Handbook of Quality of Life in Enlargement Europe

Handbook of Quality of Life in the Enlarged European Union

Recent enlargement to the east made the European Union a more diverse social space and brought it into more direct contact with the social and cultural aftermath of communism. Sound empirical knowledge on heterogeneity and homogeneity in European societies after the EU enlargement is lacking. By bringing together a collection of informative analyses of key domains of social life in the new member states and candidate countries, viewed in comparison both to each other and to the ‘old’ EU-15, this handbook will help social scientists, policy-makers and other observers cope with the unfamiliarity of this new world. In particular, it examines the implications of the new member states’ membership for the future course of EU integration. This substantial text contains 17 chapters with a focus on social conditions, such as: • • • •

poverty and living conditions; social inclusion and life satisfaction; work and labour markets; family and housing.

Making use of a range of data, this handbook will be an essential resource for undergraduate and postgraduate students and researchers of Sociology, Social Policy and Welfare, European Studies and European Union Policy. Jens Alber is Professor of Sociology at the Free University Berlin and Director of the Unit ‘Inequality and social integration’ at the Social Science Center, Berlin, which chaired the international consortium that analyzed the European Quality of Life Survey. He is the author of numerous books and articles in the field of comparative social policy and political sociology. Tony Fahey is Professor of Social Policy at University College Dublin. He has published extensively on the family, religion, demography, the elderly, housing and various aspects of social policy. Chiara Saraceno is Professor of Sociology at the University of Turin, Italy, and Research Professor at the Social Science Center, Berlin. She has written extensively on family changes and family policies, on poverty and social policies, and on gender and women’s issues.

Praise for Handbook of Quality of Life in the Enlarged European Union A superb collection that delivers accessible, balanced and penetrating analyses of social conditions among the old and new member states – Alber, Fahey and Saraceno have created an indispensable reference for social scientists, policymakers and students concerned about the quality of life in the enlarged European Union. Neil Gilbert is Chernin Professor of Social Welfare at the University of California, Berkeley. This Handbook will prove invaluable to all who want to understand the social realities of the enlarged European Union. It is an excellent source of data and analysis for those who want to compare and contrast the social situation in the Member States that joined in 2004 and 2007 as against the much more comprehensively researched EU-15. It will prove a “must read” for all scholars and students of social trends in European societies and is essential background for anyone wanting to understand the depth of the future challenges of integration and cohesion in the EU-27. Roger Liddle worked for seven years as European adviser to Tony Blair and recently, as a prinicipal adviser in the think tank of European Commission President, Jose Manuel Barroso. He is now Vice Chair of the progressive international think tank, Policy Network, and a Visiting Fellow at the London School of Economics.

Handbook of Quality of Life in the Enlarged European Union

Edited by Jens Alber, Tony Fahey and Chiara Saraceno

First published 2008 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 Routledge is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2007. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” © 2008 Jens Alber, Tony Fahey and Chiara Saraceno Typeset in Sabon by Keyword Group Ltd Printed and bound in Great Britain by TJ International Ltd, Padstow, Cornwall All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Handbook of quality of life in the enlarged European Union / edited by Jens Alber, Tony Fahey, and Chiara Saraceno. p. cm. 1. Quality of life—European Union countries. 2. Social indicators— European Union countries. 3. European Union countries—Social conditions—21st century. 4. European Union countries—Economic conditions—21st century. 5. European Union. I. Alber, Jens. II. Fahey, Tony. III. Saraceno, Chiara. HN374.H35 2007 306.3072′04—dc22 2007017233 ISBN 0-203-93630-2 Master e-book ISBN

ISBN 10: 0–415–42467–4 (hbk) ISBN 10: 0–203–93630–2 (ebk) ISBN 13: 978–0–415–42467–7 (hbk) ISBN 13: 978–0–203–93630–6 (ebk)

Contents

Figures Tables Contributors Preface

vii xi xiv xv

ROBERT ANDERSON

Introduction: EU enlargement and quality of life: the context and purpose of this book

1

J E N S A L B E R , T O N Y FA H E Y A N D C H I A R A S A R A C E N O

PART I

Fertility, families and households

25

1

27

Fertility patterns and aspirations in Europe T O N Y FA H E Y

2

Patterns of family living in the enlarged EU

47

CHIARA SARACENO

3

Is there a generational cleavage in Europe? Age-specific perceptions of elderly care and of the pension system

73

WOLFGANG KECK AND AGNES BLOME

4

Family policy patterns in the enlarged EU

100

THOMAS BAHLE

PART II

Employment and working conditions

127

5

129

Employment patterns in the enlarged EU JENS ALBER

6

Working conditions and quality of work: a comparison of Eastern and Western Europe C L A I R E WA L L A C E A N D F L O R I A N P I C H L E R

162

vi 7

Contents Extension through dilution? European integration, enlargement and labour institutions

175

JELLE VISSER

PART III

Material living conditions 8

Poverty, deprivation and economic vulnerability in the enlarged EU

199 201

C H R I S T O P H E R T. W H E L A N A N D B E RT R A N D M A Iˆ T R E

9

Minimum income policies in old and new member states

218

B E A C A N T I L L O N , N ATA S C H A VA N M E C H E L E N A N D B E R N D S C H U LT E

10

Housing conditions

235

´ SKI H E N RY K D O M A N

11

Institutional drivers of housing inequalities in the enlarged EU

254

MICHELLE NORRIS

PART IV

Social capital and social cohesion

277

12

279

Patterns of sociability in the enlarged EU M A N U E L A O L A G N E R O , PA O L A T O R R I O N I A N D C H I A R A S A R A C E N O

13

Feeling left out: patterns of social integration and exclusion

304

PETRA BÖHNKE

14

The perception of group conflicts: different challenges for social cohesion in new and old member states

328

JAN DELHEY AND WOLFGANG KECK

PART V

Processes of Europeanisation

353

15

355

Migration and mobility culture: an analysis of mobility intentions H U B E RT K R I E G E R

16

Where we stand in Europe: citizen perceptions of European country rankings and their influence on subjective well-being

385

JAN DELHEY AND ULRICH KOHLER

17

Assessing the quality of European surveys: towards an open method of coordination for survey data

405

ULRICH KOHLER

Index

425

Figures

1.1 2.1 2.2 2.3 2.4 2.5 3.1

3.2 3.3 3.4 3.5 3.6 3.7 3.8 4.1 4.2 4.3 4.4 4.5 4.6 4.7 5.1 5.2

Relationship between actual number of children and personal ideal number 38 of children among women with completed fertility (ages 40–64) Household status of young men (18–34): three distinct patterns 57 Incidence of unemployment among European young singles living 58 with their parents, by country Households of elderly people: differences between men and 60 women (65 and over) in the three clusters Households of 35- to 64-year-olds: differences between the 35–49 63 and the 50–64 age groups, by gender and country cluster Distribution of working-age couples, by working status in Europe 66 Share of individuals who care for elderly persons within the 78 household and preference of working-age family members as care givers for own parents Share of individuals in different age groups who opt for family 79 support as the preferred care solution, by country Difference between men and women according to their preference 81 for family care Family orientation in care issues and public care efforts, by country 82 Preferences on care financing, by country 84 Trust in pension system, by country 89 Trust in pension system, by country and age groups 90 Perception of tensions between young and old, by level of trust in 94 public pension system Synopsis 1: Conceptual map of family policy forms in Europe 103 Structure of social expenditure on family benefits, Europe 2004 108 Variations in family allowances, Europe 2004 111 Maternity and parenting benefits, Europe 2004 112 Childcare services and parenting policies, Europe around 2000 115 Synopsis 2: Variations in family policies – a ‘cultural map’ 116 of Europe Family policy problem constellation, around 2000 118 Development of economic output during the Great Depression 131 and post-communist transformation Employment rates of the population aged 15–59 in post-communist 131 member states and in Eastern European countries outside the EU, 1989–2004

viii 5.3

Figures

Employment rates in consumption-related and production-related services, 2003 5.4 Development of real GDP and employment rate in post-communist NMS-8 5.5 Economic growth rates and change in employment rate in EU-25, 2000–2004 5.6 Development of male and female employment rates in Europe and the United States, 1979–2003 (% of population aged 15–64) 5.7 Employment rates of older (55–64) and younger (15–24) age cohorts in EU-25, 2004 5.8 Employment rates of people with high and low educational attainment in EU-25, 2004 (population aged 25–64) 5.9 Social expenditure ratio and employment levels in EU-25, 2004 5.10 Level of social assistance benefits and low-skill employment in EU countries, 2004 5.11 Level of unemployment insurance benefits and employment rate, 2004 5.12 Burden of social security contributions and low-skill employment in EU countries, 2004 6.1 Perceptions of pay: gender and occupational differences across Eastern Europe 6.2 Perceived job security: gender and occupational differences across Eastern Europe 6.3 Perceived intrinsic rewards: gender and occupational differences across Eastern Europe 6.4 Job satisfaction: gender and occupational differences across Eastern Europe 7.1 Bargaining coverage and union density 7.2 Earnings inequality and bargaining coverage 7.3 Union presence, 2003–2004 8.1 Median monthly household equivalised incomes in purchasing power standards, by country 8.2 Median monthly household income, by income quartile across countries (in purchasing power standards) 8.3 Mean deprivation (10 items) across countries 8.4 Mean deprivation (10 items), by income quartile, by country 8.5 Share of persons in poverty at 60% median equivalised household income, 2003 8.6 Vulnerable class size, by economic clusters 8.7 Conditional probabilities for latent class models for income deprivation and economic stress 9.1 Expenditure in % of GDP, social exclusion (NEC), means-tested cash benefits, 2003 9.2 Monthly net disposable income of social assistance recipients 224, (working age), 2004 9.3 Cross-country correlations between net incomes of various household types on social assistance, 2004

143 145 146 148 148 149 150 152 153 155 166 168 168 171 181 183 191 204 204 205 206 207 212 212 221 225 227

Figures

ix

Net disposable income of social assistance recipients (working age, average of four family types), 2004 10.1 Tenure status according to accommodation, by country grouping 10.2 Relationship between GDP per capita and % of ownership with mortgage 10.3 Relationship between GDP per capita and index of housing quality 11.1 Number of dwellings and housing output per 1,000 inhabitants in European countries, 2000 11.2 Ratio of outstanding mortgage debt to GDP in European countries and annual growth to mortgage debt, 2004 11.3 % of dwellings in high-rise buildings in European countries, 2004 12.1 Distribution of countries along two dimensions of the secondary and tertiary sphere of sociability 12.2 Patterns of sociability in Europe (country clusters) 12.3 Distribution of sociability clusters with regard to selected structural indicators 13.1 The prevalence of belonging and marginalisation in the EU (mean index values) 13.2 Mean index value of belonging in three population groups 13.3 Perceptions of belonging and the availability of social network support (mean index values) 13.4a Social support outside the household as a buffer? 13.4b Family back-up as a buffer? 13.5 The impact of long-term poverty on perceived social exclusion across countries according to their welfare level 14.1 Impact of perceived tensions between rich and poor people on social trust 14.2 Intensity of group tensions as perceived by the population 14.3 Intensity of perceived vertical and ethnic tensions (country clusters) 14.4 Cross-national correlates of vertical tension perception 338, 14.5 Cross-national correlates of ethnic tension perception 341, 15.1–15.4 Stock of macro-indicators and change in intended 375, migration, 2001–2005, for EU-25 15.5–15.8 Change in macro-indicators and change in intended 377, migration, 2001–2005, for EU-25 15.9–15.10 Stock of subjective indicators, 2001, and change in intended migration, 2001–2005, for EU-25 16.1 Country–EU comparisons (share of missing answers, by survey country) 16.2 Outcomes of country–EU comparisons 16.3 Cross-check between reality and perceptions (economic situation) 16.4 Cross-check between reality and perceptions (employment situation) 16.5 Cross-check between reality and perceptions (overall quality of life) 16.6 Impact of country–EU comparison on life satisfaction 17.1 Components of the sampling process, by country and survey programme

228

9.4

236 238 246 262 263 267 289 291 293 308 312 316 318 318 324 331 333 335 339 342 376 378 380 390 393 395 396 397 400 412

x

Figures

17.2 17.3

Proportions of women (differences between survey and official sources) Proportions of women among gender heterogeneous couples

415 419

Tables

1.1 The ideal–actual fertility gap in three broad age groups of women in European countries, 2001–2002 1.2 Mean personal ideal number of children and actual number of children by school-leaving age among women with completed fertility (ages 40–64) in the EU-15, NMs and CC-3 1.3 Fertility ideal attainment and school-leaving age among women with completed fertility (ages 40–64) 1.4 Mean general ideal family size 1.5 Fulfilment of general ideal number of children among women with completed fertility (ages 45–64), pooled data for 8 European societies, 1981, 1990, 2001 A1.1 Sample Ns for female samples by countries and age groups, combined Eurobarometer dataset, 2001–2002 2.1 Household patterns in an enlarged Europe 2.2 Household statuses of the young in Europe (ages 18–34) 2.3 Elderly men and women (65 and over) living with their children, by patterns of family formation by the young 2.4 Use of time among working Europeans 3.1 Age-group differences in perceived financing responsibility for care 3.2 Logistic regression on tension perception by age and trust in pension scheme 4.1 Social expenditure on family benefits, Europe 2004 4.2 Eligibility for family allowances and variations in benefit rates, Europe 2004 4.3 Childcare coverage rates in Europe, around 2000 5.1 Indicators of employment development in the post-communist new member states 5.2 Recent employment records in EU-25, employment levels, 2004, and changes, 2000–2004 5.3 Measures of labour market inclusion and exclusion, 2004 5.4 Sector-specific employment rates in the enlarged EU A5.1 Possible determinants of employment development in EU-25 6.1 Working conditions in Eastern Europe I (working hours, unemployment, wages, job security, pay, intrinsic rewards and job satisfaction, means and standard deviations)

34, 35 39

40 41 42

44 51 54, 55 61 68 86 95 107 110 114 132 134 136 141 156 165

xii 6.2

Tables

Working conditions in Eastern Europe II (workload, tight deadlines, dangerous and unhealthy work environment, autonomy and influence, future prospects of promotion and career and job satisfaction, means and standard deviations) 7.1 Key changes in employment, unemployment and contracts, 1998–2004 7.2 Collective bargaining (private sector) and wage setting, recent years 7.3 Statutory minimum wages in Europe 9.1 Statutory adjustment mechanisms relating to social assistance benefit standards in EU countries, 2004 10.1 Housing tenure 10.2 Persons living in own homes, by age, income, area of residence and occupational status 10.3 Average floor space 10.4 Number of persons per room, by ownership status, income, region and age 10.5 Households that reported deficits in accommodation 10.6 Households that do not have an indoor toilet, by age of respondent and area of residence 10.7 Good housing (having at least one room per person and perceiving none of five housing deficits 10.8 Complaints about environment 10.9 Relationships between satisfaction with accommodation and general satisfaction with life and objective characteristics of housing 10.10 Relationships between trust and objective characteristics of housing 11.1 Occupied dwellings, by tenure, in European countries, 1980, 1990, 2000 11.2 Sources of funding for new construction in seven longstanding EU member states, 1994–1995 11.3 Age distribution of the housing stock in European countries, various years 11.4 Institutional drivers of housing inequalities in an enlarged Europe 12.1 Dimensions of sociability, by country clusters 12.2 The secondary sphere of sociability (type of support and 283, contacts), by country 12.3 The tertiary sphere of sociability (activity in political and 286, associative domains) 12.4a Involvement in public sociability, by cross-cluster 12.4b Involvement in private sociability, by country cluster 13.1 Perceived social exclusion, mean index value and distribution of 309, single indicators per country (agree and strongly agree summarised) 13.2 What determines perceived marginalisation and belonging? 313, 13.3 Socio-economic precariousness and social support as 320, determinants of belonging 13.4 The independent effect of country characteristics on the individual experience of social exclusion 14.1 Vertical tension perception in different social groups

170

177 180 185 223 237 239 240 242 243 244 245 248 250 251 256 261 265 272 282 284 287 294 295 310

314 321 322 343

Tables

xiii

14.2 Ethnic tension perception in different social groups 345 A14.1 Documentation of indicators 348 15.1 Mobility intentions in EU member states, 2001 and 2005 359 (population aged 18–64) 15.2 Mobility intentions (regression models for 2005) 362–364 15.3 Mobility intentions (regression models for 2001) 367–369 15.4 Comparison of the effect of macro-economic and subjective 381 conditions on the dynamic intention to migrate between EU-25 and EU-15 16.1 Explaining difficulties of rating one’s own country (coefficients 392 of logistic regressions of missing answer vs. non-missing answer) 16.2 Explaining personal life satisfaction: the influence of economic 400, 401 country–EU comparisons with various control variables 17.1 Target population, number of countries and number of 407 observations, by survey programme 17.2 Minimum, average and maximum response rate, by survey 409 programme 17.3 Number of countries for which complete information about the 411 specified component of the sampling process is available, by survey programme 17.4 Fraction of countries with best available sampling process, 413 by survey programme 17.5 Mean difference from the true value and fraction of countries 417 with values outside the confidence bounds, by survey programme 17.6 Mean difference from the true value and fraction of countries with values outside the confidence bounds among gender heterogeneous couples, by survey programme 420 17.7 Overall sampling quality of survey programmes 421

Contributors

Jens Alber, Social Science Research Center Berlin, Germany Robert Anderson, European Foundation for the Improvement of Living and Working Conditions, Dublin Thomas Bahle, University of Mannheim, Germany Agnes Blome, Social Science Research Center Berlin, Germany Petra Böhnke, Social Science Research Center Berlin, Germany Bea Cantillon, University of Antwerpen, Belgium Jan Delhey, Jacobs University Bremen, Germany Henryk Doman´ski, University of Warsaw and Polish Academy of Sciences, Poland Tony Fahey, University College Dublin, Republic of Ireland Wolfgang Keck, Social Science Research Center Berlin, Germany Ulrich Kohler, Social Science Research Center Berlin, Germany Hubert Krieger, European Foundation for the Improvement of Living and Working Conditions, Dublin Bertrand Maître, Economic and Social Research Institute, Dublin Michelle Norris, University College Dublin, Republic of Ireland Manuela Olagnero, University of Turin, Italy Florian Pichler, University of Aberdeen, UK Chiara Saraceno, University of Turin, Italy, and Social Science Research Center Berlin, Germany Bernd Schulte, Max Planck Institute for Foreign and International Social Law, Germany Paola Torrioni, University of Turin, Italy Natascha Van Mechelen, University of Antwerpen, Belgium Jelle Visser, University of Amsterdam, Netherlands Claire Wallace, University of Aberdeen, UK Christopher T. Whelan, Economic and Social Research Institute, Dublin

Preface

For the 2007 Spring European Council the European Commission presented a first stocktaking of social reality in an enlarged Europe. This report argued that there is a lack of consensus on the common social challenges facing Europeans, and that there is a need for better analysis and understanding of the social situation. To support debate on the social issues and challenges facing Europe, the Bureau of European Policy Advisers (a Directorate General of the European Commission) has issued a consultation paper, which begins: How can the social well-being of all Europe’s citizens be best advanced within a globalising world? This question should be at the heart of everything the EU and its Member States do. Public policy imperatives, such as “Growth and Jobs”, the Lisbon strategy, and the drive for greater competitiveness are not ends in themselves – but means to an end – the well-being of European citizens. (Liddle and Lerais 2007) The challenges arising from social exclusion, an ageing population, changing family structures and gender roles, and now enlargement, have pushed quality of life issues to the fore in the EU policy debate. Their impact is direct on people’s everyday lives, families, communities and society. The recent work of the European Foundation for the Improvement of Living and Working Conditions has thus focused on monitoring trends and changes in living conditions and quality of life across the EU and in candidate countries. ‘Living conditions’ clearly embraces a very wide area of policy interest, with a particular need to map and understand disparities associated with age, gender, health, ethnicity and region. The Foundation’s four-year programme points to the need to link the assessment of living conditions to the changing nature of employment, work organisation, and working conditions on the one hand, and to the modernisation of social protection and social welfare services on the other. Quality of life for Europe’s population is increasingly at the centre of the Foundation’s work. The enlargement of the EU to incorporate twelve new member states over the past four years has increased not only size and population, but the diversity of people, lifestyles and culture in Europe. This diversity is undoubtedly enriching daily lives but, as with other developments in economy and employment, not for all European citizens. The flipside of diversity is inequality, which is evident between member states, but often as much or more within countries and regions. The European institutions have a range of policies and programmes that impact on key ‘quality of life’ issues: employment conditions; health and safety; social inclusion;

xvi

Preface

mobility; equal opportunities. The elaboration of policy responses to established and emerging social challenges will depend upon information, but more so on insight and understanding of the living conditions and experiences of people in the EU. Appropriate measures will demand intelligence not only on objective conditions or the social situation, but regarding how people feel about these conditions, their concerns and priorities. Monitoring and analysis should cover key domains that influence quality of life – income, health, family – but should also examine how people assess the quality of the society in which they live, including conflicts and tensions in society, the quality of the environment and their satisfaction with services of general interest. This volume takes forward the analysis and interpretation of the complex social challenges in an enlarged Europe. I welcome its contribution to the debate on quality of life in Europe and believe it provides an informed basis for both further research and the identification of priorities for public policy. Robert Anderson European Foundation for the Improvement of Living and Working Conditions

Reference Liddle, R. and Lerais, F. (for the European Commission) (2007) ‘Europe’s social reality: a consultation paper from the Bureau of European Policy Advisers’. Online. Available (accessed March 2007).

