

EUROPE: A QUESTION ONTO ITSELF?

471

Valéry holds that it is the Greek mind, or philosophical reflection, that came into play to help and deal with this impossible assignment. Rome created the framework, the matrix as it were, for the Christian ap­ peal and Greek reason to merge and mingle so miraculously into the re­ flective and self-questioning European soul. In this sense, Rome just conquered to be conquered: to be overtaken by the question it had be­ come onto itself and to which it could no longer find an answer. Besides, if America is the true heir of this European spirit, one may wonder if this modern-day Rome does not just conquer to be conquered. Will it not be overtaken by the question that it is destined to become onto itself and will its successful pragmatism not be reminded of the European soul, which remained a question onto itself despite all? Will the reflective Greek keep nagging at the back of the present-day Roman’s mind? In his posthumous Heretical Essays in the Philosophy of History (1990), Jan Patocka reduces the question of all values and moral con­ cepts of a culture to that one European question: ‘How can human be­ ings be themselves’. Again, the answer has to be sought in the question itself. What would human life have to be if it were in truth other than as it appears to itself? That things appear differently than they are is a func­ tion of their always presenting themselves one-sidedly, at a distance, in a perspective, and, as a result, can assume an appearance they share with other things. That we appear to ourselves as other than we are must be based on something else. Humans are not alien to themselves as things. Humans are themselves as being other to themselves. If they are to ap­ pear to themselves as other, they must become estranged from them­ selves, and this process of estrangement must be something intrinsic to their own mode of being. Humans are themselves in being a question onto themselves. So a life can be said to be decadent and inauthentic when it loses its grasp on the innermost nerve of its functioning, when it is disrupted at its inmost core, so that while thinking itself full, it is actu­ ally draining and laming itself with every step and act. A society can be said to be decadent if it functions so as to encourage a decadent life, a life addicted to what is inhuman by its very nature. Inhuman by its very nature is the perspective from without, the perspective of a disengaged observer. The most primitive way of human self-estrangement is ecstasy and the experience of sacral orgies and rites. They contain a differentia­ tion of two worlds, a double reality. It appears as if one has been carried over into a different world from the one in which one normally lives. But a person is never in both worlds at one and the same moment, and, there­ fore, they cannot have the fundamental importance to the knowledge of the self that the ontological experience of philosophical self-reflection can have. It is Greek philosophy which introduces that unity of the soul,

472

CHAPTER NINETEEN

achieved by thought: an inner dialogue and a dialectic, which is the pro­ per method of insight and the essence of reason. It is the Platonic phi­ losophy of the soul which constitutes that first criterion of an authentic self-identity. By perceiving that the authentic, transcendent being differs from our reality of mere transient, changing opinion by virtue of its character of eternally immovable being, it first gains its own unity core, capable of resisting the pressure of various questions and problems. Care for the soul is inseparable from care for death, which becomes the true care for life; life (eternal) is born of this direct look at death, of an over­ coming death. Whereas Valéry seems to consider the Greeks as the final pedagogues of Europe, Patocka introduces Christianity to overcome this Platonic solution by an about-face. Although he also emphasizes—in Plato and Europe, his privately given Prague Seminar of 1973—that Europe is built on one single pillar, the Greek one, and that the Jewish tradition had to turn Hellenic in order to become important for Europe, he also stresses the definite correction of the Christian perspective. Christianity, after all, understands the good differently than Plato as a self-forgetting goodness and a self-denying love. The soul is no longer related to an ob­ jective truth, but rather to a Subject who looks into the soul without be­ ing itself accessible to its view. This inscrutable relation of the soul to the absolute reveals its inmost content as guilt towards an infinite love and as a responsibility from which there is no escape. The ultimate na­ ture of the soul, by nature wholly incommensurate with the mystery of its eternal being, has to do with this dialogue with its own being. By vir­ tue of this foundation in the abysmal deepening of the soul, Christianity remains thus far the greatest, unsurpassed but also un-thought-through human secret that enabled humans to struggle against the decadence of man and his values, and represents the birth of Europe. However, ac­ cording to Patocka, Europe never explicitly thematized and never grasped philosophically that fundamental abyss of its own identity as a central question. The European soul, therefore, had to become a question for itself. The context of political Rome, the self-reflection of the Greeks, and the ap­ peal of the Christian God to the people’s own inalienable conscience have turned the European into a lasting question for himself. That appeal is universal in that it applies to everybody indiscriminately and utterly singular in that it cannot be answered in general terms. According to Pa­ tocka, however, we have not asked ourselves that question; rather, we have tried to sidestep it in general answers that are no answers.

EUROPE: A QUESTION ONTO ITSELF?

473

3 Answers that are no answers Have we, who are a question to ourselves, turned into mere majorities or minorities in surveys? Have we calculated the sacrifice of objective truth, which we, as Christian European souls, have been forced to make time and again, in the economy of values, with which we hope as yet to buy an identity of our own? Have we bought off the endless guilt of our bottomless personal responsibility with a Christianity that is nothing but the return of objective Platonism for the people? What is a survey but an examination of opinions, conducted by asking people questions and re­ flecting the answers one wants to hear, so that one can reckon with them and capitalize on them? Instead of the inscrutable appeal of the Infinite to our inalienable responsibility, the European Values Study reveals who we are in general, how we behave, and how we should use that informa­ tion. It is the inhuman perspective from without, the perspective of the disengaged observer. According to Patocka, the Christian European soul rejected the Pla­ tonic solution of a general objective standard, but always accepted ex­ tensive elements of a solution launched along Platonic lines. Platonic ra­ tionalism, the Platonic effort to subject even responsibility itself to the objectivity of knowledge, continues to affect the nether layers of the Christian conception. The problem of the individual, the problem of the human person, was from the start the problem of transcending the ordi­ nary. It implied simultaneously that humans cannot be identified with any role they may assume in the world, with any answer they give to the fundamental question of who they are. But socialism and the consumer­ ism of liberalism, totalitarianism and democracy, the claim of a political correctness and the overruling force of a market-oriented way of thin­ king, for all their profound differences, equally overlook everything that is neither objective, a standard, nor a role, all that is the real Greek and Christian core of a European answer to the most important question of human life: Je est un autre. As Weber has already suggested, the management of our personal identity by means of a functional capitalization on every mystery of life can only result in a complete disenchantment of our world. Boredom then, says Patocka, does not retreat, but rather forces its way to the fore­ front. Nor does it assume only the refined forms of an aesthetic and ro­ mantic protest, but also the clear form of consumer offerings and the end of utopia, brought about by technical or scientific means, the violent re­ action of irrational religious fundamentalism and the return of the most primitive way of self-estrangement in ecstasy or the experience of pagan orgies and rites. In the form of a compulsory recreation, consumerism