Introduction: EU enlargement and quality of life The context and purpose of this book Jens Alber, Tony Fahey and Chiara Saraceno Introduction The fall of the Berlin wall in 1989 and the subsequent collapse of communism were a major advance in healing the divisions of Europe left after World War II. The continent was no longer at war with itself, either in the form of the hot war that ended in 1945 or the cold war that endured ten times longer. The recent eastern enlargements of the European Union decisively advanced this process of reconnection. Eight former communist states joined the Union in 2004 (along with the small Mediterranean states of Cyprus and Malta) and two more – Bulgaria and Romania – joined in 2007. This development consolidated the move to democracy and the freemarket economy in a whole swathe of the continent where these possibilities had long been snuffed out. It also enhanced the long-term prospect of peace within Europe and strengthened the EU’s capacity to hold its place in the world. In view of these gains, the eastern enlargement could be ranked among the EU’s greatest achievement so far. It is also notable for having been brought about by means of an inter-state machinery, embodied in the institutions of the EU, that has proved surprisingly effective and has consistently defied predictions that it would collapse under the weight of its own ungainliness. While the EU cannot claim the credit for bringing communism to an end, it can claim to be a central mechanism by means of which European states have been enabled to come to terms with both the possibilities and the stresses of the post-communist era. Great though these gains have been, their very ambition has stretched the political and institutional capacities of the EU to their limits, as is sketched out later in this chapter. How well the Union will cope with these strains is partly a matter of macrolevel factors such as the evolution of relationships between member states, the adaptation of EU institutions and the impact of macro-economic forces. However, it also depends on the micro-level impact of the EU on the daily lives of its people. The geo-political significance of EU integration will not of itself ‘sell’ the project to ordinary citizens. It must also deliver improvements in their day-to-day well-being and quality of life – or at least not be seen as a threat to what the various national populations have already achieved in these areas. The eastern enlargement has a particular significance in this regard. For people in the new member states (NMS), the hope of improvements in economic opportunities and standards of living was a major motivation for joining the EU in the first place. For people in the incumbent states, the very poverty of the new members and the fear that they would drag down living conditions and social cohesion in the richer parts

2

Jens Alber, Tony Fahey and Chiara Saraceno

of Europe, through either low-cost competition or floods of migrants, gave rise to concerns that enlargement may have gone too far, too fast. The candidate countries that have been listed as possible members in the future – of which Turkey is the most significant – accentuate these contrasts further, as they promise to add further to the already high level of diversity and inequality across countries in the EU. The hopes of the new members and the worries of the incumbents thus place quality of life at the centre of debate about the future of the European project. The ‘quality of life’ perspective on people’s living conditions, as we describe further below, aims to broaden the scope of what is considered beyond the economic to include a range of social domains of life, such as family life, the social dimensions of work, neighbourhood and community, housing and so on. It also, crucially, is not just concerned with objective indicators but also with how people themselves view their circumstances and evaluate the quality of their lives. The usefulness of this broader view is that it approximates more closely to how people experience and react to their own situation. It thus taps into what are likely to be core concerns of the EU as it seeks to deliver benefits that will establish its appeal and legitimacy among citizens. The concern that motivates this book is that, important though these issues may be for the future of the EU project, knowledge of quality of life in the European Union is seriously incomplete, particularly with regard to the situation in the NMS and candidate countries. Social reporting has a long tradition in Western Europe and has been built upon by the EU to develop wide-ranging data and analyses on social conditions (for example, through the European Community Household Panel survey, which ran from 1994 to 2002). However, this work has largely been confined to the EU-15 or even to sub-sets of older member states within the 15. The NMS and candidate countries lack a similar tradition, particularly at the level of a cross-national picture that would enable one to view these countries alongside each other rather than in isolation. As the eastern enlargement took place, therefore, assessment of living conditions in the NMS and candidate countries relied on a narrow range of economic and demographic indicators, such as GDP per capita, labour market measures and mortality data. Valuable though these are, they do not amount to the multidimensional approach required to capture broader quality of life. A further relevant gap in knowledge concerns the institutional context within which various domains of social life function. These contexts are a particular concern in former communist countries since despite the transition to democracy and the free-market economy, an institutional legacy of the old regime survives in many areas of life. Thus, for example, housing conditions in these countries are not only of distinctive quality (usually in the direction of being quite poor) but also are the product of distinctive housing systems, which in turn are the outcome of institutional histories unlike anything experienced in the west. Quite often, in areas such as these, the details of day-to-day living conditions are difficult to make sense of unless one also understands something of government policies, state regulation, market mechanisms and civil society institutions that together make up the institutional framework within which they function. Most, if not all, welfare state typologies, which are used to understand national differences in living conditions and behaviours, have been developed looking at Western Europe. We still lack a similar comprehensive exercise including also the formerly socialist countries. The purpose of this book, which consists of a series of chapters written by authors who are experts in their field, is to try to fill these gaps in knowledge on quality of

Introduction

3

life in the EU, with particular reference to the NMS and candidate countries, and to draw some implications on the future of the EU project. In later sections of this chapter, we set out in more detail what this purpose entails and how it is fulfilled in the book. We first set the scene by considering a number of macro-level consequences of the enlargement process for the EU.

1. Background: an enlargement crisis? The accession of 12 new member states has profoundly transformed the nature of the European Union. The Union’s official motto as specified in Article I-8 of the Treaty establishing a Constitution for Europe – ‘United in diversity’ – has since 2004 certainly become more true with respect to diversity. But how much unity still remains after enlargement? Will Europe grow together or will a lasting cleavage between old and new member states put a stop to further integration? For a number of reasons, many policy pundits expect the enlargement process to lead to a lasting chasm and to the hitherto most serious challenge to the unity and cohesion of the European Union. First of all, there are huge differences in economic wealth as well as in population size. Adding roughly 74 million people, the 2004 enlargement augmented the number of EU member states by 66 per cent, increased the population size of the Union by roughly 20 per cent, while raising total economic output as measured by the Union’s Gross Domestic Product by a mere 5 per cent. Even when measured in purchasing power parities, the GDP per head in the ten new member states at the time of accession was less than one half of the EU-15 average. The entry of Bulgaria and Romania in 2007 added another 30 million citizens (+ 6 per cent), but an increase of total economic output of only less than 1 per cent (Berié and Kobert 2006: 575). Among the new member states, only four – Poland, Romania, Hungary, and the Czech Republic – have double-digit population sizes. Bulgaria and Slovakia are the only other countries to exceed a population size of 5 million. The other six newcomers are small with population sizes between 400,000 (Malta) and 3.4 million (Lithuania). The addition of several countries that combine demographic smallness with economic poverty is bound to exacerbate the always difficult task of the European Union to strike a balance between the equality of all sovereign member states on the one hand and their different economic and demographic weight on the other. Second, the accession process itself is likely to leave lasting bad memories. During the accession negotiations, several representatives of the aspiring countries complained about a marked power asymmetry in the accession process (Heidenreich 2006). Experiencing the accession criteria as an encroachment upon their national sovereignty, they felt they were being asked to meet expectations that not all of the established members of the club fulfilled themselves. Such concerns were repeatedly and perhaps most forcefully voiced by the Czech president and former Prime Minister Vaclav Klaus, but also by Polish politicians. Even after accession, the new member states were not put on a par with the old members. Due to various transitional arrangements, basic rights and freedoms will only be fully extended to NMS citizens after the end of various transitional periods (e.g. 2016 for the free movement of capital, 2011 for the free movement of persons, 2008 for the free movement of goods). The farmers of the new member states were initially entitled to only 25 per cent of the direct payments granted to their peers in old member states, and they will reach

4

Jens Alber, Tony Fahey and Chiara Saraceno

equal status only gradually up to 2013 (von Baratta 2003: 119). Whereas the farmers of France (21 per cent), Spain (14 per cent), Germany (13.5 per cent), and Italy (11 per cent) received roughly 60 per cent of the payments made under the Common Agricultural Policy (CAP) in 2004, the agricultural sector in the ten new member states initially had to settle with a combined 4 per cent of total expenditure (Berié and Kobert 2005: 553). Third, after many decades under the shadow of the Soviet empire, the post-communist countries seem to look more to the US than to the EU as a social and political model. This is so for a number of reasons. First, the entry of the former communist countries into the market economy and the restructuring of their social security systems have been strongly shaped by the requirements – and ideology – of the IMF and the World Bank and of the so-called Washington consensus, with its priority for rapid privatisation and lack of attention to building social and organisational capital (Stiglitz 1999a, 1999b). Second, the US welcomed these countries as members of NATO long before the European Union opened up its ranks (Ash 2004). During the Iraq war, American newspapers and government officials such as Donald Rumsfeld, then Secretary of Defence, frequently made a polemical distinction between ‘old and new Europe’. Originally coined to distinguish between proponents and opponents of the war against Iraq, the concept was soon extended to distinguish a group of supposedly dynamic countries actively facing the future from a passive and nostalgic group clinging to the past and supposedly coping only reluctantly with the modern globalised world.1 For all these reasons, many observers see the European Union in a profound crisis after enlargement (e.g. Bach 2006 has a particularly gloomy vision). It is claimed that due to the failure of the constitution, the decision-making structures which were originally designed for a small number of fairly homogenous nations do not suit a much bigger and more heterogeneous Union. For sheer reasons of size, in this view, finding consensus and making decisions in the Commission or in the European Council has become a far more difficult task now that there are 27 rather than 15 members with differing interests. In addition, the poverty and economic backwardness of the new members will strain EU Structural Funds, which were never designed to cope with this degree of regional inequality or this big an agricultural sector. Consequently, conflicts of interest between the net contributors to the EU budget – above all the Netherlands, Sweden and Germany – and the net beneficiaries – especially the new member states and Greece, Portugal and Spain – are likely to grow. In 2005, policymakers in the old member states such as the former Swedish Prime Minister Goran Persson or the former German chancellor Gerhard Schröder complained that the EU’s big net contributors were effectively subsidizing the low tax rates of their competitors in the new member states, who could afford to cut taxes because hefty EU subsidies compensated for any lost revenue (http://euractiv.com/en/taxation/ flat-tax/article-136190, accessed 31 January 2007). Of course, there are also more positive assessments of the enlargement process. First of all, the promise made to the aspiring countries that joining the European Union would lead to freedom, democracy and economic development has basically been fulfilled. In international comparisons of democratic progress or the level of corruption as measured by Freedom House (2006) or Transparency International (2005), the new member states usually score much better than other post-communist countries. Economically, the new member states have recently begun to catch up,

Introduction

5

experiencing higher growth rates than the old members. The three Baltic countries even had double-digit growth rates in some years, thus coming close to the growth rate of the Chinese economy. While the post-communist transformation put the transition countries into a much deeper economic slump than western countries had experienced during the Great Depression, all new member states have by now exceeded the material standard of living (as measured by GDP per capita) which they had reached under the communist regime. The idea that the EU institutions could not cope with the post-enlargement situation unless fundamentally overhauled proved too pessimistic. For many years the Union has already proved capable of practising a pattern of differentiated integration where selected groups of countries rather than all members opt for closer cooperation in specific policy fields. Thus, only five countries originally signed the Schengen agreement on the elimination of border controls, from which Ireland and the United Kingdom still abstain, thus leaving not only the new member states outside the scope of the agreement. The single currency was adopted by only eleven of the old member states (absent: Denmark, Greece, Sweden, the United Kingdom), and is now valid in 13 countries including Greece and Slovenia. Thus, the Union has been capable of living with different degrees of integration in different policy realms, relying on a gradual or stepwise expansion of cooperation to new participants. This underlines the notion that the European Union resembles a tricycle that can halt the process of deepening at any moment rather than a bicycle that has to keep moving forward in order not to falter or fall (Moravcsik 2004). Yet, even if one adopts a more pragmatic and less gloomy outlook, there are two problems which, having not yet been solved successfully, will require some innovative solutions, because they are exacerbated by enlargement. The first one concerns growing disunity among political elites, the second one declining compliance among taxpayers and voters. Given the divergent political interests of net contributors and net beneficiaries and vast differences in economic development, consensus building in budget negotiations has already become visibly more difficult. As pointed out most forcefully by the British Prime Minister Tony Blair (2005) during the British Presidency, the EU budget is predominantly geared to provide support for declining sectors of the economy such as the coal and steel industry at the very beginning and agriculture today, instead of focusing on investments in education, science, and research that would help to bolster European productivity and competitiveness. Indeed more than 40 per cent of the EU budget is still spent on various supports for agriculture and more than another third is spent on structural funds supporting the development of poor regions whose GDP per head is less than 75 per cent (in the case of so-called objective 1 areas entitled to subsidies), or less than 90 per cent of the EU average (in the case of the so-called cohesion fund). Even though they are only gradually being incorporated into this system of support, the demand of the new member states is particularly high, because 92 per cent of their population live in poor objective 1 regions. Overall, 36 regions in new member states and 32 regions in old member states belong to this group (Süddeutsche Zeitung 17.03.05). In order to cope with the growing demand, the European Commission had proposed to raise the EU budget to 1.24 per cent of the Union’s Gross National Income in the period 2007–2013. But the camp of net contributing countries – led by Germany, France, the United Kingdom, the Netherlands, Austria, and Sweden – insisted in the 2005 budget negotiations on limiting it to 1 per cent. A compromise

6

Jens Alber, Tony Fahey and Chiara Saraceno

was reached after long and tough negotiations in December 2005 by setting the budget ceiling at 1.045 per cent of EU output. The compromise also included a 20 per cent decrease of the UK rebate which Prime Minister Blair had agreed upon in exchange for France’s agreement to endorse a full and wide-ranging review of the budget for farm subsidies in 2008–2009. Of the total budget of E864.3 billion for the period 2007–2013 which was formally endorsed in May 2006, E157 billion or roughly 18 per cent will go into aid for the new member states (http://news.bbc.co.uk/ 2hi/Europe/4538100.stm, accessed 31 January 2007; Berié and Kobert 2006: 577). The fact that more than one-third of the EU budget goes into funds for regional policies is usually justified as an indispensable means for fostering cohesion in the European Union. In the Commission’s reports on economic and social cohesion, cohesion is understood as a synonym for the similarity or equality of standards of living across European countries and regions. From this perspective, the more similar countries and regions become with respect to the standard of living, the more social cohesion there will be in the European Union.2 In short, EU regional policy sees convergence as resulting in cohesion, and the structural funds serve to promote cohesion by redistributing resources in a way that helps to make regions more similar. Defining cohesion as a synonym for equality poses a few problems, however. From a theoretical perspective, sociologists in the tradition of Emile Durkheim tend to see growing equality also as a platform for increased competition. They relate the origins of solidarity under modern conditions to the division of labour, which makes unequal individuals mutually dependent on each other. From this perspective, cohesion is not a synonym for equality, but a concept capturing the strength of social bonds, the degree of connectedness and felt solidarity among individuals and groups in society, and the degree to which the members of a society develop shared values and a common identity. If we keep the concepts of equality and cohesion conceptually apart, it will be possible to examine the extent to which people in equal or unequal positions actually engage in social relations, share a common sense of identity and similar norms of appropriateness. The question whether equality of living conditions fosters social cohesion then becomes a matter of empirical analysis rather than a matter of definition. Once we accept that convergence of living standards is not the same as cohesion, the relevant question concerning social cohesion on the European level is to what extent Europeans have moved beyond the nation state when it comes to defining the territorial unit to which they attribute a sense of belonging, of collective identity and of shared responsibilities. In a recent book and a series of articles based on comparative data from surveys, the sociologist Jürgen Gerhards (2005; 2006) has studied these issues. His result is that there is a marked discrepancy between the official institutionalisation of a European citizenship on the one side and the ideas of ordinary Europeans on the other. The latter continue to frame citizenship rights in national terms and hesitate to extend notions of solidarity beyond the boundaries of the nation state. In this sense, there is a widening gap between the policy visions of top-level Eurocrats and ordinary citizens. One of the specific findings is that the idea of equal access to national labour markets for all Europeans anywhere in the Union has not caught on with the citizens of Europe, least of all in new member states. According to the European Values Survey, there are only four EU member states in which a majority of respondents support equal treatment for foreigners in the labour market when jobs for co-nationals are

Introduction

7

scarce (Sweden, Netherlands, Denmark and Luxembourg). In ten of the old member states large majorities of two thirds or more advocate making a distinction between their co-nationals and foreigners. In the new member states, these majorities are even larger, varying between 80 and 96 per cent in all countries bar Estonia3 (see the data in Gerhards 2006). If this is the case, redistributive attempts to produce more equality across Europe will not automatically foster more social cohesion among Europeans, but may instead invoke problems of legitimation and citizen backlash. There are several additional signs of a widening gap between the supranational structure of the European Union and the predominantly national mindsets of Europeans. First of all, voter turnout at European elections has always been much lower than in national elections. This so-called ‘Euro-gap’ (Rose 2004) is even higher – about one-third – in new member states (29 points) than in old ones (22 points Wessels 2006). In the eight post-communist new member states only 31 per cent of the citizens – compared to 53 per cent in the old EU-15 – participated in the 2004 elections to the European Parliament. Over time, average turnout at European Parliament elections has shrunk from 66 per cent in 1979 to 48 per cent in 2004 (Wessels 2006). Second, Eurobarometer time-series data on perceptions of Europe among Europeans show that identification with Europe was highest in the early 1990s, but has since been declining. The percentage of Europeans considering the EU membership ‘a good thing’ decreased by almost 20 percentage points from a peak of 72 per cent at the beginning of the 1990s to 53 per cent in autumn 2006 (European Commission 2002: 19; European Commission 2006a: 6). There is also much scepticism regarding further enlargement. In Eurobarometer 66 of autumn 2006, the percentage of respondents opposing further enlargement (42 per cent) was almost as high as the percentage of proponents (46 per cent). In five of the old member states, less than 40 per cent expressed themselves in support of further enlargements (with a variation from 36 per cent in the UK to 30 per cent in Germany, and with France, Luxembourg and Austria in between - European Commission 2006a: 29). All this reflects the fact that European integration has primarily been an elite project, driven by the war experience of political leaders and technocratic elites who sought to overcome the tradition of European Civil Wars (as Göran Therborn, 2000, characterised the two world wars of the twentieth century). Protracted postwar prosperity and security concerns during the Cold War helped the project to gain citizen consent even though effective chances for citizen participation remained very limited, as politics were designed in rather non-transparent networks of experts which largely left the people out. The common threat posed by the Soviet Union has now ceased, and the long years of prosperity and welfare state expansion have given way to sluggish growth, welfare state restructuring and increased international competition in a globalised world. As a result, further integration will presumably need actively to win citizens consent by convincingly demonstrating that concrete welfare gains are associated with the European project. As Mark Kleinman described the new political climate concisely even before the time of enlargement: ‘The time of permissive consensus to European integration among European citizens is over’ (Kleinman 2002: 214). The failure of the French and Dutch referendums on the European constitution is convincing proof of this thesis. The European Commission acknowledged the problem of declining citizen consent when it noted already in a 2001 White Paper on European Governance that ‘people

8

Jens Alber, Tony Fahey and Chiara Saraceno

increasingly distrust institutions and politics or are simply not interested in them’ and identified these problems as ‘particularly acute at the level of the European Union’. This was followed by the self-critical diagnosis that: ‘Many people are losing confidence in a poorly understood and complex system to deliver the politics that they want. The Union is often seen as remote and at the same time too intrusive’ (Commission of the European Communities 2001: 3). One of the Commission’s proposals for change was to communicate more actively with the general public on European issues and to reach out to involve regional and local stakeholders as well as civil society actors more actively into the development of policy proposals. The Commission report on European governance (2003) reiterated the importance of these objectives and described strategies for implementing the White Paper. In a Communication published in 2005 the Commission also responded to the negative votes on the European Constitution in France and the Netherlands by declaring a ‘period of reflection’ and by publishing its ‘Plan-D for Democracy, Dialogue and Debate’ (Commission of the European Communities 2005). In order to find out how Europeans conceive of the future of the European Union, a special Eurobarometer was launched in 2006. Among its many complex results, four findings are particularly worth mentioning: (1) A majority of Europeans is in favour of shifting more decision-making to the European level, particularly so with respect to the fight against terrorism (European Commission 2006b: 40). (2) The proportion who consider that things are moving in the right direction in the European Union is only 39 per cent, compared to 27 per cent who feel that things are going in the wrong direction, and 34 per cent who are undecided (2006b: 17). (3) In most policy areas, the interviewees have a critical assessment of the European Union’s performance; particularly poor grades are given for the fight against unemployment, the protection of social rights, and ensuring economic growth (ibid: 35). (4) When asked what they would consider to be most helpful for the future, most respondents declare comparable living standards as the key element for the future of Europe. Particular urgency is given to this point in the new member states, where 74 per cent of the respondents hold this opinion, compared to 47 per cent in the old member states (2006b: 37). The proponents of European integration and EU enlargement have so far often settled with rather idealistic celebrations of diffuse general progress rather than demonstrating concrete welfare gains linked to the European project. Thus, EU official documents usually refer rather vaguely to ‘productivity gains’ which the European Social Model supposedly implies, while Sandra Kalniete, the former Latvian Foreign Minister and EU Commissioner, hailed the 2004 enlargement as ‘Europe’s triumph over the twentieth century’ (Verheugen 2005). In similarly diffuse terms the Italian banker Padoa-Schioppa praised the EMU for restoring the ancestral monetary unity which Europe had once enjoyed (Kleinman 2002). Winning the future consent and support of sceptical voters who demand comparable living standards will probably require much more concrete forms of ‘output-legitimation’ (Scharpf 1999), which demonstrate that the EU is indeed a government by the people and for the people and promotes the common welfare of Europeans. This is where the social monitoring of the development of economic and social conditions in Europe will play an important role. If serious attempts at social accounting show that the further development of the EU is not only associated with continuing or increasing economic growth, but also with measurable progress in the quality of

Introduction

9

life of Europeans, then the process of further enlargement and integration is likely to gain more active and widespread support from European citizens. So far the European project has integrated economies and to some extent also political systems, but it has done fairly little to integrate societies (Bach 2006). The Copenhagen criteria required the fulfilment of economic (functioning market economy), political (democracy with rule of law and protection of minorities) and administrative criteria (adoption of the acquis communautaire). They did not imply any requirements in terms of social policy, social structure or living conditions of the population. As a consequence, even comparative knowledge about living conditions and quality of life in the enlarged EU remained in scarce supply. This book is one of the first attempts to fill the gap based on empirical research.

2. Objectives and approach of book As just indicated, the overall purpose of this book is to present a picture of key aspects of quality of life in the EU, with reference especially to the NMS and candidate countries viewed in comparison with the EU-15. It seeks to examine both people’s experience of selected domains of social life in EU states and the institutional contexts in which those experiences arise. It is interested in particular in the degree to which the eastern enlargement has increased heterogeneity and inequality in quality of life in the EU and the implications that may arise for EU integration. Most chapters have a large descriptive component, as they try to present a comparative picture of social conditions and institutional contexts across a large number of states in Europe. There is also an analytical focus on the relationship between daily conditions and institutional contexts on the one hand and implications for integration on the other. The book owes its origin to a comparative research project initiated in 2001 by the European Foundation for the Improvement of Living and Working Conditions with the purpose of comparing living conditions and quality of life in old and new member states of the European Union. A centrepiece of this project was the first European Quality of Life Survey (EQLS), which was carried out in 2003 on behalf of the European Foundation in 28 European countries – the EU-15, the 10 new member states (NMS) and what were then the 3 candidate countries (Bulgaria, Romania, and Turkey – CC-3). The national samples covered people aged 18 and above. The sample sizes in each country were approximately 1,000 persons, apart from the smaller countries (Luxembourg, Cyprus, Malta, Slovenia and Estonia) where they were approximately 600 (see Kohler’s methodological appendix in this volume (ch. 17) for further details and methodological assessment of this data source). Other datasets were also assembled for the project, including Eurobarometer data covering more or less the same countries from 2001 and 2002 (for an overview of results from the latter data, see Alber and Fahey 2004). Over the years since the project was launched, a range of senior sociologists from various European countries has been engaged in the analysis of these data-sets, most of whom are among the authors in this book. From this body of work, six main domains of quality of life issues have been selected as the focus of this book on the basis (a) that they are important issues that figure prominently in current policy debates, (b) that relevant comparative data are available for all, or nearly all, of the EU-27 countries and Turkey, and (c) that high-level experts are available to provide analysis and comment on the domain. The six domains are: (1) fertility,

10

Jens Alber, Tony Fahey and Chiara Saraceno

family and household structures; (2) employment and working conditions; (3) income and living standards; (4) housing; (5) sociability and social cohesion; (6) patterns of Europeanisation. A seventh technical domain has also been included, dealing with the quality of EU-wide data on social conditions. All of the six substantive domains are covered by chapters dealing with objective conditions and their subjective perception, while the first four also have parallel chapters that examine the institutional context and policies. 2.1 Fertility, families and households The chapters in Part I, Fertility, families and households, address from different perspectives the role of family relationships and of patterns of family formation in shaping quality of life. The degree to which people succeed in having the number of children they want is not often spoken of as a quality-of-life issue. Yet it clearly relates to fundamental human aspirations and also, in view of the crisis of fertility that many observers now ascribe to Europe, it has a strong bearing on the future vitality of Europe at the collective level. The relationship between aspirations and outcomes as far as numbers of children are concerned thus warrants the treatment Tony Fahey gives it in his chapter (1). It has often been observed that Europeans now generally have fewer children than they say they want (among the countries examined here, it is only in Turkey that the opposite holds true – women in Turkey on average have more children than they say they want). Fahey’s concern is to explore this gap in detail and assess the common claim that it provides a promising basis for future pronatalist policies – that is, policies designed to remove or counterbalance the obstacles that inhibit people from attaining the family sizes they want. He finds that while under-attainment of family size ideals is the norm on average in EU states, there is a spread around that average such that for some women the gap between ideal and actual is substantial, for others there is no gap at all (women have the number of children they want) and for yet others the gap is in the direction of over- rather underattainment – they have more children than they want. Furthermore, he finds that patterns of over-attainment and under-attainment are linked to patterns of social advantage and disadvantage: over-attainment is concentrated among less educated women while under-attainment is the dominant experience of the well educated. This points to a policy dilemma for those who might wish to introduce pro-natalist supports aimed at raising actual fertility closer to preferred levels. Insofar as they would be likely to work at all, they would be most effective if targeted on better educated women, since it is they who have the widest gap between actual and ideal fertility. But such upwards targeting would run counter to the traditional focus of social and family policy on less well-off groups. Patterns of family formation in Europe are increasingly diverse within each country as well as cross-nationally. After a long period in which they seemed to converge, family forms at the turn of the century have ‘recovered their complexity’ (as Göran Therborn once observed). In her chapter (2), Chiara Saraceno argues that these differences occur at three levels: in the institutional forms of family life, in the shape and the timing of life course transitions, and in everyday living arrangements. Her analysis focuses mainly on the second and third aspects. With regard to lifecourse transitions, given the lack of comparative longitudinal data, she focuses on the household

Introduction

11

arrangements of three broad age groups: the young, the elderly and those at midlife. The age of exiting the parental household shows high variation across Europe, thus producing the most striking differences with respect to the young, though these have an impact on the other two age groups also. For the elderly, the main differences are gender-related, as European men and women differ with respect to their life expectancies and the age at marriage. Some of the cross-country differences in patterns of family formation have their roots in old anthropological and historical divides, while others are new and are currently reshaping the European map ‘of families of families’. In several respects, the Southern European countries stand most clearly apart from other EU nations, as a high prevalence of conventional marriage and legitimate births coincides with very low fertility and late exit out of the parental household. The new member states do not form a distinct pattern. Some joint properties emerge, however, with regard to everyday living arrangements and particularly with respect to the high incidence of dual earner households. Here all former communist countries except Poland resemble the Nordic and Continental countries and stand apart from the Southern European ones. The time structures and gender tensions that underlie these apparently common patterns are different, however, especially because dual earner households in the former communist countries stand out for their very heavy overall workload. Families are not only about households, but also about kinship and intergenerational ties. Chapter 3 by Agnes Blome and Wolfgang Keck explores this issue from a limited, but crucial, perspective: that of relationships of solidarity and obligations between generations in ageing societies. The authors empirically test a widespread notion that informs much of public discourse, namely, that there is an inevitable conflict between generations over the redistribution of resources and responsibilities. Using comparative survey data (Eurobarometer and EQLS), they test this hypothesis with respect to three specific policy issues: care arrangements for the frail elderly, financing responsibilities for care, and the assessment of pension systems. Their findings show the persistent strength of intergenerational solidarity at the kinship level, but also at the societal level. In the field of care there is widespread recognition across all age-groups that adult children have a responsibility to look after their elderly parents when the parents can no longer manage to live on their own. Younger and middle-aged adults also stress the collective responsibility to pay for non-family care out of the public budget, while the elderly more often argue for taking the financial responsibility themselves, in order to unburden the younger generations. Perceptions of public pensions are more stratified by age, as younger age groups profess less trust in pension schemes than older generations. Low levels of trust in pensions are associated with a higher awareness of tensions between the generations, suggesting that the perception of intergenerational conflict is strongly mediated by the degree of trust in pension systems. Comparing old and new member states, the authors find no consistent country differences that would correspond to this distinction. Trust in the pension systems tends to be lower in new member states, but nation-specific differences within the old EU-15 are much bigger than the difference between the two country-group averages. In the field of care, countries with a stronger tradition of public care services have lower proportions of citizens who feel that care for a frail elderly is mainly a family responsibility. With their strong reliance on family provision, the new member states here resemble the Southern countries more than the Nordic ones.

12

Jens Alber, Tony Fahey and Chiara Saraceno

Interacting strongly with patterns of everyday organisation within households and families, public policies and welfare state arrangements also shape perceptions of family responsibilities. Thomas Bahle’s chapter (4) deals specifically with family policies. Usually ignored in mainstream comparative analysis and typologies, family policy probably constitutes the most differentiated and differentiating area of social policies in the EU. It is one where convergence is totally left to self-initiated processes of mutual learning, as the EU level lacks even an open method of coordination in this field. On the other hand, the maternity leave directive, the target concerning childcare coverage for the under three agreed upon at the Barcelona summit, and the various directives and recommendations concerning equal opportunities and work-life reconciliation might be defined as a nucleus of EU family policy. This joint nucleus interacts, however, with national policies that vary widely with respect to aims, means, and levels of generosity. Bahle shows that the stark differences are rooted in different ideas concerning a state’s legitimacy to develop a family policy, which in turn are based on the interaction of two phenomena: national-historical specific patterns of family formation and different church–state relationships. Having developed different ways of dealing with state–church competition, Protestant, Catholic and Orthodox countries formed different ‘families of family policy’. In the former communist countries one must add the impact of communism with its idea of the ‘working family’ and of the respective division of labour between state and family. The complex historical past thus produced four ‘families of family policy’ in Europe which the author identifies by looking at child allowances, childcare services, maternity and parental leaves. He applies three criteria to generate this typology: the degree of universality, the generosity of benefits and services, and the definition of family and state responsibilities. Bahle distinguishes four resulting types which he terms work compatibility, subsidiarity, mixed, and no family policy. Rather than developing a specific model of their own, the post-communist Central and Eastern European countries have joined Western European countries on the basis of their pre-communist religious culture and family history. Most of them joined the Central European ‘subsidiarity model’. This model is currently changing, however, as family policies all over Europe, except in the south, are moving closer to the Scandinavian model. From the perspective of family policies, enlargement thus has not increased European diversity to any sizeable degree. 2.2 Employment and working conditions The chapters in Part II on employment and working conditions deal with a central dimension of quality of life. Constituting the main access to income and social status in contemporary societies, employment crucially affects individual well-being as well as social integration. Having a job structures everyday time organisation, while work conditions affect the everyday living conditions of individuals and families. The question then is to what extent the enlarged European Union reaches the goal of highquality jobs for all. Based on official statistics, the chapter on employment (5) by Jens Alber describes and analyses the development of employment patterns in the enlarged European Union. Even though the new member states tend to have lower levels of employment on average, recent labour market developments cut across the East–West divide, as the groups of good and poor labour market performers consist of old and new member states alike. Whilst group-specific risks of labour market exclusion tend

Introduction

13

to be similarly concentrated on vulnerable groups across Europe, the new member states stand out for two peculiarities: several of them have very high levels of youth unemployment, and low-skilled groups face particularly steep impediments. The structures of employment still show much national diversity. Whilst industrial employment has declined to similarly low levels in Western and Eastern European countries, the new member states resemble the Southern European old members in having higher levels of agricultural employment and lower levels of employment in the service sector, where there is most future growth potential. After having gone through a protracted phase of jobless growth, the Central and Eastern European new member states have translated their above-average economic growth rates into sizeable employment increases, especially in the Baltic nations. The chapter concludes with an analysis of the impact of welfare policies on employment. Even though data are not available for all new member states, this analysis suggests two preliminary conclusions. (1) The new member states did not pursue social dumping policies, but stand out for particularly high burdens of social insurance contributions which are associated with low levels of employment in the low-skill sector. (2) The group of successful labour market performers includes nations with generous and stingy welfare benefits alike, suggesting that there are two paths to full employment in Europe, i.e. the liberal model of the Anglo-Saxon countries and the more state-centred model of the Scandinavian countries where public services function as the engine of employment growth in the service sector. Claire Wallace and Florian Pichler shift attention from the overall structure and functioning of labour markets to the workers’ experience of their jobs (Chapter 6). They assess that experience from a quality-of-life point of view, focusing especially on the contrast between Eastern and Western Europe. The communist era in Eastern Europe was in theory a worker’s utopia, with guaranteed full employment and a range of services, such as health, education, holidays, pensions and extensive childcare, provided in return for work effort. The reality of work and its rewards was less rosy, as captured in the adage from the communist era quoted by Wallace and Pichler – ‘we pretend to work and they pretend to pay us’. The transition to the market economy shattered old work relations and the benefit systems that went with them. Today, the problems experienced by workers in Eastern Europe are many, and while being of a similar kind as in Western Europe, are much more severe in the new member states. Eastern Europeans in general work longer hours (partly because they often have secondary jobs), they are badly paid, their employment is insecure and their jobs are often dull, boring and located in unpleasant working environments. However, there is considerable variation around the average, both between and within countries. The accounts of labour market developments and of the quality of jobs are complemented by Jelle Visser’s Chapter 7 on changes in labour market institutions. Visser’s account centres on the effects of enlargement on regulatory structures and on the relationship between unions and employers. Visser considers a future convergence towards what used to be the mainstream Western European model of labour relations based on sectoral bargaining and upward harmonisation as highly unlikely, because the extension of a common market with the free movement of workers and services has constraining effects on unions and regulations. On both sides of the old EU-15 border, more flexible regulatory approaches allow exit from standards that can no longer be sustained in the face of increased international competition. Under the

14

Jens Alber, Tony Fahey and Chiara Saraceno

motto of ‘flexibilisation’, contractual arrangements at variance with standard employment contracts have proliferated across the enlarged EU. With independent unions and autonomous employer organisations absent, and larger shares of employment located in small firms, the new member states lack decisive preconditions for the implementation of the Western European model of collective bargaining. Bargaining at the enterprise level has become more important almost everywhere. Minimum wages which exist in 18 of the 25 member states and in all former socialist countries counteract the trend towards rising wage inequality only partly, because minimum wage adjustments lagged behind average wage growth in many member states. Whereas several EU-15 countries retained or reintroduced social pacts as instruments of macro-economic coordination, Slovenia is the only new member state where social pacts played a role. However, all CEE countries have founded tripartite councils in an attempt to comply with the European idea of social dialogue. Yet this dialogue remains particularly underdeveloped on the transnational level, as key sectors of the European economy are without transnational employer associations. Visser does not expect framework directives on the European level to have strong standardising effects, because their impact depends on how the legislation integrates with national structures. Within weakly developed national institutions for the collective representation of employees at the firm or national level, the harmonising effect of EU legislation will remain limited. 2.3 Material living conditions The four chapters in Part III on material living conditions focus on two crucial dimensions that affect quality of life: material deprivation and housing conditions. These chapters offer an integrated overview of objective circumstances and of policies in these two fields. Although there is no consensus on what should be included in measuring quality of life, all agree that living standards are central – it is difficult to live a good life if one is beset by poverty, deprivation and economic vulnerability. However, in taking up these topics in Chapter 8, Christopher T. Whelan and Bertrand Maître are faced with the difficulty of deciding who should be considered as suffering from inadequate living standards in the enlarged EU. The problem here is that member states differ so widely in their level of economic development that those who are conventionally defined as poor in some member states countries are better off in absolute terms than those who are considered well-off in others – the classic difficulty arising from cross-national comparisons of relative notions of poverty. Whelan and Maître show that this difficulty is not only a matter of income but is evident also in direct measures of material hardship: in many of the NMS, even middle-income groups experience shortfalls in basic consumption resources that are unknown amongst most of the ‘poor’ in the richer member states. Whelan and Maître’s contribution is to recommend a novel way of coping with this contradiction. They reject the view that a pan-European approach to poverty measurement would help illuminate this issue and present instead a complex, multi-dimensional measure of economic vulnerability that combines relative and absolute indicators using latent class analysis. Their key result, similar to the findings in the chapter by Petra Böhnke, is that the economically vulnerable do indeed make up a larger share of the population in the poorer member states than in the richer, but that they are more excluded – more cut off from the mainstream – in the richer countries.

Introduction

15

In order to be effective, the fight against income poverty, deprivation, and social exclusion requires more than a single policy measure. Employment and local development policies, support for working women, adequate family transfers and old age pensions are among the most effective preventive measures in this field. Most European countries also provide an entitlement to minimum income support in the case of poverty as a fundamental element of their social rights package. On the European level, a 1992 resolution promoted a minimum resources guarantee. Its actual implementation was originally left to the single member states, but later incorporated into the ‘soft’ framework regulation of the Open Method of Coordination. Bea Cantillon, Natascha Van Mechelen and Bernd Schulte in their Chapter 9 examine the effect that the EU resolution and the monitoring process had on the persistent differences in implementation of minimum income policies. Differences persist because of the complexities of national social security systems (old age pensions, invalidity benefits, unemployment benefits, family transfers, tax allowances and so forth), but also because there are implicit and explicit assumptions in national welfare cultures concerning the deserving and undeserving poor, the role of family solidarity, or the risks of welfare dependency. Minimum income support measures, therefore, differ not only with regard to eligibility rules and generosity, but also with respect to administrative styles and the application of discretion rules. Most countries increasingly link income support to some kind of ‘activation’ goal, but the way this goal is implemented and the rights and obligations attached to it add a further element of diversity among the systems. Once again, the largest differences are found not between old and new member states, but across them. The only two countries that do not have a general system of income support for the poor are two ‘old EU’ members: Italy and Greece. In all former communist countries, however, the issue of poverty surfaced dramatically in the transition period in an institutional framework that traditionally did not include any specific poverty policies. They have started slowly to set them in place only recently. The broad-ranging treatment of living standards and income maintenance issues contained in Chapters 8 and 9 is followed in Chapters 10 and 11 with a focus on a specific but highly important element of material living conditions, namely, housing. In Chapter 10, Henryk Domanski ´ describes and assesses the housing conditions of the populations of the enlarged EU. The starting point here is the widespread perception that this is one area where the communist legacy in Eastern Europe has been unfortunate in that it has left people badly housed. Domanski’s ´ account confirms that this perception is largely accurate: housing conditions are generally poor in the eastern countries, as evidenced by a range of indicators from over-crowding to lack of indoor plumbing. Furthermore, these differences are reflected in how people in the worse-off countries feel about their situation – knowing that the majority of one’s compatriots are in the same situation does not really make it much easier to live in cold, cramped and badly built apartments. However, Domanski ´ also highlights some interesting shades of variation in this broad picture. One is that the east–west divide in housing standards is far from clear-cut: some parts of Southern Europe have as many, or more, problems in housing conditions as do the better-off of the former communist countries. Another is the extraordinary situation with regard to housing tenure that suddenly emerged in Eastern Europe with the collapse of communism. The huge stock of state-owned housing that had been a feature of social provision in the communist era was quickly handed over to the ownership of its occupants as the

16

Jens Alber, Tony Fahey and Chiara Saraceno

communist system fell apart, largely because the new state administrations lacked the capacity to manage or maintain the masses of shoddy buildings previously under state control. The consequence is that the NMS have home ownership rates generally exceeding 70 per cent – but in a situation where home ownership has none of the connotations of accumulated wealth and success in life that often attach to it in the older member states. Rather, for most newly minted home owners in the NMS, ownership of their dwellings has simply brought onto themselves the massive task of upgrading decrepit buildings that formerly they might have hoped would be undertaken by the state. In Chapter 11, Michelle Norris takes the analysis of the housing situation in the NMS a step further by examining the institutional context within which housing is provided in these countries, viewed in contrast with the housing systems of the ‘old’ EU. Her main focus is on the institutional legacy left in the wake of the ‘East European housing model’ created in the communist era, particularly as it affects housing inequalities within and between countries. She identifies four key strands to that legacy – housing tenure policy, housing finance and subsidy systems, housing construction systems, and governance. For each strand, she provides a brief sketch of the communist-era background, the present-day situation in the NMS and the contrasts offered by the rest of EU – contrasts which differ broadly between the northern and southern states of the EU-15. The key historical features of the East European housing model can be summarised as: (1) high levels of state ownership and an aspiration for universal social provision of housing that could not really be sustained; (2) inadequate investment in housing as available capital was concentrated on the build-up of heavy industry; (3) large monopolistic state construction companies that were hugely inefficient and unresponsive to consumer needs; (4) poor governance at local level, resulting in inadequate management and maintenance of dwellings and poor policies on such things as rent setting and housing allocations. The rush to the market that occurred after the fall of communism transformed the surface character of some of these features – as in the massive privatisation of housing referred to in Domanski’s ´ chapter and the opening up of housing markets to private finance. However, these surface changes did not solve the underlying difficulties. Privatisation did nothing to rehabilitate dwellings and private finance made scarcely any difference to most housing since mortgage markets, the key channel of private financing for housing in other countries, remained underdeveloped. Not all of these problems were confined to Eastern Europe since some of them, such as the lack of ready access to private finance and poor management structures, were found in other parts of Europe, especially in the Southern European states. 2.4 Social capital and social cohesion The chapters in Part IV on Social capital and social cohesion address the issue of subjective perceptions of one’s own quality of life and of the quality of society including the perception of tensions which may threaten social cohesion. Chapter 12 by Manuela Olagnero, Paola Torrioni and Chiara Saraceno focuses on the quality of social networks, with particular regard to the balance of integration in family and non-family networks, including investments into the civic and public sphere. They find distinct patterns of sociability that identify six country clusters. These overlap only partly with known welfare state typologies such as Esping-Andersen’s

Introduction

17

(Western Europe based) distinction between three types of welfare regimes. Being distributed across several clusters, the former communist countries are as internally differentiated as the old EU-15 member states. The findings furthermore suggest that patterns of public solidarity and social cohesion can explain (or be explained by) patterns of private sociability and feelings of social belonging only partly. High standards of living, high rates of employment, and high levels of technological literacy are drivers of a multi-dimensional kind of sociability, whereas high rates of unemployment and of poverty restrict the range of social networks. There is no indication that a high involvement in family networks crowds out involvement in non-family ones or that involvement in the private sphere limits involvement in public life. There is, however, indication that disadvantageous circumstances which are more frequent in some Southern European countries and most of the new member states restrict the range and kind of social networks to close family and kin. Focusing on individuals’ own perceptions, the chapter by Petra Böhnke (13) analyses how widespread feelings of marginalisation are in the European Union and how differences between nations in this regard can be explained. Given the heterogeneity of living conditions in the enlarged EU, the author’s central concern is to find out if similar disadvantages bring about identical perceptions of marginalisation even if the macro-contexts are different. On average, Böhnke finds the prevalence of feelings of belonging to be lower in the new member states. However, the rank-order of member states does not fully correspond to the distinction between old and new members, as self-assessed marginalisation is found to be least prevalent in Scandinavia, Cyprus and Slovenia and most widespread in the Baltic countries, the Czech Republic and Slovakia, and in Bulgaria and Romania (as well as in Turkey). Across Europe perceived social exclusion is strongly related to unemployment and poverty. The lower the income the more likely it is that people feel at the margins of society, especially if they do not have a job. Yet unemployment or economic strains are not the only determinants of perceived marginalisation. The absence of family back-up and social network support are similarly important drivers. Thus, social back-up helps to considerably soften the consequences of material deprivation. This suggests that quality of life requires a share in economic prosperity as well as access to meaningful social relationships. Beyond individually experienced disadvantages, it also matters in which country or social context people experience social or economic hardship. Adverse conditions are most likely to give rise to feelings of marginalisation in the more prosperous countries where deprivation is least common. Rich countries thus have lower levels of perceived social exclusion, but higher degrees of polarisation in the perception of exclusion between marginalised and privileged groups, because vulnerable groups are likely to feel more stigmatised and marginalised in countries where the overall level of well-being is high. Dealing with the perception of group conflicts in Europe, the chapter by Jan Delhey and Wolfgang Keck (14) focuses on an aspect of social cohesion which is central to the concept, yet frequently overlooked: the degree to which different groups share common feelings of belonging and have cooperative rather than hostile relations. Based on the 2003 EQLS, the authors examine to what extent Europeans perceive strong tensions between various kinds of social groups. Focusing on vertical tensions such as rich vs. poor or management vs. workers on the one side, and on ethnic tensions on the other, they show that old and new member states do not differ much in the overall level of perceived social conflicts and that throughout Europe perceptions of

18

Jens Alber, Tony Fahey and Chiara Saraceno

tensions are weakly stratified, as privileged and disadvantaged groups within single nations show a similar awareness of strains. However, the citizens in old and new member states are preoccupied with different kinds of conflicts. Whereas people in the new member states worry most about distributional conflicts, the citizens in EU-15 nations are most concerned about ethnic tensions. Hence the authors conclude that the Eastern enlargements will shift the conflict agenda within the European Union towards distributional issues which have been toned down in the more affluent old member states. This means that successful cohesion policies in the enlarged European Union will have to be two-pronged. In the post-communist new member states – and to a similar extent also in France and Greece – the priority must be to smooth vertical conflicts about the distribution of material living conditions. In most of the old member states, and also in Hungary and the Czech Republic, the focus must be on mediating ethnic tensions which have become a major citizen concern in these countries. 2.5 Processes of Europeanisation Part V on Processes of Europeanisation deals with the extent to which Europeans are moving beyond the nation state and beginning to think and act in transnational European terms. The possibility of such a shift arises especially in connection with migration. The image of an EU where workers move freely between member states as they seek to advance their careers is dear to the hearts of EU policy makers and is part of the core vision of the EU project. The problem is that EU citizens do not share in this vision, not only because they dislike having migrants move into their areas but also because, as Hubert Krieger points out in his Chapter 15 on migration intentions, they are slow to move themselves. Krieger sets out to examine the latter aspect of Europe’s lack of ‘mobility culture’. He focuses on people’s stated intentions to move across regions or to other countries in the EU as a means of probing mobility culture. This focus on intentions rather than actual migration is adopted partly for the pragmatic reason that data on mobility intentions are available for the whole of the EU (something which is not true for actual migration), but also because they reveal something of people’s perceptions of the desirability of moving, even if obstacles of various kinds may inhibit them from acting on those perceptions. He uses data on mobility intentions for all the present EU-27 states drawn from 2001 and 2005. A key result is that there is an upward shift in people’s intention to move in the future that occurred between 2001 and 2005. Krieger interprets this as evidence of a growing mobility culture in Europe, but he also points out that the shift is so variable across countries that it is hard to view it as constituting a single consistent underlying trend. In looking at the influences on mobility intentions, he finds a range of socio-demographic factors to have reasonably stable effects over time. Subjective quality of life perceptions such as the general life satisfaction are even stronger predictors of cross-border migration intentions than objective macroeconomic indicators. Not surprisingly, the analysis also reveals the strong disincentive effects which language barriers exert on cross-border migration intentions in Europe. Chapter 16 by Jan Delhey and Ulrich Kohler probes into transnational aspects of perceived quality of life by examining how citizens judge the quality of life in their own country relative to the EU as a whole. Departing from sociological reference group theory and based on 2004 Eurobarometer survey data for the old and new

Introduction

19

member states, the authors take up three questions: Do citizens see their own country as offering better or worse life chances than the EU average? To what extent do their beliefs match reality? Which impact do perceptions of the European ranking of their country have on quality of life as measured by individual life satisfaction? The authors show that with respect to the material standard of living, the employment situation, and the overall quality of life, Europeans have a fairly precise idea of where their country stands in the league table of member states. They also demonstrate that perceptions of collective deprivation relative to the EU average impact negatively on personal life satisfaction, even if other determinants are statistically controlled. Their results thus suggest two major conclusions: (1) As also shown in Chapter 13, macro-conditions have an effect on well-being independent of individual life circumstances. (2) Having become a relevant yardstick for ordinary citizens who evaluate the social conditions they live in, the EU increasingly constitutes an integrated social space that provides a common cognitive frame for European citizens. As citizens are no longer only nationally oriented, but develop European aspiration levels, economic growth and improvements of material living conditions in single nations cannot be expected to produce similar improvements in subjective well-being as long as the relative position of the country within the European Union remains constant. Hence if Europeans are to become similarly satisfied across the Union, some convergence mitigating regional disparities and improving the lot in poorer countries will be required. Finally, Chapter 17 by Ulrich Kohler, gives a methodological quality assessment of different European Surveys including the European Quality of Life Survey on which most articles in this book are based. Examining the ‘European Value Study 1999’, the ‘European Social Survey 2002’, the ‘International Social Survey Programme 2002’, the ‘European Quality of Life Survey 2003’, and the ‘Eurobarometer 62.1’, he discusses to what extent certain methodological quality criteria were met and proposes some guidelines for future rounds of comparative European surveys.