474

CHAPTER NINETEEN

becomes the law of engineered amusement ending up in boredom, infil­ trating even the world of science and research. The most sophisticated inventions are boring if they do not lead, by what we discover, to the solving of the mystery concealed. The powerful penetrating ability of the human mind mainly uncovers what is directly seized by the everyday experience and discovers only what is already fully uncovered and cleared. We know that today’s findings turn into tomorrow’s common gossip and triviality, and that today’s remaining mystery will be the tru­ ism of tomorrow. Of course, we still are a question to ourselves and we are looking for an answer, mostly concealing the core of the question. In Europe: The exceptional case. Parameters of faith in the modern world (2002), Grace Davie underlines that the empirical data from the European Values Study confirm the notion that the value systems of modern Europe have a considerable amount in common, a fact easily explained by a shared religious heritage. No one disputes that the Judeo-Christian tradition— massively present in all parts of Europe—was a crucial factor in the for­ mation of European values. However, European citizens mainly behave as though ‘believing without belonging’. The historical churches are sys­ tematically losing their capacity to discipline the religious thinking of large sections of the population, especially the young. Belief, however, not only persists, but is also becoming increasingly personal, detached, and heterogeneous. Is Europe finally becoming European? The unwill­ ingness to attend to religious institutions on a regular basis, due to the di­ rect legacy of the official—in a sense Platonic—status of Christianity does not necessarily mean a parallel negative personal answer to Chris­ tian belief. However, by and large, Europeans still regard their churches as public utilities. They look at them with benign benevolence, since they consider them useful social institutions, which the large majority of the population is likely to need at one time or another in their lives. In fact, Europeans look upon their churches as institutions maintaining the values and having at their disposal the useful answers to the questions they are not raising yet but that they might be confronted with some day. In that sense, the European Values Study survey functions as a consola­ tion and a reassurance: a reflecting mirror of the general answers which are still there, should we raise a question. A perfect objective Platonism for the European people in general. In La liberté de l’Esprit (1939), Paul Valéry outlines what is happen­ ing nowadays to that European spirit, to personal and individual free­ dom, and singular responsibility. Once, this was the unique foundation of all values; nowadays, it is a value besides petroleum, real estate, or gold. The fact is, all values have become fluctuating consumer goods;

EUROPE: A QUESTION ONTO ITSELF?

475

they become interchangeable. They become part of a fever, and all we can say about them is their current, momentary quotation. In that case, must the European values study not confront us with an inescapable dilemma? The only and decisive question is the question about the state of the individual and inalienable sense of responsibility of the single person and the absolute value of the freedom of the European mind, since without them, the answers make no sense whatsoever. They are no more than second-hand opinions, parroted convictions, and vari­ ables that are subject to the fluctuations of the intellectual and economic climate of the day. But we do not ask that fundamental question differ­ ently than any of the many questions from the questionnaire. ‘Do you feel you have completely free choice and control of your life?’ is of the same order as ‘Is it important to have a job respected by people in gen­ eral?’ Thus, the possible probing into that one question that is the be-all and end-all of our existence is lost in what Valéry calls the language of the stock exchange, with freedom at most getting a mention as a ‘mind­ value’ beside other values of the stock exchange, where everything can be exchanged and reinvested. Nothing there can be of absolute value, nor can any answer concerning current values something valuable. All of it has informative value, but none of it has any meaning whatsoever. One may wonder, therefore, whether a European Values Study survey, not only revealing the values we entertain but revealing who we are in gen­ eral, really constitutes an answer to that one question that can truly be called European: ‘Who am I and what must I do’. 4 One final question In The Gift of Death (1999), Jacques Derrida wonders what would have become of Europe, its values, and its politics if that view, which Patocka himself calls heretical, had become reality with respect to the neverabandoned Platonism in Christianity. For the question about the univer­ sal and, at the same time, utterly singular mystery that a man is unto himself, the singular anxiety of man as to what his value and his inalien­ able own values are, has been dodged by Europe in its science and poli­ tics in general considerations about European values, a natural law, and international politics. To get close to this mysterium tremendum, Europe should first of all make a clean break with its suppressed and everreturning Roman/Greek Platonism. The by now heretical message might then be heard that Europe will only have become what it ought to be if in a sense it has become entirely Christian. Derrida also wonders whether it will not at the same time understand itself as truly Jewish and standing under the influence of Islam, or perhaps as the philosophical-religious

476

CHAPTER NINETEEN

continent without a religion. Such questions, incidentally, would not only imply a critical backlash for Europe’s religion and ethos, but also for law, politics, science, and the task of the university. However, Europe prefers to cover up these questions with the fragmented mirror in which it still recognizes itself. This lasting recognition, by the way, of Europe in its own Roman-Greek-religious tradition is at least as valuable for the cohesion of society as the conclusion that Europe, despite every­ thing, has remained a periodically returning question for itself. But it should cause no surprise that this cohesion is being experienced more and more as the perspective from without: the perspective of the disen­ gaged observer.

ABOUT THE AUTHORS Wil Arts is a professor of general and theoretical sociology and head of the sociology department at Tilburg University. Egidius Berns is a professor of philosophy and dean of the faculty of philosophy at Tilburg University. Sjoerd Beugelsdijk is a research fellow at the Center for Economic Re­ search (CentER) at Tilburg University. Arnoud-Jan Bijsterveld is a professor of culture and cultural heritage in Brabant at Tilburg University. Lans Bovenberg is a professor of economics and scientific director of the Center for Economic Research (CentER) at Tilburg University. Paul Dekker is a professor of civil society at Tilburg University and a senior research fellow at the Social and Cultural Planning Bureau in The Hague. Wim van de Donk is a professor of public administration at Tilburg Uni­ versity and a member of the Scientific Council of Government Policy in The Hague. Veerle Draulans is an assistant professor of practical theology at Tilburg University and an associate professor of gender studies at the Catholic University of Leuven in Belgium. Peter Ester is a professor of general and theoretical sociology and direc­ tor of the Institute of Labor Studies (OSA) at Tilburg University. John Gelissen is an assistant professor of social science methodology and statistics at Tilburg University. Jaap Goedegebuure is a professor of theory and history of literature and dean of the faculty of arts at Tilburg University.