3. Conclusions From the wide range of analyses presented in the chapters just outlined, no simple picture of quality of life in the enlarged EU emerges. Yet some brief overall conclusions can be drawn. The first concerns the value of the quality of life perspective itself as a way of evaluating the condition of societies in Europe. There are many instances where the multi-dimensional assessment that characterises the quality of life approach does no more than confirm what we already knew from narrower old-style indicators such as GDP per head. It is not just that conventional measures of economic output are good predictors of those aspects of quality of life that have a largely economic character, such as living standards and the risk of material hardship, or housing conditions, or working conditions. In all these areas, not surprisingly, there is a rough correspondence between how well societies in Europe measure up and how advanced their economies are. However, a similar relationship also seems to hold in connection with some important social aspects of quality of life, such as feelings of marginalisation and to some degree also patterns of social networks. It might be argued, of course, that in European states measures based on GDP are not just about the economy but capture also the whole gamut of civic, social and political characteristics that are particularly favourable in the old, deeply rooted

20

Jens Alber, Tony Fahey and Chiara Saraceno

democracies of north-west Europe and that weaken somewhat as one goes south (Portugal, Spain and Greece, after all, escaped dictatorship only in the 1970s), and weaken further as one enters the former communist new member states. This gradient in the age of democracy across Europe is similar to that in GDP per head. It would be a difficult matter to disentangle how the elements of this correlation fit together and which of them are most important as influences on quality of life. In any event, as a general rule in present-day Europe, it might justifiably be argued on the basis of evidence presented here that strong economies and good quality societies tend to go hand in hand and together provide a strong under-pinning for a high quality of life for individuals. All that being said, however, there are many ways in which the chapters in this book take us beyond a simple identity between the level of economic development and the quality of life, and it is here that the value of the broader multi-dimensional assessment shows its worth. For one thing, there are some central underpinnings of people’s well-being that are only weakly influenced by economic conditions – or at least where variation across Europe does not run in tandem with level of GDP. The most important of these is family life with its different kinds of gender and intergenerational relations. Differences across the member states do not vary along rich–poor lines, nor indeed between east and west. Although there are some indications of a distinctive Southern European pattern, the similarities and differences between countries in this area are little associated with current economic conditions, or indeed with any other contemporary societal features, so that one might speculate on a possible link with older historically rooted patterns of cultural difference. In any case, these different patterns of family formation suggest that there might be different understandings across Europe about what quality of life is about, at least in the domain of private, family relationships. They also suggest that cross-country differences in resources available to pursue one’s idea of a good life – e.g. achieving autonomy when young, leaving a bad marriage, balancing family and work obligations – might differ not only because of GDP levels but because of the overall way society is organised through its welfare state and gender and intergenerational arrangements. As indicated in the chapter on family policies, these arrangements differ in ways that cannot at all be explained by GDP levels. The analysis of the gap between actual and ideal family size also reveals complex differences between countries on similar levels of economic development, though – standing back a little – one might also be struck by how uniform European countries are in this area, with the near-universal pattern of small actual family sizes and larger ideal family sizes, especially in contrast to Turkey. The important cleavage in this respect is not between countries but between educational categories within countries, with the well educated across all countries generally having as much or more in common with each other than with the less educated in their own countries. Complexities along somewhat similar lines arise in connection with patterns of poverty and social marginalistion, an area where, as already mentioned, the richer EU states have less severe problems in absolute terms than the poorer ones. For the former communist countries, the experience of mass poverty has been one of the unforeseen results of the transition process, radically modifying people’s expectations and conditions of living. Yet, in richer countries the superior absolute level of wellbeing goes hand in hand with relative problems that are worse. Those on the lowest rungs of the ladder of advantage and social inclusion in the richer states are located

Introduction

21

on ground that in certain senses is quite elevated, since it places them as high as those well up the ladder in the poorer countries. But they are further away from the mainstream in their own societies and so, relatively speaking, are worse off and more marginalised in their own national contexts than the lower rungs in the poorer countries. Here again, subtleties are revealed that are important to be aware of, even if it is not at all clear how they should be reckoned up in trying to arrive at overall judgements of quality of life. As well as examining people’s direct experience of major domains of quality of life, part of the purpose of this book is to provide an idea of the institutional context within which individual experiences arise. This purpose was adopted with particular reference to the new member states, where key institutions are likely to reflect a mix of legacies from the communist era, of importation from the western democracies and of improvisation arising from the admixture of the two. Furthermore, it is sometimes neither the old nor the new regime that is the crucial influence on people’s well-being but the manner of the transition between the two. With respect to employment, the shock of transition from the communist to the market economy brought about a period of turbulence in the NMS that was of a similar order of severity as the great depression of the 1930s in the west. Since labour market institutions are weak overall in the former communist countries, workers have less means than in most old EU countries to negotiate their position in increasingly flexible labour markets. But also within the EU, labour market conditions and institutional mechanisms remain quite differentiated, and the most successful cases in terms of employment development represent two almost opposite social models (the Anglo-Saxon and the Nordic approach to full employment). Institutional transformation in the NMS was and is to a great deal shaped also by the accession process itself. Originally, under the impact of the Copenhagen criteria, emphasis was placed almost exclusively on establishing a market economy and democratic forms of governance. Later, following the Lisbon summit, there was also a new emphasis on the European social model with developed social policies. Yet, since this European model is to a large degree a kaleidoscope of nationally specific arrangements and welfare regimes, which the EU promotes mostly through means of ‘soft law’ while leaving its concrete implementation to national governments, it is not surprising that the NMS also contribute to this kaleidoscope, rather than forming a homogeneous group of their own. More than in employment policies, this is apparent in the way anti-poverty/social inclusion policies are developed across the old and enlarged EU. The fact that the former communist countries had no policy legacy to build on in this field would have provided the most favourable condition for the development of largely similar policies. Yet, each country pursued different approaches with regard to aspects that are important for a ‘social model’: generosity, duration, activation. These differences partly overlap and partly add to those already existing in the old EU, where two countries (Greece and Italy) even still lack such an income support policy for the poor. None of these findings point to a clear answer to the question posed earlier as to the likely impact of the eastern enlargement on the future cohesion and viability of the EU. Viewed from a quality of life perspective, there are many signs of convergence and of successful re-integration of the former communist states into the west European model of democratic, market-economy states. But there are also many signs of deep, persistent heterogeneity and of gulfs in quality of life between east and west.

22

Jens Alber, Tony Fahey and Chiara Saraceno

And there are persistent differences also within the West both in everyday arrangements (patterns of family formation, gender arrangements, the balances between autonomy and interdependence of generations) and in institutions. Judged on the basis of the evidence presented in the chapters that follow, whether or in what way the EU will be able to hold its now very different parts together and move towards closer union remains an open question. Interestingly, as suggested in Chapter 16, European citizens are beginning to become European in their framework of expectations and of comparative evaluations. This may be an important resource in the construction of Europe. But it may on the contrary fire back on this very construction if the conditions of living and the national patterns of social citizenship remain too far apart. In preparing this volume, the editors had valuable help from several staff members at the Social Science Center Berlin, especially Florian Fliegner, Bettina Mertel, and Marion Obermaier to whom we wish to express our gratitude. Our special thanks are due, however, to Martina Sander-Blanck, also from the Social Science Center, who transcribed all manuscripts into the template form for the publishers with admirable care, diligence, and speed, as well as with never-failing good humour.

Notes 1 The problem, of course, was that this distinction never fitted the situation even with respect to the war in Iraq. The ‘coalition of the willing’ who rallied behind the United States in their ‘letter of eight’ written in 2002 included five old member states (Denmark, Italy, Portugal, Spain and the United Kingdom), as well as three acceding countries (Czech Republic, Hungary, and Poland). Only later did the ten Central and Eastern European countries who were applying to join NATO unite as the ‘Vilnius Ten’ to sign a letter expressing even more explicit support for the United States (Ash 2004). 2 One of the implications of this approach is that, judged by this standard, the United States stands out as far more cohesive than the European Union, because regional differences are less pronounced. 3 The fact that almost one-half of the respondents in Estonia welcomed openness may perhaps be related to the mixed Estonian/Russian ethnic composition of the country.

References Alber, J. and Fahey, T. (2004) Perceptions of Living Conditions in an Enlarged Europe, Luxembourg: Office for Official Publications of the European Communities. Ash, T.G. (2004) Free World, London: Penguin. Bach, M. (2006) ‘The enlargement crisis of the European Union: from political integration to social disintegration?’, pp. 11–28, in M. Bach, C. Lahusen and G. Vobruba (eds), Europe in Motion: Social Dynamics and Political Institutions in an Enlarging Europe, Berlin: edition sigma. Baratta, M. von (ed.) (2003) Der Fischer Weltalmanach 2004, Frankfurt a. M.: Fischer Taschenbuch Verlag. Berié, E. and Kobert, H. (eds) (2005) Der Fischer Weltalmanach 2006, Frankfurt a. M.: Fischer Taschenbuch Verlag. Berié, E. and Kobert, H. (eds) (2006) Der Fischer Weltalmanach 2007, Frankfurt a. M.: Fischer Taschenbuch Verlag.

Introduction

23

Blair, T. (2005) ‘Vision for the UK presidency of the EU’, speech given by Tony Blair to the European Parliament, 23 June. Online. Available http: (accessed February 2006). Blair, T. (2006) ‘Europe is falling behind’, Newsweek, special edition (issues 2006): 26–27. Brown, G. (2005) Global Europe: Full-employment Europe, London: HM Treasury. Commission of the European Communities (2001) European Governance: A White Paper, Brussels, COM (2001) 428 final. Commission of the European Communities (2003) Report from the Commission on European governance, Luxembourg: Office for Official Publications of the European Communities. Commission of the European Communities (2005) The Commission’s Contribution to the Period of Reflection and Beyond: Plan-D for Democracy, Dialogue and Debate. Communication from the Commission to the Council, the European Parliament, the European Economic and Social Committee and the Committee of the Regions, Brussels, Com (2005) 494 final. European Commission (2002) Standard Eurobarometer 56: Full Report, Brussels. European Commission (2006a) Eurobarometer 66: First Results, Brussels. European Commission (2006b) The Future of Europe: Special Eurobarometer 251, Brussels. Freedom House (2006) Nations in Transit 2006: Democratization from Central Europe to Eurasia. Online. Available http: http://www.freedomhouse.hu/nitransit/2006/ corruption2006.pdf (accessed 23.10.2006). Gerhards, J. (2006) ‘Europäische versus nationale Gleichheit. Die Akzeptanz der Freizügigkeitsregel für Arbeitskräfte in den Mitglieds- und Beitrittsländern der Europäischen Union’, pp. 253–278, in M. Heidenreich (ed.), Die Europäisierung sozialer Ungleichheit zwischen nationaler Solidarität, europäischer Koordinierung und globalem Wettbewerb, Frankfurt a. M.: Campus. Gerhards, J. (with M. Hölscher) (2005) Kulturelle Unterschiede in der Europäischen Union. Ein Vergleich zwischen Mitgliedsländern, Beitrittskandidaten und der Türkei, Wiesbaden: Verlag für Sozialwissenschaften. Giddens, A. (2007) Europe in a Global Age, Cambridge: Polity Press. Heidenreich, M. (2006) ‘The decision-making capacity of the European Union after the fifth enlargement’, pp. 29–57, in M. Bach, C. Lahusen and G. Vobruba (eds), Europe in Motion. Social Dynamics and Political Institutions in an Enlarging Europe, Berlin: edition sigma. Kleinman, M. (2002) A European Welfare State? European Union Social Policy in Context, Houndmills: Palgrave Macmillan. Moravcsik, A. (2004) ‘The myth of a European “leadership crisis”’. Online. Available http: (accessed February 2006). Official Journal of the European Union (2004) ‘Treaty establishing a Constitution for Europe’, C310, Vol. 47, 16 December. Online. Available http: (accessed February 2006). Rose, R. (2004) ‘Voter turnout in the European Union member countries’, pp. 17–24, in International Institute for Democracy and Electoral Assistance (ed.), Voter Turnout in Western Europe, Stockholm: Publications Office, International IDEA. Scharpf, F. (1999) Regieren in Europa. Effektiv und demokratisch?, Frankfurt a. M., New York: Campus. Schmitt, H. and Scholz, E. (2005) The Mannheim Eurobarometer Trend File 1970–2002, Data Set Edition 2.00 Codebook on Unweighted Frequency Distributions, Mannheim, Cologne: MZES, ZUMA and ZA. Online. Available HTTP: (accessed February 2006). Stiglitz, J.E. (1999a) ‘Whither reform? Ten years of the transition’, in B. Plesovic and J.E. Stiglitz (eds), World Bank Annual Conference on Development Economics, Washington, DC: World Bank.

24

Jens Alber, Tony Fahey and Chiara Saraceno

Stiglitz, J.E. (1999b) ‘The World Bank at the millennium’, Economic Journal, 109, 459: F577–97. Therborn, G. (2000) Die Gesellschaften Europas 1945–2000. Ein soziologischer Vergleich, Frankfurt a. M.: Campus. Transparency International (2005) Corruption Perception Index (CPI). Online. Available http: (accessed 25.10.2006). Verheugen, G. (2005) Europa in der Krise. Für eine Neubegründung der europäischen Idee, Köln: Kiepenheuer & Witsch. Wessels, B. (2006) ‘“Founding elections” der erweiterten EU und europäische Integration’, pp. 231–251, in J. Alber and W. Merkel (eds), Europas Osterweiterung: das Ende der Vertiefung?, WZB-Jahrbuch 2005, Berlin: edition sigma.

Part I

Fertility, families and households

1

Fertility patterns and aspirations in Europe Tony Fahey

Introduction The European Commission’s Green Paper on demographic change expressed the view that the EU no longer has a ‘demographic motor’: the Union now has nearly as many deaths as births per year and without inward migration would soon be headed towards population decline (European Commission 2005). The implications for the future of Europe are hard to predict but are unlikely to be good. It is difficult to see how the EU can become the most dynamic and competitive economy in the world, as is the aspiration of the Lisbon agenda, while its population is greying and its workforce shrinking. The key problem is the very low birth rates now found in Europe. At present, the total fertility rate (TFR) in the EU-25 is at about three-quarters of the level needed to replace the population (the TFR is estimated at 1.52 for 2005, while the replacement TFR is conventionally defined as 2.1 – New Cronos 2005). This is one area where the divide between new and old member states is not that significant a part of the EU picture. The ten member states that joined in 2004 have an even weaker reproductive performance than the EU-25 average, with a TFR in that year of 1.27 (Bulgaria and Romania are at a similar level). But a number of EU-15 states also have an equally low TFR so that in this area diversity within the ‘old’ member states is as great as any gap between the old and the new. While many aspects of low fertility in Europe have been examined by researchers (for recent overviews, see Billari 2005; d’Addio and Mira d’Ercole 2005; United Nations 2003), the feature focused on in this chapter is the gap that has emerged between actual and preferred fertility: the number of children people have is, on average, less than the number they would like to have (Bernardi 2005; d’Addio and Mira d’Ercole 2005: 41–44; Goldstein et al. 2003; Bongaarts 2002: 426–427; van Peer 2002; van de Kaa 1998). Chesnais (1998), for example, points out that while women in Europe on average say they want an average family size of 2.2, the actual total fertility rate is only 1.45. This shortfall, which emerged historically in the course of the transition to low fertility, is the reverse of what is found in high fertility countries, where women typically have more births than they say they want (United Nations Populations Division 1995: 59–67; Bongaarts 1998: 8–11). The gap between actual and preferred fertility in low fertility countries is not usually included among the quality of life issues studied by researchers who work within the quality of life approach. However, it is known that the family context is one of the strongest social influences on the quality of people’s lives (see Chapter 13: Böhnke, this volume), and having or not having children relates to quite profound aspects of

28

Tony Fahey

family life. The actual–preferred fertility gap could thus be thought of as having fundamental significance for quality of life. It has also been read by some as having considerable policy significance in the light of the increasing interest in revitalising the EU’s demographic performance. It suggests that governments might be successful in getting Europe’s demographic motor moving again if they adopted appropriate pronatalist policies. For Chesnais (1998), for example, the actual–preferred fertility gap reflects a ‘latent demand for family support’; while for Sleebos (2003: 30) it ‘provides a window of opportunity for policies aimed to increase fertility and to bring it into line with individual preferences’. The European Commission’s Green Paper asserts that ‘if appropriate mechanisms existed to allow couples to have the number of children they want, the fertility could rise overall’ (European Commission 2005: 5). Others may doubt that great weight should be attached to people’s stated preferences for children. Conventional economics assumes that we all want more than we have of all good things, and the compromises we settle for through our behaviour are a better guide to the mix we really want than are our stated preferences. Nevertheless, stated preferences are worth taking into account in the present instance since they indicate that most people still regard children as a good they want more of. This is quite a significant fact, since it indicates that people feel some unease or dissatisfaction with how few children they have had. This is good news for governments concerned about low fertility, since, at the very least, it suggests that family policies aimed at supporting birth rates are likely to be swimming with rather than against a tide of popular preferences. It is another matter whether such policies are likely to be effective, since research has shown that state supports for child-rearing at best have only modest effects on fertility outcomes and are less important than macroeconomic influences, of which buoyant demand for female labour seems to be the most important (Sleebos 2003; d’Addio and Mira d’Ercole 2005). Nevertheless, the actualpreferred fertility gap is worth taking some note of, since it is a part of the context within which the problem of Europe’s demographic weakness might be addressed. In spite of the interest in the gap between preferred and actual fertility, it has been little analysed from the point of view of its significance for present or future pronatalist policy in low fertility countries. Demographers have regularly included measures of desired fertility in fertility surveys since the 1950s but they have been preoccupied with their value for predicting the future fertility behaviour of women still in their childbearing years rather by the final gap between preferred and actual number of children among those with completed families (from a large literature, see e.g. Freedman et al. 1980, Thomson and Brandreth 1995, Thomson 1997, Schoen et al. 1999; for a rare example of a focus on the gap between preferred and actual family size as an object of interest in its own right, see van Peer 2000). It is this final gap between outcomes and preferences, rather than the predictive value of the preferences, that becomes a matter of central interest when, as here, the concern is with the specific issue of very low fertility rather than with broader explanations of reproductive behaviour. This chapter first provides some contextual information on fertility trends and patterns in Europe, focusing especially on the question of whether there are broad regional distinctions to be found in these patterns with regard either to west–east distinctions or to differences between ‘families of nations’. The chapter then turns to the gap between actual and preferred family size in Europe. It examines evidence of the extent of the gap and how it has changed over recent decades, and assesses the

Fertility patterns and aspirations

29

relevance of that gap from a family policy point of view, with particular reference to its significance for potential pronatalist policy.

1. Context: general patterns While below-replacement fertility is now the norm in the developed world, and is increasingly common in poorer countries, there is considerable diversity in how far fertility has fallen below replacement levels. In Europe, Billari (2005) classifies countries into those with ‘lowest low’ fertility (a total fertility rate below 1.3), ‘very low’ fertility (below 1.5), and ‘low’ fertility (below 2). This categorisation roughly coincides with a spatial gradient from north-west to south-east Europe (Council of Europe 2005). ‘Low’ fertility is found in a band of countries running roughly along the north-west edge of the continent – the Scandinavian countries, Britain, Ireland and France – while ‘lowest low’ is found along the south and east of Europe – the Mediterranean countries and some of the former communist countries of Central and Eastern Europe. The countries lying geographically between these two extremes are in the intermediate ‘very low’ fertility band, with Turkey in the south east forming an exceptional case of relatively high fertility. The family formation patterns from which these fertility patterns emerge also differ across Europe and indicate that there are different ways of arriving at similar low fertility outcomes. For example, delayed childbearing is often cited as a contributor to very low fertility (Billari 2005) but in eastern Europe countries, where fertility is at the bottom of the range, women give birth at a younger age than they do in northern European countries such as Sweden, Finland, Ireland, Britain and France where fertility is higher (United Nations 2003: 51). A rise in the incidence of childlessness is an important contributor to the emergence of low fertility in some countries (such as Germany, where over 25 per cent of the 1960 birth cohort remained childless compared to less than 10 per cent of birth cohorts of the late 1930s) but not in others (such as France, where the birth cohorts around 1960 had less childlessness than those of the 1930s) (United Nations 2003: 66). One reasonably pronounced feature of the pattern of lowest low fertility found in the former communist new member states is that it is relatively recent (Billari 2005: 59). In 1990, the total fertility rate in countries like Poland (2.20), the Czech Republic (1.90) and Hungary (1.87) was more or less similar to that of the UK (1.83) and France (1.78) but by 2003, the Czech Republic had the lowest fertility rate of the present EU countries (1.18), while Poland (1.22) and Hungary (1.28) were not much higher. Those countries thus made the transition from low to ‘lowest low’ fertility quickly and recently, reflecting an impact on family formation associated with the transition from communism. Italy and Spain, by contrast, countries with similarly low fertility today, were already close to their current position by 1990 (with total fertility rates in that year of 1.33 and 1.36 respectively).

2. Data The key indicators focused on in this chapter consist of responses to the following questions in surveys of the adult population in Europe (these questions produce the variables as labelled here in brackets): •

What is the ideal number of children for a family? (general ideal number of children)

30 • •

Tony Fahey What is the ideal number of children for you personally (personal ideal number of children) Have you had any children? If yes, how many? (actual number of children)

These three questions were asked in Eurobarometer 56.2 carried out in EU member states (EU-15) in 2001 and the Candidate Country Eurobarometer carried out in 2002 in what were then the 10 accession states (now the new member states – NMS) and the three candidate countries, Bulgaria, Romania and Turkey (CC-3). The common variables from these two surveys were compiled into the single 28-country dataset used here by the Social Science Research Centre, Berlin (WZB), as part of a project on living conditions in Europe carried out for the European Foundation for the Improvement of Living and Working Conditions in 2003 (Alber and Fahey 2004). This combined dataset is the principal data source used in this chapter. The main focus is on the second and third of the three variables listed above (personal ideal number of children and actual number of children), since they allow us to examine the gap between people’s personal family size preferences and their actual family size for 28 European countries. The chapter also draws on two further data sources – the European Values Study (EVS) surveys carried out in 1981 and 1990. Eight countries are common to both of these surveys and to Eurobarometer 56.2 mentioned earlier. For these countries, the first variable listed above (general ideal number of children) is measured at each of the three time points. This variable is of limited interest for our purposes because it taps into general social norms about family size rather than personal preferences. However, it has value because it provides a rare trend measure of family size ideals for the two-decade period between 1981 and 2001–2002. It is examined briefly below on that account. 2.1 Classification by age In examining the gap between ideal and actual fertility across adults of all ages in cross-sectional data, it is necessary to distinguish between what might be termed the ‘final’ gap between ideal and actual fertility among those whose childbearing is completed and the interim gap that arises among younger adults who may yet have more children. The final gap is more important than the interim gap for long-term demographic outcomes and therefore will occupy most of our attention here. For this reason also, the present analysis concentrates on women, since in their case it is possible to identify an age at which, in biological terms, childbearing can be said to be complete. Men’s fertility is not age limited in the same way and so it is more difficult to speak of completed fertility in their case. In much of the analysis below, we concentrate on the youngest possible segment of women who could be said to have completed fertility, while at the same maintaining reasonable sample size at the country level in the data at our disposal. We therefore frequently focus on women with completed fertility, defined as those who are aged 40–64 and who, if they are aged under 50, have said they plan to have no more children. Even though the age-range of this group is narrowed down as far as sample size considerations will allow, the childbearing time span of the women involved is quite wide. In the case of the 2001 data, for example, the oldest women in the age group 40–64 would have entered their childbearing years in the late 1950s, while the

Fertility patterns and aspirations

31

youngest would be arriving at the end of their childbearing years at around the time of the survey. The childbearing time span represented by these women, therefore, amounts to most of the second half of the twentieth century. This indicates the difficulty of linking cross-sectional data of this kind to short-term temporal patterns of fertility, and so drawing conclusions about time trends that might be of interest to policy.

3. The role of education The effect of educational level on the ideal-actual fertility gap is viewed here as a useful means of elucidating what that gap signifies. The measure of education level used is respondents’ age when they left full-time education. While this is a crude measure, it provides a serviceable basis for classifying education in a comparable way across countries. For the present analysis, we use a three-fold classification of the age at which people left full-time education: under age 16 years, between 16 and 19 years, and 20 years or over. The focus on education is adopted here partly for methodological reasons, reflecting the technical character of education as a proxy measure of socio-economic status that is stable over the adult life course and independent of fertility. In cross-sectional data such as are used here, education can be used as a measure of socio-economic background not only at the point of observation but also prior to or during a person’s childbearing years: for the most part, people complete their education before they begin childbearing, and their educational level does not change as their family formation proceeds. There may be exceptions to this rule (e.g. young women who drop out of education to have children) but in the fertility regimes found in Europe in the latter part of the twentieth century, these exceptions are unlikely to be common enough to render invalid the assumption that education is independent of fertility. Other possible measures of socio-economic background, such as current employment status, occupation and income, are not independent of fertility in the same way since they are subject to change through the family cycle and in particular may be affected by the number or timing of children that a person might have. In addition to the methodological significance of education, it is also important because, as Cleland (2003: 187) states, ‘education of adults consistently emerges as the single most powerful predictor of their demographic behaviour’. In less developed societies, women’s education first causes a short-term rise in fertility, because of increased fecundity, reduced risk of foetal death, and the decline of traditional practices such as prolonged breastfeeding and postpartum abstinence (United Nations 1995: 23). As societies begin to develop, education causes fertility to fall both because of individual-level effects on resources and incentives and community-level effects on cultural norms regarding family size (Caldwell 1980; Castro Martin 1995). As societies make the transition to very low fertility, fertility differentials by education get narrower at the individual level (United Nations 1995: 23; Cleland 2003), while at the aggregate level the relationship changes direction – societies with higher levels of female education come to have marginally higher fertility rates than those with lower levels of education (Sleebos 2003: 19-20). The latter pattern echoes the recent reversal of the female employment–fertility relationship found in industrialised societies. Where low female employment was associated with relatively high fertility in industrialised societies in the 1960s and 1970s, the relationship reversed during the

32

Tony Fahey

1980s and low female employment came to be associated with very low fertility. Today, the industrialised countries with the lowest fertility now generally also have low rates of female labour force participation (Brewster and Rindfuss 2000; Sleebos 2003: 20; Ahn and Mira 2002). There is a paradox here because at the individual level, women with higher education and with stronger attachment to the labour force continue to have smaller families (though this effect is not strong in advanced industrial societies – Cleland 2003). In other words, when we look at the situation within countries, we find that women with weaker labour market and educational profiles have somewhat larger families, but when we compare developed countries with each other we find that those that enhance the overall labour market and educational profiles of the female population have the higher birth rates (Castles 2003). Our concern here is with the effect of education on the ideal–actual fertility gap. This effect can be broken down into two components: the effect on the ideals themselves (whether the better educated have larger or smaller ideal family sizes than the less educated) and on the degree to which those preferences are fulfilled (whether the better educated are more or less likely to attain their ideal family size). Lower family sizes among better educated women could be the result of either of these two components: better educated women might prefer smaller families or they might be less likely to attain whatever family size they might prefer. An effect that operates through a preference for smaller families would suggest that the cultural influence of education on fertility ideals and aspirations is the important causal mechanism, while an effect that operates through reduced attainment would point to opportunity costs and the problems of reconciling work with motherhood as the key influences. In the latter instance, better educated women might have more resources in an absolute sense, but the scope for careers, income and other uses of time open to them could be experienced as an obstacle to motherhood and a constraint not faced to the same degree by less educated women. The paradox here is that the range of options open to better educated women might confront them with choices that less educated women do not face. As a result they might experience these options as sources of strain and tension that might make motherhood seem less rather than more possible.

4. The ideal–actual gap We first consider the gap between personal ideal family size and actual family size among women in 28 European countries in 2001–2002. As well as looking at women with completed fertility as defined above (i.e. those aged 40–64), we also consider in this section women still in their childbearing years (aged 18–39) and older women (those aged 65 and over) as an indirect means of assessing trends over time (see the appendix (Table A1.1) for details of sample numbers in each age category). In the case of the two older age groups of women (those aged 40–64 and 65 and over), childbearing can be assumed to be complete and family size is measured by women’s own reports of their actual number of children. In the case of the women aged 18–39, among whom further childbearing is likely to occur, the relevant comparable variable is eventual completed family size rather than current family size. As this variable is not directly measurable, we use the total fertility rate (TFR) in 2000 as a proxy (the TFR is the number of births a woman would have during her reproductive life if she

Fertility patterns and aspirations

33

had children at the same rate as the various childbearing age groups in a particular year). The TFR is unlikely to match exactly the eventual completed family size of the age groups on which it is calculated but it provides the best available approximation to a final outcome measure and is used here for that reason. Table 1.1 sets out the data. The countries are ordered according to the ‘families of nations’ categorisation described in the introduction to this volume and are also clustered into the EU-15, the new member states (NMS) and the candidate countries (CC-3). 4.1 Ideal number of children We can look first at the differences in ideal number of children across the three age groups of women. In all countries except France and Turkey, ideal number of children declines consistently with age, with the differential on average between the oldest and the youngest of the three age groups being less than half a child. This is an indication of a definite, though not large, decrease in ideal family size over time. In France, one of the exceptional cases, there is no real differential. The mean ideal family size for women aged 65 and over in France is 2.52. This falls to 2.48 for women aged 40–64 but rises to 2.54 for women aged 18-39. In Turkey, the other exception (which, as we shall see further below, differs from the other countries in more ways than just this), the youngest age group has a larger ideal number of children than the oldest (2.24 versus 2.07) but taking the possibility of sampling error into account, the difference is not very great. It is worth noting that among the youngest age-group, ideal number of children exceeds 2.1 (the conventional definition of the replacement fertility rate) in 20 of the 28 countries, even though the actual fertility rate as measured by the TFR in 2000 is below 2.1 in all countries bar Turkey. Although sub-replacement fertility outcomes are now found in all of the countries bar Turkey, sub-replacement fertility ideals among younger women have not become the norm, contrary to what Goldstein et al. (2003) suggest may be happening. In many of the countries with sub-replacement ideals, those ideals are only barely below the 2.1 threshold, though in Germany and Austria, where ideal number of children among 18–39-year-old women is 1.75 and 1.7 respectively, they are quite far below. Romania, at 1.91, is also quite low. (The low figure for Malta – 1.91 – has to be viewed with caution as the sample N on which it is based is small – see Table A1.1 in the Appendix.) Further breakdowns of these data presented in Fahey and Spéder (2004: 30) show that younger women in Germany and Austria have larger-than-normal minorities who choose ‘none’ as their personal ideal number of children (17 per cent and 13 per cent of women aged 1839 in Germany and Austria respectively, compared to a weighted average of 5.5 per cent for the EU-15). A personal preference for having no children is thus held by a distinctively large minority of young women in Germany and Austria. The main contrast with Germany and Austria is provided by the Scandinavian countries, Ireland, the UK and France, in all of which ideal number of children among women aged 1839 exceeds 2.4. In this instance, there is as much contrast within the EU-15 as there is between the EU-15 and the NMS, thus reinforcing the point made earlier that contrasts between east and west are not a dominant feature of the fertility-related landscape of the current EU.

Scandinavian Denmark Finland Sweden Western Ireland UK Continental Austria Belgium France Germany Luxembourg Netherlands Mediterranean Greece Italy Portugal Spain

2.43 2.42 2.44 2.43 2.43 2.61 2.42 2.09 1.7 2.2 2.54 1.75 2.16 2.1 2.16 2.28 2.14 2.1 2.16

1.65 1.76 1.73 1.54 1.65 1.89 1.64 1.57 1.32 1.65 1.89 1.34 1.78 1.72 1.27 1.3 1.25 1.54 1.22

0.78 0.66 0.71 0.89 0.77 0.72 0.78 0.5 0.38 0.55 0.65 0.41 0.38 0.38 0.89 0.98 0.89 0.56 0.94

2.51 2.45 2.53 2.52 2.49 3.02 2.42 2.23 2.19 2.29 2.48 2.04 2.28 2.34 2.42 2.84 2.33 2.52 2.4

2.01 1.94 2.14 2.03 2.37 3.07 2.33 1.99 1.91 1.91 2.31 1.85 1.91 2.1 2.04 2 1.86 2.61 2.25

Actual number of children (mean)

Personal ideal number of children (mean)

Ideal– TFR

Personal ideal number of children (mean)

Total fertility rate, 2000

Women aged 40–64

Women aged 18–39

0.5 0.51 0.39 0.49 0.12 −0.05 0.09 0.24 0.28 0.38 0.17 0.19 0.37 0.24 0.38 0.84 0.47 −0.09 0.15

Ideal– actual

2.63 2.58 2.95 2.51 2.88 3.84 2.83 2.37 2.41 2.4 2.52 2.2 2.43 2.74 2.74 3.04 2.75 2.66 2.65

Personal ideal number of children (mean)

Women aged 65+

Table 1.1 The ideal–actual fertility gap in three broad age groups of women in European countries, 2001–2002

2.36 2.26 2.72 2.22 2.73 3.67 2.67 2.53 2.59 2.77 3.05 2.11 1.86 2.83 2.6 1.87 2.54 2.86 2.83

Actual number of children (mean)

0.27 0.32 0.23 0.29 0.15 0.17 0.16 −0.16 −0.18 −0.37 −0.53 0.09 0.57 −0.09 0.14 1.17 0.21 −0.2 −0.18

Ideal– actual

2.16 2.65 1.91 2.05 2.11 2 2.07 2.16 2.1 2.09 2.18 2.06 2.08 2.17 2.09 1.91 2.24 2.11 2.08 2.17

0.72 0.82 0.01 0.8 0.82 0.86 0.74 0.82 0.9 0.78 0.79 0.82 0.75 −0.02 0.84 0.61 −0.26 0.62 0.78 0.56

0.496* (excl. Turkey: 0.588**)

1.45 1.83 1.9 1.25 1.29 1.14 1.33 1.34 1.2 1.31 1.39 1.24 1.33 2.08 1.25 1.3 2.5 1.49 1.30 1.61

2.48 3.2 1.98 2.31 2.28 2.14 2.21 2.36 2.32 2.26 2.21 2.23 2.27 2.24 2.1 2.07 2.38 2.33 2.29 2.32

0.54 0.59 0.18 0.55 0.19 0.18 0.22 0.13 0.29 0.37 0.37 0.41 0.32 −0.29 0.18 0.18 −0.75 0.26 0.22 0.19

0.633* (excl. Turkey: 0.762**)

1.94 2.61 1.8 1.76 2.09 1.96 1.99 2.23 2.03 1.89 1.84 1.82 1.95 2.53 1.92 1.89 3.13 2.07 2.07 2.13

Note: For sample Ns, see Table A1.1 in the Appendix.

Sources: Eurobarometer 56.2 (2001); CC Eurobarometer 2002 data files; total fertility rate 2000: Eurostat

Correlation of ideal and actual

West NMS Cyprus Malta Slovenia Visegrad Czech Rep. Hungary Poland Slovakia Baltic Estonia Latvia Lithuania CC-3 Bulgaria Romania Turkey EU-15 NMS All countries

2.81 4.1 2.9 2.47 2.56 2.48 2.29 2.75 2.4 2.37 2.12 2.44 2.46 2.3 2.1 2.52 2.07 2.58 2.55 2.57

0.52 0.49 −0.12 0.6 0.08 0.28 0.26 −0.02 −0.28 0.45 0.43 0.65 0.34 −0.38 0.1 0.35 −2.01 0 0.12 0.22 0.492* (excl. Turkey: 0.734**)

2.29 3.61 3.02 1.87 2.48 2.2 2.03 2.77 2.68 1.92 1.69 1.79 2.12 2.68 2 2.17 4.08 2.58 2.43 2.55

36

Tony Fahey

4.2 The ideal–actual gap We now turn to the gap between ideal and actual fertility, looking again at the three age groups in Table 1.1. Among women aged 65 and over, actual number of children exceeded ideal number of children in 10 of the 28 countries, suggesting a widespread lack of adequate control over their own fertility by women in these countries. This excess was by far the widest in Turkey, where older women had quite large families on average (4.08 children) but had low ideal family sizes (2.07 children). It is striking to note that there was no similar excess of actual over ideal number of children in either Ireland or Cyprus, the two countries that were closest to Turkey in levels of actual family size among older women. In both these countries, women’s expressed preference was for even larger families than the relatively large families they actually had. Thus, where older women in Turkey experienced ‘traditional’ fertility outcomes while holding ‘modern’ family size ideals, their counterparts in Ireland and Cyprus were traditional in both outcome and ideal. Apart from these countries, however, the most common experience even among older women was that actual number of children fell short of the ideal. In most cases, the shortfall was quite small, but the very fact that it occurred so widely indicates that the present generation of women of childbearing age is not the first to fail to attain its ideal number of children. Moving on to women aged 40–64, we find that ideal number of children was almost universally lower than among those aged over 65 and the excess of actual over ideal number of children had disappeared in all but three countries (Ireland, Portugal and Turkey). This suggests that effective contraceptive practice had become more widely established among women of this generation in a larger number of countries. We will examine the ideal–actual gap for this group of women in more detail further below. The gap between ideal and actual fertility among the youngest group of women – those aged 18–39 – strictly speaking cannot be compared that of the two older age groups since, as mentioned earlier, their childbearing is not yet complete and we therefore use a different measures of actual fertility in their case (the total fertility rate in 2000). The gap between ideal family size among 18–39-year-old women and the TFR in 2000 is wider than the ideal-actual gap among 40–64-year-olds in all countries bar Malta and Turkey. Although we cannot be sure whether this indicates a real increase in the extent of unfulfilled fertility aspirations among younger women or is an artefact of differences in measurement, the most plausible interpretation is that a real widening of the gap between ideal and actuality has occurred.

5. Composition of the ideal–actual gap The overall gap between ideal and actual fertility for any category of the population has to be interpreted carefully since it is a composite arrived at by averaging quite different outcomes. These outcomes can be classified as under-attainment (having ‘too few’ children relative to one’s ideals), over-attainment (having ‘too many’ children) and ‘just right’ outcomes (where the ideal and actual number of children match). Even in a country where there was no gap on average between ideal and actual numbers of children, substantial proportions of women could have family sizes that either exceeded or fell short of their ideals.

Fertility patterns and aspirations

37

Figure 1.1 summarises the distributions across these three categories for women with completed fertility in the age group 40–64. In the majority of countries, the ‘just right’ outcome was achieved by between half and two-thirds of women aged 40–64, which means that the women who had other outcomes amounted to significant minorities. Turkey is an exception, in that the ‘just right’ outcome was achieved only by a minority (33 per cent) of Turkish women in this age group. In most cases, as we would expect from findings already presented, under-attainment of family size was the most common other outcome: in the EU-15 and NMS, onethird of women on average had ‘too few’ children relative to their ideal. Greece had particularly high levels of such under-attainment (56 per cent), though the sample on which this estimate for Greece is based is somewhat small (see Table A1.1 in the Appendix). Turkey had particularly low levels of under-attainment (15 per cent). In some countries with especially low fertility (such as Spain), the proportion of women whose number of children fell below their ideal was not especially large. Alongside such under-attainment, there were smaller but significant levels of overattainment (‘too many’ children relative to the ideal). In the EU-15 and NMS, 10 to 11 per cent of women aged 40–64 had excess fertility in this sense. Again, Turkey was an exception: here excess fertility was the most common situation and was present among half of women. The level of excess fertility found in Turkey was three to four times greater than that found in most other countries. Although excess fertility was a dominant feature only in Turkey, its widespread if less prominent presence in other countries is notable. It indicates that even in countries where on average actual family size falls short of the ideal, there can be significant minorities whose family size exceeds the ideal.

6. Influence of education We now turn to differences in the ideal–actual fertility gap by educational level as a means to shed further light on what that gap signifies. Table 1.2 examines the relationship between ideal and actual family size for women with completed fertility, classified by educational level. If we look first at ideal family size, we find that differences by educational level are absent or slight in both the EU-15 and the NMS and become substantial only in the CC-3. In the EU-15, for example, the lowest and highest educational categories have the same ideal number of children (2.46), while in the NMS a difference between the two is present but slight (2.39 among the less educated compared to 2.20 among the best educated). Substantial differences in ideal number of children by educational level are evident only in the CC-3 (2.33 among the less educated compared to 1.93 among the best educated). Separate analysis not shown in the table indicates that the distinctiveness of the CC-3 in this regard is more or less wholly driven by Turkey, where the gap in ideal family size between the least and best educated is wide (2.41 compared to 1.61). Apart from the Turkish case, therefore, Table 1.2 suggests that the effect of education on ideal family size is weak. This is a finding of some interest, as it suggests that whatever influence education might have on women’s fertility behaviour, it does not operate primarily through an effect on family size preferences. We get an indication of where education does make a difference when we look at the column for actual family size in Table 1.2. Here it is clear that differences by educational level are substantial and are present in each of the three regions: the higher

38

Tony Fahey Greece

56

Cyprus

43

Sweden

43

Italy

39

Denmark

38

Slovenia

41

3

43

14 53

5

51

10

56

38

5

57

5

Latvia

36

57

7

Estonia

36

57

7

Belgium

35

France

35

53

12

49

16

Luxembourg

33

50

17

Slovakia

33

51

16

31

Hungary

58

11

Poland

29

57

14

Finland

28

60

11

Lithuania

28

Czech Rep.

28

Romania

28

Portugal

26

Netherlands

25

Malta

25

Spain

25

Austria

25

Great Britain

24

Ireland

22

Bulgaria

22

8 15

61

11

56

18

63

12

50

25 62

13

66

10

60

24

Germany

Turkey

64 57

16

64

12

56

22 67

16

11

33

51

EU-15

32

56

12

NMS

32

57

11

CC-3

22 Too few

52 Just right

26 Too many

Figure 1.1 Relationship between actual number of children and personal ideal number of children among women with completed fertility (ages 40–64) Sources: Eurobarometer 56.1 (2001), CC Eurobarometer 2002 Note: Countries ordered in descending importance of ‘Actual children less than ideal’.