478

ABOUT THE AUTHORS

Jacques Hagenaars is a professor of social science methodology and sta­ tistics and vice-dean of the faculty of social and behavioral sciences at Tilburg University. Loek Halman is an associate professor of sociology at Tilburg Univer­ sity, secretary to the Board and the Steering Committee of EVS, and program director of the 1999 EVS study. Frank Hendriks is an associate professor of law, public administration and informatization at Tilburg University. Ernst Hirsch Ballin is since 1981, interrupted by his term as Minister of Justice from 1989 to 1994, a professor of law at Tilburg University, and, since 2000, member of the Council of State. Matthijs Kalmijn is a professor of sociology at Tilburg University. Donald Loose is a professor of philosophy at Tilburg University. Ruud Luijkx is an assistant professor of sociology and data manager at the Work and Organization Research Institute (WORC) at Tilburg Uni­ versity. Guy Moors is an assistant professor of social science methodology and statistics at Tilburg University. Ruud Muffels is a professor of socio-economics and social security and a research director of the Institute of Labor Studies (OSA) at Tilburg Uni­ versity. Niels Noorderhaven is a professor of organization and strategy at Til­ burg University Wim van Oorschot is a professor of sociology and social policy at Til­ burg University. Anton van Schaik is a professor of economics and director of the pro­ gram in economics at Tilburg University. Sjak Smulders is an associate professor of economics at Tilburg Univer­ sity and a research fellow at the Royal Netherlands Academy of Arts and Sciences (KNAW) in Amsterdam.

ABOUT THE AUTHORS

479

Karel Veraghtert is a professor of economic and social history at Tilburg University. Henk Vinken is a scientific director of the Institute for Research on Inter­ cultural Cooperation (IRIC) at Tilburg University. Marga van Zundert is a freelance science writer.

This page intentionally left blank

AUTHOR INDEX Abramovitz, M. 115, 143

Alestalo, M. 309, 370

Alm, S. 309

Almond, G.A. 246, 274

Alwin, D.F. 337

Amin, A. 115

Anheier, H.K. 274

Arber, S. 246, 308

Arrow, K. 143

Arts, W. 11, 215, 308, 369, 447, 448

Atkinson, A.B. 308

Attias-Donfut, C. 246, 308

Audretsch, D. 115, 117

Avineri, S. 399

Baden-Fuller, C.W.F. 116

Bailey, J. 11, 215

Baker, K. 178

Baker, W.E 12, 179, 215, 247, 399, 419,

448

Barber, B.R. 246, 274

Barnard, B. 85

Barnes, S. 246

Barr, N. 308

Barrett, D.B. 399

Barro, R.J. 115, 143, 178

Bartlett, R. 85

Baumol, W. 115, 143, 178

Beck, U. 399

Becker, G.S. 369, 418

Beckford, J.A. 399

Beer, S. 274

Begg, I. 115, 143

Belk, R.W. 178

Bell, D. 448

Bellah, R.N. 246

Benassi, M. 144

Berger, P.L. 246

Berger, S. 274

Bernard, J. 369

Berry, J.M. 246

Bertrand, M. 143

Beugelsdijk, S. 115, 116,144, 178, 274,

418, 419, 246

Bils, M. 178

Binmore, K. 418

Bisin, A. 418

Blockmans, W. 85

Boekema, F. 145

Boggs, C. 178

Boix, C. 178

Boje, T. 308

Bond, M. 145

Bonoli, G. 448

Booth, A. 336

Bosch, A. 85

Bosker, R. 370

Bourdieu, P. 274

Bovenberg, A.L. 178, 418

Brady, Th.A. 85

Brady, H.E. 247

Brandstätter, H. 115

Braun, M. 337

Brockhaus, R.H. 115

Bryson, C. 309

Burgers, J. 309

Burt, R. 144, 178

Camburn, D. 337

Carland, J.C. 116

Carland, J.W. 116

Chambers, S. 246

Charlemagne 11

Chenery, H.B. 115

Cherlin, A. 336

CN 431

Coleman, J.S. 144, 178, 274

Colombo, M. 115

Cooke, P. 115

Cotran, E. 431

Coughlin, R. 309

Cowling, M. 115

CPB 418

Crittenden, J. 309

Crowder, K. 370

Dafoe Whitehead, B. 369

Dalton, R. 178

Davie, G. 399, 474

Davies, N. 85

Davis, K. 336

Dawson, S. 179

De Graaf, N.D. 215

De Groot, H.L.F. 144

De Hart, J. 309

De Jong, A. 418

De Kadt, R. 144

De Moor, R. 11, 48, 215, 337, 399

482

AUTHOR INDEX

De Renzio, P. 179 De-Shalit, A. 399 De Swaan, A. 309 De Tocqueville, A. 145 De Wit, Th.W.A. 432 Deakin, N. 246 Dekker, P. 178, 246, 274, 309 Delmastro, M. 115 Den Boer, P. 85 Derrida, J. 461, 475 Deutsch, K.W. 11 Diani, M. 246 Diepstraten, I. 246 Dirven, H.J. 448 Dobbelaere, K. 399 Douglas, J.A.T. 274 Dulles, A. 431 Dunkerley, D. 85

Geertz, C. 116 Gelissen, J. 215, 308, 309, 369, 447, 448 Georgellis, Y. 116 Gijsberts, M. 215 Ghoshal, S. 179 Glaeser, E.L. 418 Golding, W. 467 Goldman, G. 274 Golodner, A. 116 Goode, W.J. 370 Goodin, R.E. 309, 448 Graafland, J. 418 Granato, J. 144 Granovetter, M. 144 Greif, A. 418 Grier, K.B. 144 Grunebaum, G.E. 431 Gulati, R. 144, 179

Eberly, D.E. 246 Edmunds, J. 246 Edwards, B. 246 Ehrenberg, J. 246 Engbersen, G. 309 England, P. 336 Esping-Andersen, G. 309, 369, 370, 448 Ester, P. 48, 246, 247, 274, 309, 399 Etxebarria, G. 115 Etzioni, A. 48, 246, 431, 399 European Commission 144 European Investment Bank 144 Evans, D.S. 115 Evink, E. 461 Fagerberg, J. 144, 179 Farkas, G. 336 Fedderke, J. 144 Ferrera, M. 309, 448 Ferry, J.M. 461 Fine, B. 179, 274 Finke, R. 400 Fiorina, M.P. 247 Flap, H. 309 Flora, P. 309 Florida, R. 115 Fogarty, M. 11, 309 Foley, M.W. 246 Forma, P. 309 Francois, P. 179, 418 Fukuyama, F. 144, 179, 246, 274, 309 Füstös, L. 216