Fertility patterns and aspirations

39

the educational level, the smaller the actual number of children. In the EU-15, for example, women with completed families who finished their education at 15 years or younger had 2.34 children on average, while women who finished their education at 20 years or older had 1.81 children on average, a differential of 0.53 children. The difference in family size by education level is somewhat wider in the NMS, where the less educated have 2.5 children and the better educated 1.68 children (a differential of 0.82). It is wider still in the CC-3 (at 1.34 children). This lower level of childbearing among better educated women, when taken in conjunction with the small or absent differences in family size ideals by educational level noted earlier, means that the better educated have a larger gap between actual and ideal family size than do the less educated. In the EU-15, for example, the shortfall of the actual from the ideal number of children is five times greater among the best educated compared to the least educated (0.65 compared to 0.12). This ratio is somewhat greater in the NMS and greater again in the CC-3. As before, we need to be conscious that the gaps between ideal and actual fertility reported in Table 2 are composites of under-attainment (‘too few’), over-attainment (‘too many’) and ‘just right’ outcomes. Table 1.3 shows the distribution of these outcomes by age at which respondents left school in the EU-15, the NMS and the CC-3. There is a striking similarity between the better educated in all three regions in the degree to which they are likely to have ‘too few’ rather than ‘too many’ children: 41–2 per cent of women in this educational category have ‘too few’ children, while only 5–8 per cent (depending on the region) have ‘too many’. The pattern is quite different, and is less uniform across regions, for those in the lowest educational category. In the EU-15, the less educated are less likely to have ‘too few’ children than the better educated (26 per cent versus 41 per cent) and they are more likely to have ‘too many’ (16 per cent versus 8 per cent). Nevertheless, in this region the ‘too few’ outcome is more common than the ‘too many’ outcome even for the less educated. In the NMS, by contrast, the balance among the less educated women between those with too few and too many is reversed: those with too many (22 per cent) slightly exceed those with too few (20 per cent). This reversal is even more pronounced in the less educated category in the CC-3, where women with ‘too many’ are more than twice as numerous as those with ‘too few’ (38 per cent vs. 17 per cent). Thus, the likelihood that those with low education will over-attain their fertility ideals increases as we move from the EU-15 to the NMS and to the CC-3. Further breakdowns of the data Table 1.2 Mean personal ideal number of children and actual number of children by schoolleaving age among women with completed fertility (ages 40–64) in the EU-15, NMS and CC-3 EU-15

NMS

CC-3

School leaving age

Ideal Actual Gap: I-A Ideal Actual Gap: I-A Ideal Actual Gap: I-A

Up to 15 years 16-19 years 20+ years All

2.46 2.20 2.46 2.34

2.34 1.98 1.81 2.07

0.12 0.22 0.65 0.27

2.39 2.26 2.20 2.29

2.50 1.97 1.68 2.07

Sources: Eurobarometer 56.1 (2001); CC Eurobarometer 2002

−0.11 0.29 0.52 0.22

2.33 2.05 1.93 2.20

2.73 1.93 1.39 2.32

−0.4 0.12 0.54 −0.12

40

Tony Fahey

Table 1.3 Fertility ideal attainment and school-leaving age among women with completed fertility (ages 40–64) Fertility ideal attainment (%) School leaving age

Too few

Just right

Too many

26 28 41

58 59 52

16 12 8

100 100 100

20 31 42

58 58 53

22 11 6

100 100 100

17 24 41

45 64 55

38 12 5

100 100 100

20 12 7

56 60 51

24 29 42

100 100 100

EU-15 Up to 15 years 16-19 years 20+ years NMS Up to 15 years 1619 years 20+ years CC-3 Up to 15 years 16-19 years 20+ years All Up to 15 years 16-19 years 20+ years

Total (%)

Sources: CC Eurobarometer 2002; Standard Eurobarometer 56.1 (2001)

not shown here, however, indicate that these regional contrasts can be overstated, in that over-attainment among less educated women varies considerably within regions as well as between regions (in particular, in the case of less educated women with completed fertility, over-attainment is below 5 per cent in the Scandinavian countries but is in the region of 25 per cent in Britain and Ireland). In sum, these patterns show that under-attainment of fertility ideals is not evenly distributed within the female populations in European countries but is particularly characteristic of better-educated women. It occurs among the less educated also, but less frequently and is counter-balanced to some degree among the less educated by substantial levels of over-attainment.

7. Trends since 1981 As a check on the patterns just looked at, we can briefly examine some trend data on ideal and actual family size since 1981 for a sub-set of European countries. As indicated earlier, the measure of ideal family size that is available across time relates to ‘general’ rather than ‘personal’ ideals (that is, to what people think is desirable for families in general rather than for themselves personally). The countries for which this measure is available are eight in number. The data are presented first for all adults for the eight countries in Table 1.4. These data broadly confirm the picture derived from the age comparisons for 2001–2002 set out earlier, keeping in mind that the latter relate to personal rather than general ideal number of children. There is a modest decline in general ideal family size between 1981 and 2001, with a fall in most countries from around 2.5 to something in the region of 2.2 or 2.3. Ireland again is an outlier: it had an ideal family size of 4 in 1981, and this fell to 2.8 in 2001.

Fertility patterns and aspirations

41

1

Table 1.4 Mean general ideal family size 1981

1990

2001

2.6 2.5 2.2 2.4 2.6 2.5 2.6 3.5

2.4 2.3 2.0 2.3 2.2 2.1 2.3 2.8

1990 (35–44) 2.6 2.3 2.0 2.3 2.4 2.4 2.5 3.2

2001 (45–54) 2.4 2.4 2.0 2.3 2.4 2.1 2.4 2.9

All adults France Britain W. Germany Italy Netherlands Belgium Spain Rep. Ireland

France Britain W. Germany Italy Netherlands Belgium Spain Rep. Ireland

2.6 2.4 2.3 2.3 2.5 2.4 2.9 4.0 Age cohort of women 1981 (25–34) 2.6 2.3 2.1 2.2 2.4 2.4 2.6 3.7

Sources: 1981 EVS; 1990 EVS; Eurobarometer 56.1 (2001) Note: 1 Ideal number of children for a family.

West Germany is the only one of the eight countries which moves to a sub-replacement family size ideal, again echoing earlier findings concerning the emergence of subreplacement fertility ideals in Germany. The lower panel in Table 1.2 presents the trend data for an approximate age cohort of women – those aged 25–34 in 1981, 35–44 in 1990 and 45–54 in 2001. The cohort data show that the sharp decline in family size ideals found in Ireland occurred within this age cohort across time and thus is not solely a consequence of changing attitudes from one cohort to the next. In the other countries, the change in ideals across time were generally slight and do not indicate a great deal of within-cohort change. Table 1.5 takes the pooled data for the eight countries shown in Table 1.4 and examines the incidence of under-attainment, over-attainment and ‘just right’ outcomes by educational level among women with completed fertility for 1981, 1990 and 2001. Keeping in mind that these outcomes are measured relative to general rather than personal family size ideals and relate only to eight countries, their significance should not be overstated. Nevertheless, it is of interest to note that they show no increase over time in the proportions of women at any educational level who have either ‘too few’ children or ‘too many’ children. Otherwise, the patterns confirm earlier findings: less educated women are more likely to have ‘too many’ children and less likely to have ‘too few’ children than better educated women.

8. Conclusions and implications The key empirical finding of this chapter is that under-attainment of family size preferences is a function of social advantage rather than disadvantage in Europe: well

42

Tony Fahey

Table 1.5 Fulfilment of general ideal number of children among women with completed fertility (ages 45–64), pooled data for 8 European societies, 1981, 1990, 2001 School leaving age (%)

1981 Too few Just right Too many Total 1990 Too few Just right Too many Total 2001 Too few Just right Too many Total

15 yrs or earlier

16-19

20+

All

No. of cases

37 39 24 100

41 39 20 100

44 46 10 100

39 39 21 100

704 533 311 1548

33 42 25 100

40 41 19 100

46 40 14 100

36 41 22 100

730 853 446 2029

25 52 24 100

33 44 23 100

43 47 10 100

32 47 20 100

341 662 172 1175

Sources: European Values Study 1981, 1990; Eurobarometer 56.1 (2001)

educated women are more likely to fall short of the number of children they would want than are less educated women. There is a slight effect of education on the family size ideals themselves in some countries (better educated women prefer slightly fewer children than less educated women) but the main effect of education operates on the attainment of those ideals rather than on the ideals themselves: better educated women have substantially fewer children than less educated women, even where their ideals are broadly similar. A less substantial though nonetheless notable pattern is that small but significant minorities of less educated women have too many children: their family size outcomes exceed their preferences, even in societies where on average women fall short of their preferences. These patterns do not divide along east– west lines in the enlarged EU. Differences within the EU-15 are as great as those between the EU-15 and the NMS. Turkey is sharply differentiated from all the other countries in that Turkish women still typically have more children than they would prefer. But apart from that, consistent regional variations in patterns of under-attainment and over-attainment are hard to identify in the EU. These empirical findings have a number of implications for policies that European governments might contemplate if they wish to boost fertility rates. First, they confirm that such policies might usefully focus on the constraints that prevent women from attaining their ideals rather than on the ideals themselves. Women in Europe do not need to be persuaded of the desirability of having more children – apart, perhaps, from countries such as Germany and Austria where women’s preferred family size has fallen to very low levels. Rather, they need to be facilitated in having the number of children they desire. Our findings do not tell us what kind of interventions would be most effective in achieving that end – nor even whether macro-economic measures designed to raise employment rates would be better than family-oriented measures such as child

Fertility patterns and aspirations

43

support payments, parental leave, improved childcare provision, or more flexible work practices. But they do tell us what categories of women should be the primary target of whatever interventions are chosen – namely, better educated women. These are the women most likely to fall short of their ideal number of children, presumably because the opportunity cost of childbearing is greatest in their case. While we have not tested the likely effectiveness of pronatalist interventions here, a targeting of such interventions on better educated women would probably require that the resources deployed be large, since if interventions were to be effective, their cost would have to be in proportion to the earning power and value of time (that is, to the opportunity costs of childbearing) of the segment of the population they would seek to influence. This would be so no matter what specific form the interventions would take (as between cash payments, maternity leave, etc.). It costs more for a highly educated, highly skilled woman to be given maternity leave, or to have flexible working hours, or to be provided with cash supports amounting to a meaningful share of household income, than for a less educated woman to be given supports with a similar relative value. There is an additional reason for arguing that the efficiency of pronatalist measures, evaluated in strict pronatalist terms, would be enhanced by targeting them on better educated women. This is that while less educated women also often fail to attain their ideal number of children, there are many countries where the mismatch in their case is as likely to arise in the form of having too many children as having too few. If resources were used to enable less educated women to attain their family size ideals, therefore, the result would consist of downward as well as upward movement in fertility. In many countries the downward movement would be large enough fully to counterbalance upward movement. While downward movement for women with more children than they want might be desirable for their welfare and that of their families, it would do nothing to raise fertility rates and would therefore fail to serve pronatalist objectives. The problem with pronatalist policies targeted on educated women is that, while they might work in raising fertility, they would also be socially regressive: they would give most to those who already have most. They would therefore be difficult to reconcile with the traditional socially progressive logic of family support, where the aim is to enhance child welfare and living standards among less well-off families. There is a less oppositional way of posing the alternatives: pronatalist policy could be viewed in terms of horizontal equity, where the aim is to transfer resources from those without children to those with children, rather than in terms of vertical equity, where the aim is to transfer resources from the well-off to the less well-off. (Income tax reliefs for families with children are one possible mechanism of horizontal distribution, especially in systems where income taxes in general are progressive. In those instances, income tax reliefs for children would provide the greatest benefits to high income earners with children, would impose the highest costs on high income earners without children, and would provide little or no costs or benefits on low income earners, whether with children or without.) Nevertheless, even if pronatalist measures were designed along horizontal equity lines, they would reduce the overall package of resources available for downward distribution (for example, income tax reliefs for high earners with children, while not imposing direct costs on low income families with children, would reduce nevertheless reduce the resources available for distribution to such families). It would therefore be difficult to design effective pronatalist measures that would not weaken the possibilities for socially progressive distribution, even if they did not obviate them entirely.

44

Tony Fahey

Table A1.1 Sample Ns for female samples by countries and age groups, combined Eurobarometer dataset, 2001–2* Female sample

Denmark Finland Sweden Ireland Great Britain Austria Belgium France Germany Luxembourg Netherlands Greece Italy Portugal Spain Cyprus Malta Slovenia Czech Rep. Hungary Poland Slovakia Estonia Latvia Lithuania Bulgaria Romania Turkey Total

Total sample

18-39

40-64

65+

All women

1000 1003 1000 1001 1312 999 1007 1005 2007 604 999 1002 999 1001 1000 500 500 1002 1000 1020 2000 1067 1010 1000 1015 1000 1049 2000

205 227 204 267 311 228 219 248 394 143 255 215 216 222 241 99 55 246 237 195 465 252 218 235 211 173 218 747

189 198 175 166 250 199 185 175 421 109 186 198 189 188 180 138 84 227 223 255 497 267 211 220 221 268 236 228

89 121 124 91 129 105 113 86 262 34 62 88 120 143 94 35 50 122 85 163 139 113 117 62 77 165 120 30

483 546 503 524 690 532 517 509 1077 286 503 501 525 553 515 272 189 595 545 613 1101 632 546 517 509 606 574 1005

30102

6946

6083

2939

15968

*Sources: Combined from Eurobarometer 56.1 (2001); CC Eurobarometer 2002

References Ahn, N. and Mira, P. (2002) ‘A note on the changing relationship between fertility and female employment rates in developed countries’, Journal of Population Economics, 15: 667–682. Alber, J. and Fahey, T. (2004) Perceptions of Living Conditions in an Enlarged Europe, Luxembourg: Office for Official Publications of the European Communities. Online. Available Http: (accessed February 2007). Bernardi, F. (2005) ‘Public policy and low fertility: rationales for intervention and a diagnosis for the Spanish case’, Journal of European Social Policy, 15, 2: 123–139. Billari, F. (2005) ‘Europe and its fertility: from low to lowest low’, National Institute Economic Review, 194, 1: 56–73. Bongaarts, J. (1998) ‘Fertility and reproductive preferences in post–transitional societies,’ Policy Research Division Working Paper, no. 114, New York: Population Council. Bongaarts, J. (2002) ‘The end of the fertility transition in the developed world’, Population and Development Review, 28, 3: 419–443.

Fertility patterns and aspirations

45

Brewster, K.L. and Rindfuss, R.R. (2000) ‘Fertility and women’s employment in industrialized nations’, Annual Review of Sociology, 26: 271–296. Caldwell, J.C. (1980) ‘Mass education as a determinant of the timing of the fertility decline’, Population and Development Review, 6, 2: 225–255. Caldwell, J.C. (2002) ‘Policy responses to low fertility and its consequences: a global survey’, Journal of Population Research, May 2002. Online. Available Http: (accessed February 2007). Castles, F.G. (2003) ‘The World turned upside down: below replacement fertility, changing preferences and family-friendly public policies in 21, OECD countries’, Journal of European Social Policy, 13, 3: 209–227. Castro Martin, T. (1995) ‘Women’s education and fertility: results from 26 demographic and health surveys’, Studies in Family Planning, 26, 4: 187–202. Chesnais, J.-C. (1998) ‘Below-replacement fertility in the European Union (EU-15): facts and policies, 1960-1997’, Review of Population and Social Policy, 7: 83–101. Cleland, J. (2003) ‘Education and future fertility trends with special reference to mid-transitional countries’, pp. 187–202, in United Nations Population Division of the Department of Economic and Social Affairs (ed.), Completing the Fertility Transition, New York: United Nations. Online. Available Http: (accessed February 2007). Council of Europe (2005) Recent Demographic Developments in Europe 2004, Strasbourg: Council of Europe Publications. d’Addio, A.C. and Mira d’Ercole, M. (2005) Trends and Determinants of Fertility Rates: The Role of Policies, OECD Social, Employment and Migration Working Papers, Paris: OECD. European Commision (2005) Confronting Demographic Charge: A New Solidarty between Generations, Green Paper, Communication from the Comission Com (2005) 94, final, Brussels: Commision of the European Communities. Eurostat (2002) European Social Statistics: Demography, Luxembourg, Office for Official Publications of the European Communities. Fahey, T. and Spéder, Z. (2004) Fertility and Family Issues in an Enlarged Europe, Luxembourg: Office for Official Publications of the European Communities. Online. Available Http: (accessed February 2007). Freedman, R., Freedman, D.S. and Thornton, A. (1980) ‘Changes in fertility expectations and preferences between 1962 and 1977: their relation to final parity’, Demography, 17, 4: 365–378. Gauthier, A.H. and Hatzius, J. (1997) ‘Family benefits and fertility: an econometric analysis’, Population Studies, 51: 295–306. Goldstein, J., Lutz, W. and Testa, M.R. (2003) ‘The emergence of sub-replacement family size ideals in Europe’, European Demographic Research Papers 2003, no. 2, Vienna Institute of Demography in collaboration with the European Observatory on the Social Situation, Demography and the Family. McDonald, P. (2000) ‘Gender equity, social institutions and the future of fertility’, Journal of Population Research, 17, 1: 1–16. New Cronos (2005) Eurostat on-line database. Online. Available Http: (accessed February 2007). Pearse, D. (1999) ‘Changes in fertility and family size in Europe’, Population Trends, no. 95, London: Office for National Statistics. Schoen, R., Ashtone, N.M., Kim, Y.J., Nathanson, C.A. and Fields J.M. (1999) ‘Do fertility intentions affect fertility behaviour?’, Journal of Marriage and the Family, 61: 790–799. Sleebos, J.E. (2003) Low Fertility Rates in OECD Countries: Facts and Policy Responses, OECD Social, Employment and Migration Working Paper no. 15, Paris: OECD.

46

Tony Fahey

Thomson, E. and Brandreth, Y. (1995) ‘Measuring fertility demand’, Demography 32, 1: 81–96. Thomson, Elizabeth (1997) ‘Couple childbearing desires, intentions, and births’, Demography, 34, 3: 343–354. United Nations (2003) Partnership and Reproductive Behaviour in Low-fertility Countries, ESA/P/WP.177, New York: United Nations. United Nations Population Division (1995) Women’s Education and Fertility Behaviour: Recent Evidence from the Demographic and Health Survey Series, New York: United Nations. Van de Kaa, D. (1998) ‘Postmodern fertility preferences: from changing value orientation to new behaviour’, Working Papers in Demography, no. 74, Australian National University, Canberra. Van Peer, C. (2002) ‘Desired and realized fertility in selected FFS countries’, pp. 117–142, in M. Macura and G. Beets (eds), Dynamics of Fertility and Partnership in Europe: Insights and Lessons from Comparative Research, Vol. 1, New York: United Nations.

2

Patterns of family living in the enlarged EU Chiara Saraceno

Introduction: A history of diversity Patterns of family formation and family living have a long history of diversity throughout Europe (e.g. Reher 1998; Therborn 2004). Diversity in marriage rates, age at marriage, family structures, fertility, gender relations and intergenerational power relations continues to exist. These historical differences shape patterns of change and the impact of external pressures on existing family arrangements. To some degree, they also inform the way in which needs are understood and how responsibility for these needs is allocated between individuals, families and other social actors, including the welfare state. As a consequence, difference, rather than convergence, seems to be the persistent feature of family arrangements even within the limited space of Europe. An analysis performed for the Council of Europe in 2001, for instance, concluded that ‘The results of this analysis show that some demographic patterns are diverging (fertility quantum, fertility and marriage timing); others are geographically unchanged (divorce and out-of-wedlock births) even if their levels are different, and there is no tendency towards convergence’ (Pinnelli et al. 2001: 16). In their overview of changes in family patterns across Europe since the sixteenth century, Barbagli and Kertzer (2003) argue that at the beginning of the twentieth century there remained enormous differences in patterns of family formation among the various regions of the European continent. According to their reading, the twentieth century, differently from the preceding ones, was actually marked primarily by convergence. Yet, at the end of the twentieth century, although differences had diminished, substantial differences remained in the ways in which ‘families are formed, transformed and divided, and in the relations of family members and of more distant kin’ (Barbagli and Kertzer 2003: xxxviii). Differences in particular involve the role of marriage and its relationship to both sexuality and fertility, fertility rates, patterns of intergenerational exchange, gender relations and arrangements. These differences underline what Therborn (2004: 11), looking at the world as a whole, defined as ‘geo-cultures’: ‘To view family systems as geo-cultures means to treat them as institutions or structures taking their colouring from customs and traditions, from the history of a particular area, a cultural wrapping which may remain after structural, institutional change, leaving imprints in the new institution.’ Thus, for instance, the delay in leaving the parental home in Italy is certainly a contemporary device used by the young and their families to deal with a weak welfare state and a rigid housing market. But this behaviour has its roots in a historical pattern of

48

Chiara Saraceno

family formation according to which many Italian men in the centre-north regions until the first decades of the twentieth century never exited the parental household. If and when they married, they brought their wife into the parental household (Barbagli et al. 2003). Many authors have offered their own reading of this diversity. In a famous essay, John Hajnal (1982) identified two broad areas within Europe, on the basis of three criteria: prevalence of the nuclear, rather than the complex, family; degree of universality of marriage; and age at marriage for women. East of the conceptual line that runs from St Petersburg to Trieste, in what Hajnal called the Eurasian marriage pattern, the incidence of complex households has been substantial for many centuries. In these countries marriage also was virtually universal, and the age at marriage for women was lower than in western European countries. The complex household was almost absent in western European countries, particularly in Scandinavia, the Netherlands, England and northern France. It was present – mostly in the form of the stem family, as observed by Frédéric Le Play (1875), in southern Austria, some parts of Germany and in the rural areas of northern and central Italy, but not in southern Italy. More recently, David Reher (1998) has pointed to another aspect of diversity: the relevance of, and embeddedness in kinship ties. In the Mediterranean countries, irrespective of the household structure (nuclear, stem or extended), households traditionally have been embedded in and dominated by a dense kinship network to a degree very different from that found in northern European households, even in the past. According to Reher, within Europe at least three general patterns of family formation and arrangement may be found up to the twentieth century. The first was prevalent in the central-western and Nordic countries. Here, individuals married late, at a comparatively lower rate, and had fewer children; moreover, couples were relatively unstable and had loose ties with their kinship network. The second was prevalent in some parts of southern Europe. Here, households also were nuclear, but people married earlier and at a relatively higher rate, households were embedded in dense kinship networks. The third pattern was prevalent in eastern European countries and the Balkans, as well as in parts of southern Europe (e.g. rural central and northern Italy). Households were complex, and age at marriage for women often was comparatively low. As a result, gender asymmetry was sharp, and the fertility rate was high. Hajnal’s and Reher’s typologies partly overlap, but also partly differ, designing different boundaries in traditional patterns of family formation across Europe, depending on the specific dimension of family arrangements on which they focus. They do, however, share the idea that family arrangements belong to the longue durée and involve some kind of cultural path dependence. Past differences, therefore, shape also patterns of change. To stress the relevance of long-standing arrangements in family formation and cultures does not mean that one ignores history and the relevance of changing social, economic and political conditions. Rather, the approach suggests that the impact of social, economic and political conditions on family arrangements is to some degree dependent on these arrangements themselves. This (inter)dependence helps explain the substantial differences that even today can be found across Europe in family patterns – including gender and intergenerational arrangements. It may also partly explain differences in welfare arrangements (see also Bahle, this volume Ch. 4). One might argue that long-standing family cultures offer different resources to deal with new problems or opportunities.

Patterns of family living

49

These differences are not obliterated by the construction of the EU as a political and social entity. Rather, they enter, more or less explicitly, into the formal and informal negotiations and confrontations through which Europe as a supra-national body is constructed, shaping country-specific understandings of common goals and priorities, as in the case of what civil and social rights are about, or of gender equality. They are relevant not only for demography but for understanding patterns of everyday organisation, such as dealing with needs of income and care, for gender and intergenerational relations, and for the division of labour and responsibilities. This chapter is largely based on a new comparative source: the European Quality of Life Survey (EQLS), performed in 2003 on the 15 old EU member states, the 10 new member states and three countries that have applied for EU membership: Bulgaria, Romania and Turkey. This dataset has its limitations and does not allow for a full exploration of the range of possible variations in patterns of family formation and living.1 Yet the EQLS data do offer a rich, cross-sectional picture of what kinds of households Europeans live in and of how they organise their everyday life in dealing with their needs for income and care. The following analysis starts with the political and institutional divide that stems from the different timing of access to the EU – the EU-15, the new member states, and the three countries which have applied for membership – and then moves towards a deeper exploration of differences in patterns of family formation. In the last section, the analysis addresses differences in gender arrangements and, more specifically, in the gendered division of labour – a crucial contemporary difference in family arrangements.