Habermas, J. 461 Hagemann, L. 431 Hagenaars, J.A.P. 48 Hall, R.E. 144 Haller, M. 215, 336 Halman, L. 11, 48, 215, 309, 310, 337, 370, 399, 431, 461 Hamberg, E. 399 Harding, S. 11, 309 Harriss, J. 179 Headey, B. 448 Hechter, M. 11 Heidenheimer, A. 309 Helliwell, J.F. 144 Heinen, T. 309 Hermkens, P. 215 Heunks, F. 48 Hill, C.T. 336 Hildebrandt, K. 178 Hobsbawm, E. 85 Hodginson, V. 246 Hodgson, L. 85 Hoellinger, F. 336 Houellebecq, M. 465 Hofstede, G. 48, 116, 144, 246, 274, 418 Hoppenbrouwers, P. 85 Hospers, G.J. 116 Houtman, D. 310 Huizinga, J. 11 Huntington, S.P. 85, 431 Huxley, A. 465 Hyman, H.H. 336

Gallie, D. 309 Gambetta, D. 144 Gargiulo, M. 144 Gartner, W.B. 116 Geary, P.J. 85

Iannacone, L. 400 Inglehart, R. 11, 12, 48, 144, 179, 215, 246, 274, 336, 399, 419, 448 Inkeles, A. 144 International Herald Tribune 431

AUTHOR INDEX Jackman, R.W. 179

Jagodzinski, W. 399

Jansen, M. 336

Janssen, C. 309

Jehoel-Gijsbers, G. 309

Jones, C. 144

Johnson, P.A. 145

Johnson, T.M. 399

Johnsson, F.H. 48

Kangasharju, A. 116

Kahneman, D. 370

Kalmijn, M. 336

Kamano, S. 370

Kant, I. 462

Karatnycky, A. 215

Kasza, G. 309

Kay, J. 419

Keefer, P. 144, 179, 274

Kerkhofs, J. 461

Kiecolt, K.J. 370

Kiernan, K. 370

King, D.C. 247

Klenow, P.J. 178

Kluegel, J.R. 215, 216

Knack, S. 144, 179, 274, 419

Knijn, T. 309

Knudsen, K. 370

Kohli, M. 247, 309

Komter, A. 309, 310

Konopacki, S. 85

Koopmans, R. 178

Korczynski, M. 144

Kormendi, R.C. 144

Korpi, W. 448

Kraaykamp, G. 336

Krueger, A.B. 179

Kuhnle, S. 309, 310, 370

Kurian, G.T. 399

Kymlicka, W. 246

Laibson, D.I. 418

Lawson, C. 116

Lebeaux, C. 448

Leblang, D. 144

Leerssen, J. 85

Leggewie, C. 431

LeGrand, J. 310

Lehmann, K. 431

Leibfried, S. 448

Leighton, L.S. 115

Leonardi, R. 144, 145, 179, 247, 310

Lesthaeghe, R. 336, 370

Levinas, E. 462

Levine, R. 144, 179

Lewis, B. 431

Lijphart, A. 274

Lindahl, M. 179

Lindgren, T. 465

List, R. 274

Lorenz, E. 116

Lucas, R.E. 179

Luhmann, N. 144

Luiz, J. 144

Luttmer, E.F.P. 143

MacIntyre, A. 399

Maddison, A. 215

Madesen, R. 246

Malecki, E. 116, 144

Mankiw, N.G. 116, 145

Manning, N. 370

Maraffi, M. 145, 247

Maritain, J. 432

Markusen, A. 116

Marshall, A. 116

Martin, D. 399, 400

Martin, R. 145

Mason, D.S. 215, 216

Maurseth, P.B. 145, 179

Mayes, D. 448

McCelland, D. 116

McLean, S.L. 247

Meekers, D. 370

Meeus, M.T. 145

Meguire, P.G. 144

Mellens, M. 216, 370

Menocal, M.R. 431

Mertens, T. 462

Milanovic, B. 216

Milgrom, P.R. 418

Miller, R.A. 179

Mincer, J. 179

Misztal, B.A. 247

Molina, A.M. 432

Molle, W. 116, 145, 179

Mommsen, W. 310

Moore, R.J. 85

Moors, G. 336

Moors, H. 370

Morgan, K. 116

Motyl, A. 215

Mueller, S.L. 116, 117

Muffels, R. 448

Mullainathan, S. 143

Myers, S.M. 336

Nahapiet, J. 179

Nanetti, R.Y. 145, 179, 247, 310

Nelson, R.R. 179

Neuhaus, R.J. 246

Nevitte, N. 12

Newton, K. 145, 247, 274

Nicoll, W. 85, 461

483

484

AUTHOR INDEX

Nie, N.H. 247 Nijkamp, P. 116, 145 Noorderhaven, N. 115, 178 Nooteboom, B. 145 Norris, P. 247 Nye, J.S. 247 Oberman, H.A. 85 OECD 116 Oerlemans, L.A.G. 145 Olson, M. 145 Opp, K.D. 11 Oppenheimer, V.K. 336 Ottati, G.D. 116 Paci, R. 116 Paling, K. 85 Palomba, R. 370 Parsons, T. 145 Pascall, G. 370 Patocka, J. 471, 472 Paxton, P. 145, 179, 310 Penrose, E. 116 Peplau, L.A. 336 Pettersen, P. 310 Pettersson, T. 399 Petty, R.E. 336 Pétursson, P. 48 Pharr, S. 274 Phelps, E.S. 179 Phillips, D. 11, 309 Poot, J. 116, 145 Porter, M.E. 116 Portney, P.E. 246 Pose, A.R. 116 Posner, D.N. 178 Putnam, R.D. 145, 179, 310, 247, 274 Quah, D. 116, 145 Rabellotti, R. 116 Rabin, M. 419 Rebelo, S. 179 Rein, M. 309 Renelt, D. 144, 179 Richins, M.L. 179 Rietbergen, P. 85 Riis, O. 431 Ring, P.S. 145 Rokeach, M. 179 Romer, D. 116, 145 Rothstein, B. 310 Rotter, J.B. 116 Rubin, Z. 336 Sainsbury, D. 337, 370 Sako, M. 145

Sala-I-Martin, X. 115, 143, 178 Salamon, L.M. 274 Salmon, T. 85, 461 Sandel, M.J. 247 Saris, W.E. 419 Scarbrough, E. 48 Scheinkman, J.A. 418 Scherpenzeel, A.C. 419 Schlozman, K.L. 247 Schnetzer, A. 215 Scholten, H. 309 Schulze, G. 247 Schumpeter, J. 116 Schuyt, C. 310 Schwartz, S.H. 48, 247 Scott, J. 337 SCP 247 Seligman, A.B. 247 Selle, P. 247, 310 Sen, A.K. 448 Shane, S.A. 116 Sherif, A.O. 431 Siaroff, A. 370 Sierksma, F. 432 Sinn, H.W. 419 Skocpol, T. 247 Smit, H. 116, 145, 179 Smith, A. 419 Smith P. 145 Smulders, S. 246, 418 Snijders, T. 370 South, S.J. 370 Soutter, C.L. 418 Spieker, M. 310 Spybey, T. 85 Stark, R. 400 Stewart, J.R. 116 Stopford, J.M. 116 Storey, D.J. 116 Storper, M. 116 Strømsnes, K. 247 Sullivan, W.M. 246 Sunley, P. 145 Svallfors, S. 310, 370 Swank, D. 145 Swidler, A. 246 Szakolczai, A. 216 Szydlik, M. 247 Tarrow, S. 179 Taylor, C. 400 Taylor, M. 115 Taylor-Gooby, P. 310 Teachman, J. 370 Temple, J. 145 Ten Hooven, M. 432 Ter Meulen, R. 448

AUTHOR INDEX Teulings, C.N. 418

The Treasury 447

Therborn, G. 310, 400, 448

Thomas, A.S. 116, 117

Thomas, K. 117

Thompson, A. 85

Thomson, K. 246

Thornton, A. 370, 337

Thurik, R. 117

Tibi, B. 432

Tilly, Ch. 85

Tipton, S.M. 246

Tirole, J. 419

Titmuss, R.M. 448

Toepler, S. 274

Tomaney, J. 115

Tomka, M. 48

Tracy, J.D. 85

Trent, K. 370

Triandis, H. 48

Tsakloglou, P. 448

Tullock, G. 144

Turner, B.S. 246

Turner, F.C. 48

Tversky, A. 370

Ulam, A.B. 274

UNDP 48

United nations 370

Uranga, G. 115

Usai, S. 116

Uslaner, E. 246

Uslaner, M. 310

Uzzi, B. 145

Van de Ven, H.A. 145

Van den Akker, P. 337

Van den Broek, A. 178, 216, 246, 247

Van den Dussen, J. 85

Van der Lippe, T. 337

Van Deth, J.W. 48, 145, 247

Van Dijk, L. 337

Van Harskamp, A. 400

Van Holst, B. 116, 145, 179

Van Oorschot, W. 310

Van Schaik, A.B.T.M. 115, 144, 178, 274,

418, 419

Van Snippenburg, L. 400

Van Wijck, P. 215

Valéry, P. 470, 474

Veldheer, V. 85

Verba, S. 247, 246, 274

Verdier, T. 418

Verheul, I. 117

Verspagen, B. 144, 179

Vinken, H. 246, 247, 274

Vissers, A. 309

Vloet, A. 48

Waerness, K. 370

Wall, H.J. 116

Walzer, M. 247

Warren, M.E. 247, 274

Watson, W.E. 116

Wayne, H. 116

Weber, M. 448

Weil, D. 116, 145

Weingast, B.R. 418

Wegener, B. 215, 216

Wegener, D.T. 336

Wennekers, S. 117

Wessels, L.H.M. 85

Whiteley, P. 145, 247, 310

Wilensky, H.L. 448

Williamson, O.E. 145

Wilson, B. 400

Wilson, J.Q. 310

Wilson, K. 85

Wippler, R. 11

De Wit, Th.W.A. 432

Wojciech, S. 274

Wolfe, A. 247, 310

Woolcock, M. 145

Wright, C.R. 336

Wuthnow, R.D. 247

Yankelovich, D. 310

Young-DeMarco, L. 337

Zak, P.J. 419

Zanders, H. 48

Zelikow, P.D. 247

Zibojnik, J. 418

Zijderveld, A. 310, 432

485

This page intentionally left blank

SUBJECT INDEX abortion, 28, 31, 39, 54 absolutism, 80 activity associational, 8, 120, 121, 123–125, 141, 142, 149, 169 intra-preneurial, 98, 113 rent-seeking, 152, 154, 155, 169, 176 aggregate level, 23, 25, 47, 281, 288, 290, 296, 305–308, 333, 347, 384 agnostics, 2, 19, 97, 98, 113, 149–151, 381 amoralism, 464 anomie hypothesis, 196, 214 associations non-profit, 263 inner-directed, 221 voluntary, 129, 147, 169, 178, 218–221, 234, 243, 263–265, 278, 292, 310 attitude entrepreneurial, 8, 95–97, 99, 101, 108, 109, 111, 113, 114 materialistic, 147–149, 151, 152, 159, 166 pro-social, 278, 297 sex-role, 10, 311–316, 318, 321, 322, 324–328, 330, 331, 333–335, 337 authority, 29, 30, 47, 53, 56, 73, 227, 230, 231, 304, 372, 376, 422, 460, 463, 466 collective, 372 autonomy, 16, 17, 19, 28, 30, 32–34, 36, 37, 39, 40–44, 46, 53, 227, 304 individual, 15, 19 personal, 30, 33 believing without belonging, 474 benefits, 55, 57, 99, 103, 124, 129, 147, 149, 152, 156, 157, 162, 166, 225, 226, 234, 236, 241, 269, 277, 278, 284, 285, 287, 291, 301, 302, 305, 310, 344, 380, 404, 407, 408, 410, 412, 413, 417, 434, 441 Catholicism, 82, 192, 232, 304, 346, 374, 376 Catholic, 18, 19, 35, 36, 37, 285, 303, 347, 348, 353, 355, 357, 359, 361, 364, 367, 369, 376, 381, 382, 423 center-periphery, 376 cheating, 57, 147, 148, 155, 157, 178, 269, 284, 304, 305, 380, 388, 406, 408