1. Patterns of family formation in the European Union: an overview Within the ‘old’ 15 European Union member states (hereafter, ‘EU-15’), the main divide is between the northern European pattern, including the United Kingdom, and the southern European one, with the central European countries somewhere in between. The first group of countries for a long time had a marriage pattern that can be characterised as late, with a comparatively high percentage of the population who never marries. Until the 1970s, it also exhibited the lowest fertility rates in Europe; it now has the highest, together with Ireland and France. It also is the first group of countries where cohabitation as a pattern of first couple-formation emerged in the late 1970s and then spread, to become the most common pattern. Women’s labour force participation rates already were comparatively high in the 1960s. In comparison, in southern European countries the young, in particular young men, always have exited the parental household latest, and mostly only to marry. The gender differentials in age at marriage and the marriage rates have traditionally been higher here than in the northern countries, although in recent years the latter has been declining. Moreover, after having been the countries with the highest fertility rates, they now compete with eastern European countries (and Japan) for lowest fertility. Still, in these countries, marriage instability rates are comparatively low, as are labour force participation rates among women. EU enlargement is progressively including within the EU boundaries such countries as Estonia, which clearly lie west of Hajnal’s divide; countries such as Hungary, Poland, Latvia and Lithuania, which are on the border of this divide; and still others, such as Slovenia, Cyprus and, even more clearly, Bulgaria, Romania and Turkey,

50

Chiara Saraceno

which lie east of that divide.2 Whatever their positions with regard to this divide, most of these countries share some of the features of the so-called eastern European model: nearly universal marriage, a low age at marriage for women, and high fertility (Council of Europe 2001: 55–76). In 1985, the marriage rates in Communist Eastern (excluding Russia) and Central Europe were, respectively, 0.88 and 0.90, compared to 0.83 in Greece, 0.79 in Portugal and 0.70 in Austria and West Germany. Fertility, including non-marital fertility, also was higher in these countries than in Western Europe. The pattern of universal marriage and a comparatively low age at marriage for women persisted, particularly in Bulgaria, the Czech Republic and Hungary, until the generation born after 1955 (see also Tomka 2002). Due to widespread impoverishment since the end of communism, marriage rates have declined sharply, as have fertility rates. According to some analyses (e.g. Therborn 2004), however, this phenomenon reflects a conjuncture of events: marriage rates will pick up again, remaining on average higher than in Western Europe. In addition to fertility, marriage rates and age at marriage, another aspect showing variation in the past, but particularly at present, is a mother’s partnership situation at first birth. According to the data of the United Nations Fertility surveys in the 1990s (United Nations Economic Commission for Europe 2002), substantial differences exist across Europe, reflecting in part historical differences. In the Netherlands, Belgium, Estonia, the Czech Republic, Hungary, Poland, Spain, Italy and Greece, between 89 per cent and 95 per cent of first-time mothers were married. In another group of countries (Finland, France, western Germany, Latvia, Lithuania, Slovenia), the percentage was between 70 per cent and 85 per cent. In Norway it was 62 per cent, in Austria 56 per cent. In Denmark and Sweden it was well below half of all mothers: 26 per cent and 32 per cent, respectively. These two countries had the highest percentage of first-time mothers in an unmarried partnership (50 per cent and 58 per cent, respectively), followed at a distance by Norway (23 per cent), France (20 per cent), Austria (18 per cent), Finland (16 per cent) and Slovenia (15 per cent). Denmark also had the highest percentage of single mothers (24 per cent), followed by western Germany (19 per cent), Norway (15 per cent), Slovenia (15 per cent) and Poland (15 per cent). Poland, Greece and Lithuania had the lowest percentages of first-time mothers in a cohabitant unmarried partnership (1–2 per cent), followed by Spain, Italy, the Czech Republic and Belgium (3–5 per cent). With the exception of Denmark and Sweden, with Norway following at a distance, no clear Nordic pattern can be found in these data. Within the continental countries, a similarity emerges in that a substantial minority of first-time mothers are in a cohabitant unmarried partnership. But the incidence of both single and married first-time motherhood varies. The Mediterranean pattern of the prevalence of first-time motherhood within marriages is clearer. The pattern in most central-eastern European countries resembles the Mediterranean one with regard to the dominance of married first-time motherhood and the very low incidence of first-time mothers in cohabitant unmarried partnerships. But the incidence of single motherhood is higher. Furthermore, by 2002, Estonia had the second-highest extramarital fertility rate in Europe. Latvia and Bulgaria also were among the ten European countries with the highest fertility rate for unmarried mothers (e.g. higher than that in the United Kingdom; Council of Europe 2004). The risk of a marriage breaking up also varies substantially across Europe, ranging from one in ten (in the Mediterranean countries) to one in two (in the

Patterns of family living

51

Scandinavian countries). In 2002, Sweden, Belgium, Luxembourg, Finland, Estonia, the Czech Republic and Austria were among the ten countries with the highest rate of divorce, whereas Turkey, Spain, Italy, Croatia and Spain had the lowest divorce rate in Europe (Council of Europe 2004). Given these differences, the incidence of specific household patterns at a given point in time varies across countries. Not all countries use the same definition of the household in their statistics, despite the guidelines both of the United Nations and of Eurostat. Therefore comparisons are not always precise. The main differences concern whether a household is defined primarily as a dwelling unit or as a housekeeping unit. Other differences pertain to the definition of family units within households, and whether there is an age limit to define the relationship between parents and children. This variation in definitions renders not fully comparable the official statistical data, including the census data (see, for example, De Vos and Sandefur 2002; Hantrais 2004). Differences in definition affect in particular the comparability of the data on unmarried cohabitant couples, lone parents, and extended and multiple households. In the light of these methodological and conceptual difficulties, the EQLS data have the virtue of having been generated within the same conceptual framework. Households are primarily defined as dwelling units. Respondents were asked with whom they lived. Household types were reconstructed on the basis of these individual answers, focusing on whether or not a couple, children and/or other relatives were present in the household. There is no age limit in defining the parent–child relationship. Within the EU-25 (the 25 EU member states as of 2004), only 35 per cent of European households include a couple with children of any age, compared to 54 per cent in the three other countries; 30 per cent include a couple only.3 Lone parents comprise 8 per cent of all households. One-person households make up a quarter of all households, whereas extended households, in which (usually older) relatives live with a couple and their children, constitute a very small group (see Table 2.1).

Table 2.1 Household patterns in an enlarged Europe (%) Country cluster EU-15 NMS EU-25 CC-3

Living alone 26 15 25 9

Couple without children

Couple with children1

Lone parent2

Extended or multiple household3

32 19 30 19

33 46 35 54

7 11 8 8

2.5 10 3 10

Source: EQLS 2003; layer % Question HH3 (household grid): Now think about other members of your household, starting with the oldest: What is this person’s relationship to you? Notes: 1 Includes unmarried children of all ages. 2 Includes all adults living with their children, irrespective of the children’s age, but without a partner or any other person. 3 Extended households include one or more relative, but there is only one conjugal couple. Multiple households include more than one couple.

52

Chiara Saraceno

However, the incidence of the latter household pattern reaches 10 per cent in the new member states, compared to just 2 per cent in the old EU-15, confirming indirectly the persistence of Hajnal’s long-standing geographical divide.4 Conversely, singleperson households and childless couples are fewer in number in the ten new member states (hereafter, ‘NMS’) than in the EU-15. They are even less common in the three other countries (hereafter, ‘CC-3’), which show a distribution of household patterns more similar to that of NMS than that of EU-15. Households in the NMS and CC-3 are also more likely to include unmarried children than are those in the EU-15. These distributions suggest that different mechanisms of family formation are in place. The data do not, in fact, mean that most couples in Europe never have a child, nor that one fourth of the EU-25 population live alone throughout their entire adult life. Rather, the combination of increased life expectancy, low fertility rates, the widespread but diversified popularity of the nuclear household (in which each couple sets up its own household) and increasing marital instability opens up new household and individual life phases and/or renders old ones (e.g. the empty-nest phase) more prolonged and widespread. For instance, it is no longer widowhood alone that transforms a surviving spouse into a single person or a lone parent. The transformation also may occur through divorce or separation. Individual ‘household careers’ and the life courses of households have become more diversified, if not fragmented. For this reason, the kinds of household that people live in during different phases of the life course is a more meaningful indicator of patterns of household formation than is the simple distribution of household types within and across countries. Although the EQLS data are not longitudinal, they do offer the possibility of comparing across countries, as well as across gender and economic circumstances, the kinds of household situations that individuals experience and the household statuses that they have at various ages. The household unit – whatever its definition – of course does not fully coincide with ‘living arrangements’. Some household units may simply represent an address and possibly a sleeping place, whereas the individuals inhabiting them may spend all or most of their day in other households. This is the case of elderly individuals who spend their day and eat their meals in one of their children’s households. In some cases, young adults may have their own dwelling, but are supported financially by their parents, often eating their meals with them and having their laundry done by their mothers. More generally, ‘living arrangements’ may involve more or less frequent exchanges and relationships within the kinship network. As recent research has shown, kinship – particularly intergenerational ties – is a crucial practical as well as emotional resource throughout Europe (e.g. Kohli et al. 2005; Ogg and Renaut 2005). The EQLS data also testify that contacts with kin are frequent. Moreover, kin is the main resource for expected or received practical support in all countries, irrespective of household patterns and of welfare regime type or political history (Saraceno et al. 2005). ‘Households’, therefore, do not fully coincide with ‘families’ in the practices and understanding of most people. Yet households define some kind of boundaries within kinship networks. Specific household patterns, therefore, define those boundaries, and the need and the possibilities to cross them, in different ways. One difference concerns the experience of living alone or with others. The highest percentage of adult individuals living with someone else is found in CC-3 (91 per cent), the lowest in EU-15 (75 per cent), with NMS in between (85 per cent). Within the

Patterns of family living

53

EU-15, however, the Mediterranean countries are more similar to the NMS than to the central and Nordic European countries. Not surprisingly, whereas the households of those in the middle age-groups are more similar across countries, the biggest differences across countries and country clusters are found among the young and the elderly. These differences throw a light on distinct patterns of family formation and household experience along the entire lifecycle. When children leave the parental household late or, in some cases, bring their spouse or partner into it, adults in their mature and older years may spend many years living in the same house as their adult children. Some will become old and die while still living with them. Many lone mothers in Mediterranean countries, for instance, are neither unmarried single mothers nor separated or divorced ones. They are widows still living with their unmarried adult children. On the contrary, when children exit the parental household early, the adults in their mature years are likely to live many years as a childless couple or, particularly when old, as a single person. The kinship network is affected by these patterns, for in some cases the parents of both partners live in different households. In other cases, one set of parents (or one parent only), and sometimes other relatives, live with the couple and even with grandchildren. Although the kin network is larger than any kind of extended or multiple household and nuclear households and even single individuals may be embedded in dense kinship networks, demands and forms of support, patterns of giving and receiving across kinship networks are affected by patterns of family formation and household structure in the different phases of life.

2. Patterns of family formation when young Throughout Europe, the formation of a married or unmarried cohabitant couple and the beginning of parenthood – events traditionally marking completion of the transition to adulthood – normally follow the individual’s exit from the parental home. Yet for a small number of the young, the couple is formed within the parental household of one of the partners. This phenomenon is more evident, as one might expect, in some of the eastern European countries, such as Hungary, Bulgaria, Romania and Cyprus, but also in Ireland. Thus, for a proportion of the young in these countries, family support in ‘forming a new family’ occurs in the form of including a child’s partner in the household. On average, in the EU-15 the young become parents later than in the NMS and the candidate countries. They do, however, leave the parental household at a younger age. Young persons in the CC-3 leave the parental household later than in the EU-15, but earlier than in the NMS, and they also become parents earlier. The largest crosscountry differences among the young concern the percentages of those who still live as children in the parental household. Within the EU-15, differences are substantial for both genders, but particularly for men. The incidence of men under 35 years of age still living in the parental household, without a partner or children, ranges from 12 per cent in Sweden to almost 6 times as much, 67 per cent, in Italy. The other Mediterranean countries score 20 percentage points less than Italy. Greece is more similar to France and Luxembourg than to Italy. Only Malta, in the new member states, has a percentage similar to the Italian one. The other new member states range from 36 per cent for Estonia to 57 per cent for Slovakia. The CC-3 are similar to the higher end of the NMS spectrum (excluding Malta) (see Table 2.2).

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

Men 29 28 17 13 30 21 39 25 67 35 36 48 45 12 19

Women 13 28 8 11 18 14 19 19 60 25 22 33 34 10 12

Living with parents

Household Status

Men Women 35 23 28 11 33 28 39 29 36 24 40 26 33 24 10 7 11 6 12 5 27 20 7 5 5 5 44 31 33 14

Living alone

Men Women 14 26 20 14 28 28 26 25 16 23 11 27 8 14 14 15 8 12 13 13 19 22 8 10 22 16 17 24 22 36

Living as childless couple

Table 2.2 Household statuses of the young in Europe (ages 18–34) (%)

Men 13 20 20 18 15 20 9 15 11 25 12 28 13 22 8

Women 18 33 21 26 23 23 33 18 20 39 25 30 26 23 19

Living as a couple with children

Men 4 1 0 1 1 1 2 12 0 4 3 4 5 1 4

Women 3 1 2 0 1 1 3 13 1 5 1 11 4 2 1

Living as a couple with or without children in an extended household Men 3 2 0 0 1 2 1 6 0 4 1 0 1 0 1

Women 14 11 9 6 7 5 1 15 2 7 10 10 4 5 14

Lone parent in nuclear or in extended household

Men 2 1 1 3 2 6 8 18 2 7 2 5 9 5 12

Women 4 2 5 3 3 4 6 13 0 6 1 2 10 5 4

Other kind of household

43 48 36 47 39 41 67 55 57 53 50 51 49 36 52 49

29 28 29 32 22 18 55 45 41 43 23 19 29 27 39 26

16 9 17 13 14 9 7 2 4 10 11 6 6 26 6 6

6 8 19 5 13 8 4 3 1 6 3 11 1 16 4 4

9 17 26 10 13 8 9 4 4 10 3 12 7 15 8 8

20 7 23 14 20 14 14 4 4 11 13 18 9 21 8 11

19 20 10 16 21 34 9 22 21 17 23 11 20 14 21 18

30 34 17 25 24 35 11 23 31 26 25 20 40 23 26 35

9 2 5 11 9 8 5 11 12 5 10 16 7 3 10 10

8 9 2 16 7 9 8 15 13 9 27 16 12 2 14 15

2 0 2 1 2 1 1 2 1 1 0 1 2 1 1 2

3 11 10 4 11 13 1 8 7 3 6 4 4 7 8 4

3 4 3 2 3 0 2 4 2 5 4 4 8 5 3 7

5 3 1 5 4 4 7 2 2 3 3 12 4 4 3 6

Notes: The category ‘Living with parents’ includes all respondents without a partner and childless, who are living with parents with or without other aggregate members. The category ‘Other kind of household’ includes the respondents who live with friends, brothers or sisters (without parents) or other aggregate members

HH3c: ‘Now think about other members of your household, starting with the oldest: What is this person’s relationship to you?’

Source: EQLS 2003; layer %

Cyprus Czech Republic Estonia Hungary Latvia Lithuania Malta Poland Slovakia Slovenia Bulgaria Romania Turkey EU-15 NMS CC-3

56

Chiara Saraceno

Turkey has the highest, Malta the lowest percentage of young women who are mothers: respectively, 48 per cent (including those living with a partner and children in an extended household) and 12 per cent. Ireland, the United Kingdom, Austria and Lithuania have the highest percentages of lone mothers (15–13 per cent; living alone or in an extended household) in this age group; Greece and Malta have the lowest, followed by Italy. Sweden has a percentage close to that of Spain and much lower than that of Denmark. The former socialist countries have comparatively high percentages. Thus, the EQLS data indirectly confirm findings based on demographic sources: the rise in natural births is accompanied by a rise in lone motherhood in some countries, such as the former socialist ones and some western European ones as diverse as Ireland, the United Kingdom, Portugal, the Netherlands and Denmark (though not others, such as Sweden, where natural births tend to occur within a cohabitant couple relationship).5 If one examines two of the dimensions which in the literature are considered to be crucial dimensions both for entering adulthood and for forming a family – exiting the parental household and forming a partnership – three patterns may be distinguished (see Figure 2.1 for these patterns among young men).6 They confirm the patterns described by Alessandro Cavalli and Olivier Galland (1996) for the EU-15 only (see also Billari and Wilson 2001, Schizzerotto and Lucchini 2004). The patterns are clearer among men than among women. In the first pattern, prevalent in the Nordic countries, Germany, Austria, Belgium, France and the United Kingdom, as well as in Greece,7 a substantial number of young people is outside the parental household by the age of 34. They either live alone as singles, particularly if they are males, or live with a partner, with or (more often) without children. Among young Swedish and German men 44 per cent and 40 per cent live by themselves, respectively, compared to 12 per cent and 21 per cent who still live at home with their parents. In the second pattern, prevalent in the southern European countries and in the new member states, around half, and sometimes more, of those under 35 years old are still in the parental household without a cohabiting partner, particularly if they are male. For example, 67 per cent of Italian and Maltese young men, 57 per cent of Slovakian, 55 per cent of Polish and over 40 per cent of Portuguese and Spanish young men are still living in the parental home at the age of 34, without a partner. These households usually are nuclear, but, particularly in the NMS and in CC-3, for a small number they are extended; that is, other relatives (usually one or more grandparents or a sibling’s spouse) also are present.8 If they are outside the parental household, the young are more likely than in the former pattern to be already in a partnership and, particularly if female, to have one or more children. The third pattern concerns a smaller group of the young in all countries; it is particularly concentrated in the NMS and the CC-3, but also plays a significant role in Ireland. In this case, when the young form a partnership, they remain in the parental household or enter the partner’s parental household, which sometimes includes other relatives as well. In all three patterns, women generally enter a partnership and become parents earlier than do men. In comparison with men, they are less likely to live as childless singles, but more likely to live as a lone parent, either alone with their child or children, or together with their parents in an extended household. As Cavalli and Galland (1996) have indicated, the three patterns of family formation found among young Europeans involve different kinds of exchange and forms of

Patterns of family living

57

60 56

50

48

40 35 % 30 26

19

20 16

16 14 11

10

7

7 5 2

5

9

10 7

4 2

1

0 Pattern of ʻearly exitʼ

Pattern of ʻlate exitʼ

Living with parents Living as childless couple Living as a couple with or without children in extended household Other kind of household

Pattern of ʻpartnering in a parental householdʼ

Living alone Living as a couple with children Lone parent in nuclear or extended household

Figure 2.1 Household status of young men (18–34): three distinct patterns (%) Source: EQLS 2003 Question HH3c: ‘Now think about other members of your household, starting with the oldest: What is this person’s relationship to you?’ Notes: The category ‘Living with parents’ includes all respondents without a partner and childless, who are living with parents with or without other aggregate members. The category ‘Other kind of household’ includes the respondents who live with friends, brothers or sisters (without parents) or other aggregate members. The ‘early exit’ pattern includes Austria, Belgium, Denmark, Finland, France, Germany, Greece, the Netherlands, Sweden and the United Kingdom. The ‘late exit’ pattern includes Estonia, Hungary, Italy, Malta, Poland, Portugal, Slovakia, Slovenia and Spain. The ‘partnering in a parental household’ pattern includes Bulgaria, Cyprus, the Czech Republic, Ireland, Latvia, Lithuania, Luxembourg, Romania and Turkey.

58

Chiara Saraceno

support across families and kin, as well as different options available for the young. An individual’s ability to leave the parental household early may be supported by cultural values, but also by favourable labour and housing markets as well as by welfare state provisions. If, however, the family is the main financial resource for the young and the housing market is tight, it is more difficult for a young person to leave the parental household when he or she is not yet established in the labour market. Moreover, it may be more costly for parents to help children live on their own. And, finally, if the family culture does not support the extended household pattern, or if parents are unwilling to accept a child’s spouse into their household, the young must wait not only to leave the parental household, but also to marry. The concentration of young couples living in an extended household in the countries east of Hajnal’s line, that is, in countries where the extended or multiple family pattern has a long tradition, confirms the substantial influence of historical, geo-culturally bounded family patterns. Data on the occupational status of the young in different household circumstances and in different countries support the hypothesis that decisions about forming a new household are certainly influenced, but not dictated, by labour market and welfare Cluster EU-15 NMS CC-3

TR

% of the young neither in work nor in education

40 BG UK

30 PL

20

RO

EL

IR LT

DE BE

FI

FR

PT

LV

ES

NL LU

IT EE HU

10

SW DK

10

AT

SI

CZ CY

20

SK

MT

30 40 50 60 % of the young living with parents

70

Figure 2.2 Incidence of unemployment among European young singles living with their parents, by country Source: EQLS 2003, own calculations

Patterns of family living

59

state conditions. Actually, with the partial exception of Poland, the incidence of those who are neither in work nor in education is higher among the relatively few young persons who live without a partner and childless in the parental household in some of the ‘early exit countries’ (e.g. the United Kingdom) than among the more numerous ones in the ‘late exit’ countries (see Figure 2.2). Those who live with a partner, either with or without children, in an extended household are also more likely to be in paid work: 90 per cent in the NMS, slightly less in the EU-15 and CC-3. Thus, remaining in the household of one’s parents is not necessarily a means to buffer unemployment. It might be a means to integrate otherwise low wages, to accumulate capital for establishing a new family, or to support exploratory strategies in the labour market.9 Gender differences are evident, however: young women who are not in education and are still living in the parental household are generally more likely than men not to be working. In sum, these cross-country differences suggest that paths into adulthood are shaped not only by available resources, such as publicly provided income support in the event of unemployment, scholarships for education, or access to housing. Cultural expectations about proper behaviour and the proper sequence of events are also involved.