Christendom, 3, 72, 73, 76, 80, 82, 195 Christianity, 3, 16, 19, 60, 71, 73–75, 80, 81, 87, 195, 422–426, 428, 430, 463, 470, 472–475 Christians, 18, 84, 422–425, 428, 431 church attendance, 52, 227, 234, 237, 285, 298, 299, 304, 314, 315, 320, 326, 327, 329–332, 334–336, 350, 353–355, 357, 359–361, 364, 366, 367, 369, 371, 383 citizen involvement, 217 self-made, 238 citizenship, 72, 75, 129, 276, 279, 398, 458 civic community, 124, 223 culture, 219, 223, 224, 246, 266 engagement, 9, 124, 150, 152, 219, 221, 238, 239, 243, 261, 273 morality, 9, 275, 279–282, 284, 285, 290–292, 295–297, 298, 303–308 networks, 124, 150 participation, 218, 238, 263 religion, 376 sense-versus-permissiveness, 380 virtues, 217 civilization, 2, 18, 59, 63, 70–72, 74, 76, 82, 83, 91 Christian, 71 Western, 425 clash of civilizations, 423 cohabitation, 2, 339 cohesion, 80, 84, 144, 218, 243, 260, 476 cohort, 7, 37, 40–43, 46–48, 189, 192, 197, 199, 204, 208, 209, 211, 212, 214, 239, 299, 303, 316, 326, 350, 360, 368 Cold War, 77 collectivism, 247, 470 communism, 7, 78, 185, 192, 193, 229 community, 60, 68, 72–74, 78–80, 92, 123, 124, 129, 148, 153, 155, 156, 176, 217, 218, 224, 238, 243, 244, 247, 250, 259, 260, 268, 276, 277, 279, 304, 398, 422, 426, 427, 429, 438, 440, 447, 455, 459, 460 erosion of, 217 global, 374 spirit of, 31, 398

488

SUBJECT INDEX

competition, 17, 55, 62, 91, 98, 123, 144, 178, 209–212, 307, 374, 376, 377, 387, 418, 419 confederation, 458, 459, 461 confidence, 20, 56, 63, 64, 121, 139, 224, 284, 379, 449 conformity, 409 conscience, 407, 465, 470, 472 conservatism, 42, 227, 279 conservative, 20, 40, 61, 279, 280, 283, 287, 304, 314–316, 321, 323, 326, 328, 329, 332, 334, 339, 343–345, 351, 353, 366 constitution American, 1 European Union, 1 consumerism, 277, 473 convergence, 5, 73, 85, 111, 115, 116, 137, 141, 143–145, 165, 168, 173, 174, 178, 190, 212, 412, 413, 415, 433, 436 cooperation, 3, 4, 11, 77, 89, 122–124, 213, 218, 243, 255, 261, 407, 409, 418, 450 corruption, 64, 154, 178, 195, 200, 202, 205, 206, 209, 213, 227–229, 231, 260, 285, 305 cosmopolitan, 451, 453, 454, 456, 458, 460 cultural change, 179, 190, 247, 336, 410, 411, 413, 417, 419 diversity, 1, 70, 75, 79, 81–83, 85, 226, 403, 412, 414, 433 heritage, 6, 35, 37, 47, 68, 77, 85, 192, 375, 396, 397, 410, 412, 414, 417, 418, 437 homogeneity, 10, 73, 403 zones, 5, 436, 437 culture communitarian, 265 masculine, 229 decentralization, 264, 405, 418 Declaration Franco-British Union, 67 Human Rights, 458 de-commodification, 343 democracy consensus, 9, 256, 257, 265 consensual, 273 consolidated, 200 majoritarian, 258 multi-party, 185, 186 pluralistic, 188 quality of, 220, 258 social, 192 vital, 218, 260

denomination, 35, 80, 227, 285, 299, 303, 312, 319, 347, 348, 350, 360, 374, 376, 377, 381, 382, 387, 426 despotism, 80, 459 dialogue interreligious, 10, 422–424, 426–430 differentiation, 52, 72, 222, 312, 372, 379, 396, 413, 471 discrimination, 424 disintegration social, 444 divorce, 10, 57, 267, 316, 320, 326, 327, 329, 336, 339–362, 364, 366–369, 380, 445 laws, 342, 344 patterns, 344 rate, 342, 343, 350– 356, 361, 364, 367, 369 regimes, 344, 368 duty, 51, 219, 269, 412, 423, 425, 460 Earth Charter, 59, 61, 65 economic changes, 6, 312, 413 development, 33, 34, 37, 47, 83, 95, 116, 119, 121, 129, 136, 144, 145, 151, 179, 189, 191, 192, 205, 208, 214, 218, 227, 228, 374, 384, 396, 403, 410, 419, 425, 436, 437, 450 economy dynamic, 404 statist, 188, 201 transitional, 188, 201 service-oriented, 405 egoism, 31, 304, 459 employment female, 313, 314, 316, 320, 331, 332, 335, 336, 340 full-time, 327 part-time, 327 entrepreneurship, 95, 101, 116, 117 corporate, 98, 116 equal rights, 74, 311, 333 equality, 16, 75, 102, 103, 106, 198, 247, 250, 299, 316, 336, 454, 456 Erasmus, 62, 65, 68, 421 ethical relativism, 31, 398 ethnocentrism, 260, 285, 290 ethos, 28, 51, 309, 476 immoralistic, 277 individualist, 303 European Central Bank, 411 Community, 4, 68, 116, 145, 179, 438, 442, 448 culture, 15, 18, 67, 76, 80, 81, 425 identity, 59, 70, 81, 84

SUBJECT INDEX Investment Bank, 120, 144

legislation, 90

soul, 11, 469–473

spirit, 2, 471, 474

unification, 68, 450, 459

faith, 18, 53, 63, 73, 80, 91, 304, 372, 397, 421, 423, 424, 426, 427, 430, 464, 470, 474 familialism, 339, 344 family nuclear, 339, 343 one-earner, 339 ties, 154, 158, 161, 166, 169, 173, 177, 440 federalism, 259, 450, 459 fertility, 344, 348 flexibility, 6, 123, 150, 259, 412, 443 market, 443 fragmentation, 3, 449 freedom, 16–18, 30, 55, 62, 67, 71, 75, 82, 84, 102, 103, 107, 170, 199, 227, 231, 232, 259, 267, 304, 412, 418, 439, 453– 456, 458, 474, 475 Freedom House, 200, 215, 227 fundamentalism, 60, 63, 423, 426, 427, 473 gender attitudes, 312, 323, 327, 328 General Social Surveys, 283 generation, 31, 32, 39, 43, 47, 214, 238, 239, 240, 242, 243, 245, 246, 299, 375, 384, 390, 392, 430 Baby Boom, 239–241, 243 Baby Bust, 239–242, 245 postwar, 220, 238 prewar, 238, 239 war, 239 globalization, 6, 59, 61, 62, 64, 65, 374, 383, 384, 387, 389, 396, 437, 456, 459 governance effective, 150, 255, 256 government intervention, 198, 199, 201, 205, 208, 209, 212–214, 404 transitional, 200 hedonism, 15, 18, , 277 Herfindahl index, 382 households single-person, 339 human capital, 104, 105, 110, 119, 136, 161–165, 168, 173, 405, 440, 441, 443 entrepreneurial, 95, 98 Human Development Index, 36 human rights, 84, 129, 250, 428, 429

489

humanism, 60, 62, 73, 422, 428 hypothesis communicating vessels, 279– 281, 287, 291, 293, 296, 306, 307 identity, 3, 23, 59, 70, 72, 78, 81, 82, , 271, 287, 303, 376, 383, 405, 413, 469, 470, 472, 473 collective, 377 cultural, 64, 69–72, 76, 78, 431 national, 1 personal, 473 social, 190 ideologies religious, 75, 332 imagination, 53, 98, 102, 213, 463 immigrants, 50, 58, 61, 284–286, 288– 290, 295, 297, 298, 300–303, 307, 308, 421 independence, 53, 102, 259, 264, 348 individualism, 15, 18, 20, 31, 35, 226, 247, 304, 309, 398, 465, 470 false, 470 individualization, 3, 60, 220, 304, 312, 345, 349, 371, 396, 398 inequality, 227, 231, 418, 441, 442, 444 innovation, 95, 98, 99, 100, 114–117, 145, 164, 263, 404, 440 institutions legal, 9 religious, 374, 422, 474 integration, 4, 67, 68, 69, 123, 215, 403, 411–414, 417, 418, 422, 430, 433, 439, 443, 459 economic, 4, 10, 403, 411, 413, 417 policy, 434 social, 219, 279, 339, 438, 439, 441, 443 societal, 343 International Social Justice Project, 189 International Social Survey Programme, 314 intolerance, 50, 423, 427 Iron Curtain, 68, 77, 81 Islam, 19, 80, 84, 421–426, 428–432, 475 Jews, 50, 285, 422, 424, 425, 431, 470 Judaism, 16, 82, 422 justice, 56, 67, 74, 82, 187–189, 193, 194, 196, 204, 205, 213, 308, 381, 404, 422, 429, 433, 466 principles, 189, 204, 213, 308 social, 75, 189 socialist, 9, 187, 189, 193–196, 198, 199, 205, 212, 214 socio-economic, 188, 189, 193–198, 209, 212, 214

490

SUBJECT INDEX

justification, 57, 353, 380, 439, 454 labor division of, 311, 348, 407, 412 female participation, 324, 332 labor force characteristics, 10, 97, 313 labor market, 97, 104, 312, 345, 349, 438, 442, 443 labor unions, 186, 433 League of Nations, 3, 76 legacy of Antiquity, 71 liberalism, 34, 36, 37, 40, 41, 43, 44, 74, 279, 473 liberty, 75, 76, 151, 267, 268, 454, 470 civil, 75, 227, 232 life course, 37, 41, 191, 214, 299, 316, 328, 336, 434, 443, 447 quality of, 87, 88, 228, 229, 349, 437, 441, 446 locus of control, 99, 100, 102, 104, 116 manifesto Diversity Within Unity, 424, 431 market economy, 9, 185–188, 190, 194, 196, 197, 201, 212, 213 justice, 9, 187, 189, 193–202, 204, 205, 209, 212–214 materialism, 31, 46, 55, 151–153, 157– 160, 163, 166, 170–173, 175, 177–179 membership active, 129, 130, 134, 138, 142, 221, 329, 416 group, 121, 124, 125, 129, 130, 131, 134, 138, 141, 143, 169, 261, 416 passive, 129, 139, 141 modernity, 74, 79, 255, 345 modernization, 3, 6, 74, 85, 190–195, 212–214, 220, 226, 256, 374, 376, 378, 396, 436 monarchy, 73, 453, 459 monetary union, 411, 413 monopoly, 81, 221, 376, 377, 396 moral bonds, 218, 243 Christian, 73, 299 corruption, 277, 278, 290, 301, 305, 308, 464

decay, 3

decline, 275, 309, 372, 398

evil effects, 275, 276, 278

standards, 31, 291

morality autonomous, 17 Catholic, 16

personal interest, 380, 388, 390, 391, 394, 395 sexual, 20 multiculturalism, 427 Muslims, 50, 285, 329, 361, 421, 423, 424, 425, 428, 431 national character, 79, 113, 368 network access to social, 405 closed, 148 family, 276 informal social, 256 local, 124 open, 125, 148, 151, 154, 155 norms traditional, 3, 311 obedience, 53, 102, 231, 454 Orthodox, 35, 37, 80–82, 84, 195, 315, 319, 320, 326, 327, 329, 346, 347, 350, 352, 355, 358, 359, 361, 364, 367, 369, 377, 381, 382, 389 participation culture, 239, 241–243 economic, 440 leisure, 224, 225, 234–236, 239, 244 network, 152 political, 217, 219, 224, 235 social, 218, 235, 260 path dependency, 6, 7 period effect, 40 permissiveness divorce, 342, 366, 369 personal-sexual, 46 perseverance, 53, 102, 106 personal interest, 124, 300, 380, 388, 390, 391, 394, 395 pessimists, 275, 277, 278, 306 physical capital, 109, 135, 136, 161, 163, 164, 167, 168, 404 pluralism political, 73, 186, 219 religious, 263, 372, 376, 377, 380, 382, 383, 385–387, 390, 391, 393, 394, 396, 397, 399 segmented, 376 political action, 54, 224, 226, 229, 233–235, 243–245, 250, 455

autonomy, 79

differences, 31

discussions, 233, 234, 242, 245

diversity, 227

ideology, 6, 436

integration, 4

SUBJECT INDEX interest, 219, 224, 230–234, 244 involvement, 178, 218, 234, 235, 239, 241, 244, 245, 260, 266, 269, 273 left, 285, 299, 304 parties, 131, 169 performance, 9 right, 75, 227, 304 system, 56, 57, 69, 185, 186, 230, 234, 235, 257, 259, 263, 265, 423, 450 union, 413 politics discussing, 224, 226, 227, 244, 245 post-communist, 68, 185–189, 193, 344 postmaterialism, 33, 46, 55, 151, 170, 172, 175, 226, 227 postmodern, 228, 230, 233, 244, 436, 455 values, 436 postmodernization, 226, 246, 336 poverty, 29, 51, 120, 143, 186, 250, 344, 441, 442, 444 power distance, 226, 227, 230, 231, 257 privatization, 186, 195, 200, 205, 209, 213, 276 property rights, 404, 411, 417 protest proneness, 30, 46, 54 Protestant, 35, 37, 60, 84, 227–229, 233, 244, 299, 312, 315, 319, 320, 326, 327, 329, 350, 352, 362, 372, 373, 376, 377, 382, 385, 387, 388, 393, 397, 427, 448 Protestantism, 82, 128, 192, 346, 374 rationalization, 74, 372, 473 reciprocity, 219, 259–261, 303, 408 Reformation, 60, 80, 82, 85, 372 regime familializing, 341 intermediate, 345, 346, 352, 355, 364, 366 relationship church-state, 373 relativism, 427, 455 religion institutional, 10, 372, 374, 375, 396 private matter, 427 significance of, 371, 373 religiosity, 31, 35, 46, 52, 314, 324, 372, 379, 383 personal, 379 religious conviction, 372, 426, 427, 429 decline, 372, 374 profile, 381 values, 1, 41, 47, 433, 435 responsibility, 17, 29, 53, 55, 60, 64, 88, 96, 102, 103, 105, 106–109, 130, 198, 231, 276, 277, 290, 299, 349, 412, 418, 472–475

491

revolution French, 75, 79 industrial, 74 sex-role, 10, 311–316, 318, 321, 322, 324–328, 330, 331, 333–335, 337 Velvet, 7, 185 risk, 63, 276, 299, 344, 405, 412, 440–442 sharing, 405, 412 rites of passage, 379 role male breadwinner, 328 non-specialized, 311 Roman conquest, 470 Empire, 18, 72, 73, 81 rule of law, 78, 185, 195, 200, 411, 417, 424, 428 secularization, 3, 15, 18, 19, 63, 75, 304, 312, 371, 374, 375, 383, 384, 389, 396, 397, 398, 463 security, 56, 88, 171, 284, 293, 308, 374, 378, 435, 436, 440, 443, 452 existential, 33, 34, 374, 378 self-determination, 380, 388, 392, 393, 457 self-interest, 31, 47, 69, 84, 208, 304, 408, 409 selfishness, 84, 277 social approval, 409 bonds, 217, 259 capability, 98, 120, 440–443, 446 contract, 452, 453, 456 interaction, 120, 148, 151–156, 158, 162, 163, 166, 169, 183, 219, 260, 440 social capital bonding, 148, 150, 152, 163, 170, 173, 174, 177, 246, 414, 417, 418 bridging, 148, 150, 152, 157, 159, 161, 166, 169, 171– 178, 260, 410, 414 Social Europe, 11, 215, 433, 448 socialism, 74, 279, 473 society civil, 9, 89, 115, 200, 217, 218, 220, 221, 223, 226, 227, 229, 230, 232–234, 238, 240, 243, 244–247, 260, 261, 269, 276–279, 284, 301, 309, 418 inclusive, 439, 442, 446 democratic, 263, 264, 438 multicultural, 427 patchwork, 427 solidarity informal, 9, 275, 279–283, 285–290, 294–296,298,299,300–303,306–308 sovereignty, 67, 75, 76, 451, 458–460, 461

492

SUBJECT INDEX

state formation, 81 intervention, 200, 209–212, 276, 281 of nations, 453, 459 welfare-capitalist, 441 stratification, 47, 279, 343 subjective well-being, 272, 437, 441, 446 subsidiarity, 263, 413, 434, 460 substitution hypothesis, 279–281, 287, 288, 290, 291, 296, 301, 302, 306, 307 sustainability, 61, 91 Theodor Heuss, 69 theory cultural norm, 341 game, 409, 410 middle range, 6 modernization, 6, 190–194, 212, 436 structural, 341 thrift, 53, 102, 231, 228, 233, 267, 268, 271, 340, 367, 376, 377, 413, 421, 422, 429, 430 tolerance, 10, 46, 53, 74, 82, 84, 228, 233, 267, 268, 271, 340, 367, 376, 377, 413, 421, 422, 429, 430 conditional, 266 ethnic groups, 46 qualified, 267, 272, 273 totalitarianism, 473 tradition, 6, 35, 47, 61, 67, 71, 80, 124, 150, 223, 247, 279, 310, 344, 374, 412 Christian, 47, 373 Judeo-Christian, 16, 474 religious, 18, 37, 228, 345, 373, 375, 476 transaction costs, 122–124, 150, 218, 404, 440 transparency, 285, 413 index, 285 Treaty Maastricht, 68, 411 of Rome, 67, 78 trust in institutions, 64, 284, 294, 301 intra-firm, 405, 413 lack of, 20, 64, 404 trustworthiness, 155, 219, 304, 406, 408, 409, 411, 416 truth, 82, 293, 471–473 uncertainty avoidance, 226, 228 unchurched, 377, 378, 381, 382, 385, 386, 389, 397 unemployment, 55, 103, 278, 285, 287, 302, 309, 328, 441, 442, 450 United States of Europe, 11, 77

urbanization, 75, 97, 380, 384, 386, 387, 396 utility, 6, 152, 153, 159, 162, 408, 409 function, 152, 153 theory, 6 value adjustment, 315, 316, 328, 331 change, 5, 189, 191, 192, 309, 311, 312, 313, 316, 334, 435 orientations, 4, 7, 9, 192, 256, 258, 265, 375, 449 patterns, 3, 4, 7, 8, 85, 99, 190, 192, 256, 263, 271, 273, 424 profiles, 5, 24, 39, 40 relativism, 273 structures, 433, 435, 436, 437, 446, 447 types, 228 values European, 8, 13, 15–17, 59, 60, 61, 67, 81, 83, 87, 93, 271, 272, 401, 403, 415, 474, 475 intrinsic, 407 moral, 4, 10, 11, 16, 411, 444, 470 prime guidelines, 4 public, 340, 367, 417, 418 survival, 229, 436 well-being, 228, 229, 233, 244 violence, 269, 423, 427, 453, 457, 458, 465 volunteering, 142, 217, 219, 222–224, 246 welfare regime, 9, 10, 282, 343, 344–346, 351, 358, 359, 366–369, 436–439, 441– 445 welfare state Anglo-Saxon, 279, 283 Scandinavian, 279, 344 social-democratic, 280, 283, 441, 442 Southern European, 279, 287, 289 well-being, 74, 88, 147, 148, 227–230, 233, 244, 261, 439–442 women economic role, 318, 319, 332 right for, 36, 46 work, 30, 325, 312, 332, 334 work ethic, 278 unpaid voluntary, 130, 133, 138