3. Households of elderly persons By around 2020, the main increase among elderly persons in the EU will occur within the over-80 group. This group will increase by about 50 per cent (European Commission 2003). The situation looks more balanced in the new member states, as their population is on average younger, but the trend is quite similar (Fahey and Spéder 2004). An increasing number of households will comprise only elderly people. Kinship networks will have an age and intergenerational balance skewed towards elderly people – with frail elderly persons constituting a crucial, if relatively small, proportion of them, who will have special needs and demands for care. In which kind of household elderly persons live depends not only on their own past choices concerning marriage and/or cohabitation with a partner, divorce, having children and how many children they have had, at what age, and so forth. It depends also on patterns of household formation by the young: whether they bring their partner into the parental household or they form a new household, at what age they exit their parental household, and so forth. It depends, moreover, on gender, given women’s higher life expectancy on average, which, coupled with their younger age at marriage on average, makes it more likely for them than for men to outlive their partners into old age (see also Iacovou 2000, based on European Community Household Panel data). Gender differences are systematic across countries. Throughout Europe, older men live with their partner to a much greater degree than older women. Conversely, older women live alone to a much greater degree than older men (see Figure 2.3). Thus, in old age there is a partial reversal of the pattern found among the young, among whom more men than women live alone. For instance, in Sweden and Denmark, over 70 per cent of older women live by themselves, compared to over 40 per cent of older men. In Portugal and Greece, over 55 per cent of older women live alone, compared, respectively, to 28 per cent and 10 per cent of older men. In Poland, the percentages

60

Chiara Saraceno 40 30 20 10

%

0 EU-15

NMS

CC-3

−10 −20 −30 −40 Living alone

Living as a childless couple

Living as a couple with children

Living as a couple in a next ended household

Living without partner, with children or relatives

Other kinds of households

Figure 2.3 Households of elderly people: differences between men and women (65 and over) in the three clusters (%) Source: EQLS 2003 Notes: The figure shows the relative differences between the household status of elderly men and that of elderly women (calculated as the difference between the % of elderly men who live in a given household status and the correspondent women’s % in the same household status). Negative differences show how much less men are in that status compared to women. Positive differences indicate how much less women are in that status compared to men. The category ‘Other kind of household’ includes those who live with relatives or other people, but no partner or children.

for women and men are, respectively, 39 per cent and 17 per cent; in Turkey, 23 per cent and 3 per cent. Fewer women than men still live with both their partner and their children. On the other hand, more women than men live with their children, whether or not these children have a partner, when they lose their own partner. This latter gender difference is particularly clear in NMS and CC-3, given the higher incidence of extended households in these countries. Cross-country differences are as important as gender differences. They do not, however, neatly overlap with any standard clustering based on tripartite EU membership, on Hajnal’s divide, or on patterns of family formation by the young (see Table 2.3). Certainly, the experience of living with one’s own children when old is nearly absent in such ‘early exit’ countries as Denmark, Sweden, the Netherlands, France and Germany. But in Austria and particularly in Greece, which share the same pattern, this experience is relatively widespread. As one might expect, the incidence of this

Patterns of family living

61

1

Table 2.3 Elderly men and women (65 and over) living with their children; by patterns of family formation by the young2 (country %) Patterns of family formation by the young ‘Early exit’ Austria Belgium Denmark Finland France Germany Greece Netherlands Sweden United Kingdom

10.8 7.5 2.23 6.7 3.63 1.63 19.8 3.83 2.13 7.8

‘Late exit’ Estonia Hungary Italy Malta Poland Portugal Slovakia Slovenia Spain

14.3 20.9 27.8 31.0 35.0 11.3 18.3 21.6 23.2

‘Partnering in a parental household’ Bulgaria Cyprus Czech Republic Ireland Latvia Lithuania Luxembourg Romania Turkey

22.0 21.4 12.9 15.0 16.7 14.2 9.33 19.1 44.8

Source: EQLS 2003 Notes: 1 All elderly persons living as a couple with children (in a nuclear or extended household) and lone parents living in a nuclear or extended household. 2 See Figure 2.1. 3 N < 15.

household situation is highest in the ‘late exit’ and the ‘partnering in a parental household’ countries, but with notable internal differences. Within the former, Poland, Malta and Italy have substantially higher percentages than do Portugal and Estonia. Slovenia, Slovakia, Spain and Hungary lie in between. Differences are even greater within the third group of countries, ranging from almost half of all elderly persons in Turkey living with their children – whether or not the children have a partner – to 13 per cent in the Czech Republic. This ‘fuzziness’ of clusters is probably due to the fact that household patterns in old age are the outcome both of patterns of household formation by the current younger cohort and of their own cohort’s specific socio-cultural and demographic

62

Chiara Saraceno

histories: marriage rates, age differentials at marriage, fertility, mortality and migration. The interplay of all these phenomena may offer different household ‘options’ when old, even within otherwise similar geo-cultural patterns of family formation. Furthermore, life expectancy varies significantly across Europe. Sweden, Italy, Cyprus, Greece and Norway are among the ten countries with the highest life expectancy for men in Europe (at 76–79 years). Romania, Lithuania, Turkey, Latvia, Estonia and Ukraine are among the ten with the lowest life expectancy, ranging from 63 to 67 years. For women, Sweden and Italy are still in the first group, which also includes Spain, France, Finland and Austria (at 83–84 years). Bulgaria, Romania, Ukraine and Turkey are still together in the lowest expectancy group, at 71 to 76 years (Council of Europe 2004). Given these differences in household patterns, a different proportion of elderly people in each country, as well as of women in all countries, may find themselves living alone when they become frail. This situation puts a different kind of pressure both on family and kinship solidarity and on welfare arrangements. In the EQLS, this issue has not been dealt with in detail. There was one question asking how often the respondent took care of an elderly or disabled relative (with no distinction made between co-resident relatives and relatives living elsewhere). Provision of care for elderly persons is higher in the NMS and CC-3 than in the EU-15, and also involves to a greater degree adults of all ages, not just those over 54 years old (Saraceno and Olagnero 2004). Of the entire sample, 5 per cent assumes such an obligation daily (with more women than men taking on this responsibility), another 5 per cent at least once a week, and the remainder less often. But over half of the sample never takes care of an elderly or disabled relative. This last figure is higher than that found in Eurobarometer data (Alber and Kohler 2004).10 In her study based on European Community Household Panel (ECHP) data, Maria Iacovou (2000) found that a large proportion of elderly persons living with their children receives care within the household, particularly in the Mediterranean countries. She also found that this care is to a large degree reciprocal, particularly in the case of women: the ‘younger old’ among them, in fact, provide childcare almost to the same extent as care is provided by the household to the ‘older old’. This reciprocity does not hold for men, however. Recent research on a smaller number of countries with a specific focus on the issue of intergenerational exchanges (Kohli et al. 2005) indicates that between 40 per cent and 65 per cent of those over 80 receive some kind of help (care, but also help with shopping or with bureaucratic work) from their co-resident or non-co-resident children. Contrary to many assumptions implicit in welfare regime typologies, elderly people living in Mediterranean countries do not seem to receive more help from their family than do those living in Nordic and central European countries. Welfare state regimes possibly make a difference not so much in availability of family support as in the degree to which these exchanges and forms of support are the only, or main, available option, or even a necessity.

4. Household patterns in middle adulthood (35–64) By age 35, about 90 per cent of all Europeans live outside their parental household. The large majority lives with a partner, with or without children. There are, however, noticeable differences with respect to gender and age (see Figure 2.4). Up to 49 years of age, more men than women live alone, either because they have not yet formed a

Patterns of family living

63

30

20

10

0

−10

−20

−30

EU-15: difference between 50–64and 35–49-year-old men

EU-15: difference between 50–64and 35–49-year-old women

NMS: difference between 50–64and 35–49-year-old men

NMS: difference between 50–64- and 35–49-year-old women

CC-3: difference between 50–64and 35–49-year-old men

CC-3: difference between 50–64and 35–49-year-old women

−40 Living with parents

Living alone

Living as childless couple

Living as a couple with children

Living as a couple with or without children in extended household

Lone parent in nuclear or extended household

Other kinds of households

Figure 2.4 Households of 35– to 64-year-olds: differences between the 35–49 and the 50–64 age groups, by gender and country clusters (%) Source: EQLS 2003, own calculations Notes: The figure represents the relative differences between the household statuses of, respectively, (a) men in the 50–64 and 35–49 age groups and (b) women in the 50–64 and 35–49 age groups. The difference has been calculated as that between the % of 50–64-yearold men who have a given household status and the correspondent % in the 35–49 age group; the same calculation has been carried out for the women’s sample. Negative differences indicate how much less 50–64-year-old men (or women) are in that status compared to 35–49-year-old men (or women). Conversely, positive differences indicate how much less 35–49-year-old men (or women) are in that status compared to 50–64-year-old men (or women). The category ‘Other kind of household’ includes the respondents who live with friends, brothers or sisters (without parents) or other aggregate members.

partnership or because they have left it. More women than men are lone parents.11 Actually, these two household statuses are somewhat symmetrical, suggesting that to some degree they are gender-specific outcomes of the process of partnership dissolution, particularly when children are present. Women are more likely to remain with the children. Men are more likely to leave the household to form a new one as a single person household, at least temporarily. Compared to younger women, 35- to 49-year-old lone mothers are less likely to ‘go back’ to, or remain in, their parents’ household.

64

Chiara Saraceno

There are fewer lone mothers living in extended households in this age group. These findings are similar to those from a study on the EU-15, which was based on ECHP data and used a more restricted definition of lone parenthood (i.e. the lone parent lives only with her or his dependent children; see Lehmann and Wirtz 2004). According to this study, in 2001 lone-parent households made up 9 per cent of all households with dependent children in the EU-15, ranging from 22 per cent in Sweden and 17 per cent in the United Kingdom to 4 per cent in Italy, Portugal and Greece and 3 per cent in Spain. Over 90 per cent of lone parents were women. Only in Sweden were 26 per cent of lone parents men. Of the lone parents in the EU-15, 86 per cent were between the ages of 25 and 49. Only in Greece and Portugal was there a relatively high proportion of lone parents in the age group 50–64 years: 23 per cent and 22 per cent, respectively, compared to an average of 11 per cent. These differences in the incidence of lone parenthood imply differences in the risk that European children have of spending part of their childhood and/or adolescence with only one cohabitant parent. Furthermore, some national research reported by Linda Hantrais (2004: 59) suggests that children are likely to spend a longer period of time in a lone-parent household if their parents divorce than if one parent dies or a cohabitant unmarried partnership breaks up. In the EQLS sample, 14.2 per cent of underage children lived in a lone parent household in EU-15, 11.9 per cent in NMS and 8.4 per cent in CC-3. The highest percentages were found in the UK (26.5 per cent) and Estonia, the lowest in Malta. But the numbers are too small to allow any analysis. By the age of 50, the effect of gender differences in life expectancy and in age at partnering starts reversing the differences in household statuses found among the 35- to 49-year-olds. The pattern that is so clear among elderly people starts to emerge: now women have a slightly higher likelihood than men to live alone. They continue to have a higher likelihood to be lone parents, but their children are now likely to be young adults. Thus, although they spend much of their adult life together, women and men shift their differences over the lifecourse. Or rather, their differences change sign: men are more likely to live alone when young; women are more likely to live alone when old. This is also the age at which an increasing percentage of people, particularly women, is involved in the care of a frail elderly person, cohabitant or not. In no country cluster is there a substantial percentage of 35- to 64-year-olds who live in an extended, three-generation household. This circumstance may be a lifecourse effect. As individuals age, it is less likely that their parents are still alive, while their cohabiting children may not have yet formed a partnership. Yet, given the present high life expectancy, this explanation is at best partial. In fact, although among the 50- to 64-year-olds the percentage of those who still have a child in the household decreases dramatically, the percentage of those who are in an extended household increases only slightly. Rather, this finding seems to suggest that, even in the countries where it once was widespread, the extended household pattern has been weakened as a model among the cohorts who are at present in the middle age groups. It remains, however, as a buffer or a necessity, in two phases of household formation: when the young couple does not have enough resources to set up its own household, and when elderly persons become frail and remain alone. Moreover, although co-residence between more than two generations may not be the prevalent pattern of family organisation, exchanges and support across households within the family network may be a crucial organisational as well as emotional resource. Other comparative research has indicated that it is particularly this age group, the so-called pivot generation that

Patterns of family living

65

is most likely to give financial support and care both to the older and to the younger generations (Attias-Donfut et al. 2005; Kohli et al. 2005).

5. One-worker and two-worker couples: another intra-European differentiation in patterns of household organisation? Households are not simply a matter of who lives with whom when. Because households have to deal with needs of income and care, decisions concerning household organisation involve also decisions and arrangements concerning the allocation of paid and unpaid work, particularly within the couple (if one is present). These decisions, particularly as they affect women’s and mothers’ participation in paid work, increasingly differentiate households’ organisational patterns across Europe. As findings of the European Values Survey show (Halman 2001), the decisions certainly are influenced by cultural patterns regarding what is appropriate, particularly when there are small children in the family. But they also are contingent on welfare state and labour market arrangements. As European employment surveys indicate, there are meaningful differences across Europe not only in women’s activity rates, but also in the degree to which motherhood has a negative impact on women’s labour force participation. The negative impact of having a child under six ranges from nearly 40 percentage points in the Czech Republic and Hungary to 3 percentage points in Denmark and minus 8 percentage points in Slovenia (see Plantega and Siegel 2004). EQLS data indicate the considerable variation across Europe in the distribution of three possible paid-work arrangements within couples of working age – one worker only, two workers and no workers. Differences in this case seem to depend less on patterns of family formation than on welfare regime patterns (see Figure 2.5) and on gender models. The social democratic countries and the former socialist countries show a remarkably similar distribution. In these countries, in the large majority of working-age couples both partners work. The outlier is Poland, both because of the higher incidence of couples who do not work and because of the lower percentage of two-worker couples. In the continental countries, the majority of couples also consists of two workers, but the proportion of one-worker couples is higher than in the first two clusters. The Mediterranean countries appear to be quite differentiated. Cyprus and Portugal are similar to the continental countries, and in Spain, too, slightly over half of the couples have both partners in employment. In Italy and Greece, two-worker couples make up less than half of all couples of working age. Finally, in Malta, they represent a clear minority. Among the two countries usually identified with the liberal cluster, the United Kingdom is similar to Belgium and France, whereas Ireland is similar to Italy. With respect to the CC-3, Turkey has by far the lowest percentage of two-worker couples in all 28 European countries, as well as a relatively high percentage of couples with no workers. Bulgaria’s and Romania’s percentages of two-worker couples are similar to those of the continental countries. On the other hand, Romania has a higher percentage of couples with no workers than Poland does. Two-worker couples may, however, combine quite different time schedules. European labour force surveys indicate that across Europe, on average, women work shorter days than men do, and that part-time work is largely women’s work, either

66

Chiara Saraceno CZ 2

19

79

DK

6

17

FI

5

19

76

20

75

77

SK

4

SI

4

23

SE

3

25

EE

3

72

28

AT 3

70

29

HU

6

68 27

FR 2

67

33

LV PT

73

5

64 31

3

64

33

LT

9

64 28

64

UK

7

30

63

BE

7

31

63

CY 1

38

BG

6

LU 2

60

38

DE NL

61 33

60

9

34

5

58

37

ES 2

58 47

RO

51

18

32

IT

5

48

IE

6

48

PL EL

CC-3 0%

47 42

4

45

53

42

61

TR

NMS

47

13

MT 1

EU-15

50

38

17

70

5

13

37 9

58 33

58

16

59

20%

40% No workers

25

60% Single worker

80%

100%

Two workers

Figure 2.5 Distribution of working–age couples by working status in Europe Source: EQLS 2003, own calculations

as an option during a certain life stage (e.g. in the early years of motherhood) or as a permanent condition. However, given the different incidence and availability of parttime work across Europe, couples with ‘one and one-half workers’ (Crompton 2006, Lewis 2003, Pfau-Effinger 1998, 2004) predominate in the continental countries and the United Kingdom, particularly when they have young children (Bielenski et al. 2002, Franco and Winqvist 2002, Plantega and Siegel 2004). They also make up a substantial proportion in the Nordic countries. Two full-time workers is the prevalent pattern in most of the former socialist countries. It is the prevailing pattern in the Mediterranean countries as well – when the couple has two earners – even though there is little support from the welfare state. This lack of support explains the relatively low, though increasing, proportion of two-earner households in the Mediterranean countries. Household organisation is affected not only by the number of paid workers in the couple and by their paid working-time schedule, but also by the needs for care and by the gender division of unpaid family work (Saraceno 2005). With regard to the former, in most countries the presence of children, particularly when they are very

Patterns of family living

67

young, has a negative impact both on women’s activity rates and on their working time, but to a quite different degree across countries (e.g. Plantega and Siegel 2004). Parenthood impacts negatively on women’s participation far more in new member states than the EU-15: the Czech Republic, Hungary, and Slovakia are the three countries where women are most affected by parenthood. On the contrary the impact is non-existent in Sweden and minimal in Denmark. With regard to the gender division of labour, comparative time-use data on a number of European countries (Eurostat 2004; Sabbadini and Romano 2006) indicate that unpaid family work remains mainly women’s responsibility (see Table 2.4). When women are in paid work, however, they reduce their time devoted to family work, whereas their partners increase their share of it. In most countries studied, the overall (paid and unpaid) workingtime load is higher for working women than for working men, although men on average work more hours for pay than women do. This is not the case, however, in Sweden, Norway and the United Kingdom, although the balance between paid and unpaid work may differ between men and women. Estonian, Slovenian and Italian working women have the longest working day. Italian men have the longest paid working day and are the least collaborative in unpaid family work. The EQLS data are less precise. But they give information on a larger number of countries. Generally, they confirm the picture presented above. In particular, they suggest, on the one hand, that both the former socialist and the Nordic countries are the most gender equal with regard to the overall workload, particularly in two-worker couples. But whereas in the Nordic countries the overall workload for both men and women is among the shortest in Europe, in the former socialist countries it results in a very long working day for both. The Mediterranean countries appear to be the most unbalanced with respect to gender, both when households are based on the male breadwinner model and when they have two earners. Even when women work full time, the gender division of labour in family work remains skewed (Saraceno et al. 2005).

6. Conclusions Households in Europe are differentiated in terms of who lives with whom when, timing and sequence of events (e.g. cohabitation, marriage, fertility, exit from the parental household and so forth). During the twentieth century there has been a noticeable convergence with regard to family structures, in so far as neo-locality prevails at present in the formation of households throughout Europe to a much greater extent than in the past. In addition, the widespread downward fertility trend has homogenised countries and social groups which previously had quite distinct patterns. Yet the timing of exit from the parental household and the sequencing of cohabitation, marriage and parenthood distinguish different clusters of countries that to some degree overlap with the north/south and east/west divides. This, in turn, has an impact on the average household size (greater in southern and eastern Europe) and on the likelihood that a substantial proportion of elderly people (particularly elderly women) live alone. At the same time, significant cross-country differences in marital instability rates and birthrates both out of wedlock and out of cohabitation shape different household and life course patterns for children. There are also noticeable differences in the patterns of households’ time organisation and of dealing with needs for income and care. One of the main changes in

8:01 2:35 5:03 2:15 1:43 4:23 24:00

Men Sleep Meals and self-care Paid work and/or study Unpaid family work Travel Free time/unspecified time Total 8:22 2:11 5:05 2:20 1:20 4:47 24:00

8:23 2:36 4:13 4:04 1:15 3:51 24:00

Estonia

8:12 1:55 5:32 1:59 1:17 5:06 24:00

8:22 2:02 4:20 3:21 1:16 4:38 24:00

Finland

8:24 2:58 5:44 1:53 1:10 3:51 24:00

8:38 2:57 4:32 3:40 1:05 3:08 24:00

France

8:00 2:58 5:05 1:52 1:31 3:51 24:00

8:11 2:57 4:32 3:40 1:05 3:08 24:00

Germany

8:08 2:30 5:25 2:09 1:10 4:37 24:00

8:18 2:21 4:43 3:54 1:02 3:43 24:00

Hungary

7:58 2:52 6:13 1:10 1:40 4:06 24:00

8:00 2:44 4:39 3:51 1:28 3:18 24:00

Italy

7:53 1:58 4:56 2:12 1:23 5:37 24:00

8:07 2:02 3:46 3:26 1:17 5:22 24:00

Norway

8:06 2:07 5:20 2:24 1:14 4:52 24:00

8:12 2:02 4:23 4:24 1:09 3:51 24:00

7:52 2:05 5:17 2:23 1:32 4:51 24:00

8:05 2:23 4:05 3:32 1:28 4:27 24:00

Slovenia Sweden

Sources: Eurostat 2004; for Italy: Sabbadini and Romano 2006. The years when the data were collected differ. For example, for France, they refer to 1998; for Italy, to 2003

8:16 2:36 3:53 3:52 1:30 3:51 24:00

Women Sleep Meals and self-care Paid work and/or study Unpaid family work Travel Free time/unspecified time Total

Belgium

Table 2.4 Use of time among working Europeans (hours and minutes)

8:11 1:55 5:42 1:54 1:36 4:41 24:00

8:25 2:07 3:46 3:26 1:17 4:21 24:00

UK

Patterns of family living

69

family relationships over the past half century – women’s emancipation through education and labour force participation – appears to have a differential impact on household organisation across Europe depending both on previous arrangements and cultural models and on the ability of a society to adapt to this change. In the Nordic countries, the two-worker arrangement predominates, but when children are very young a good share of women chooses part time. Couples with one and one-half workers prevail in the continental countries and the United Kingdom, particularly when they have young children. Two full-time workers is the prevalent pattern in most of the former socialist countries as well as in the Mediterranean countries – when the couple has two earners. In the Mediterranean countries, however, this arrangement still pertains to just a minority of all working-age couples; moreover, the skewed gender division of labour remains unfavourable for women. Particularly in the Mediterranean and the former socialist countries, households headed by adults in the middle age-groups may find themselves under considerable time pressure due to long working days and responsibilities to provide care that are little supported by the welfare state. Extended family solidarity may buffer these strains. But the declining fertility rate is creating a demographic context in which there will be an increasing care deficit for the older generations to come. In the former socialist countries, this care deficit might even be intensified by emigration. Furthermore, somewhat paradoxically, eastern European women as migrant care-workers in the EU-15, particularly in Mediterranean countries, are becoming the private market solution in welfare regimes that deal inadequately with the care deficit emerging from the interplay between population ageing and the increase in women’s labour force participation (Da Roit and Sabatinelli 2005). Yet the women filling these positions may open up this same deficit in their own countries. The ‘global care chain’, as it has been defined (Anderson 2000), might produce unbalances not only in individual and family lives, but also in demographic and social structures. In any case, the brief story that has been told in these pages suggests that in family arrangements both differences and similarities do not always overlap with political and institutional divisions, even if these may crystallise them. It also suggests that in this field convergence is a somewhat illusionary (or moving) target. Families and societies interact with each other. The way the latter have adapted to challenges posed by changes in the former has been and continues to be as diverse across Europe as family forms themselves.

Notes 1 For instance, some crucial questions (e.g. whether a couple was married or cohabitant; whether a lone mother previously had been married) were not included in the questionnaire. 2 Exactly which country belongs to which part of the divide is controversial. For instance, Hajnal did not include Hungary in the eastern European model, but, according to Tomka (2002), Hungarian family arrangements bear a great similarity to aspects of this model, particularly with regard to age at marriage. According to De Vos and Sandefur (2002), the same holds for Slovakia, whereas Estonia is more similar to the Nordic pattern and the Czech Republic to the west-central European one. 3 Only 30 per cent of all households across Europe include at least an underage child. 4 In Bulgaria 32 per cent of all underage children live in an extended household, compared to slightly over 1 per cent in Denmark or the UK. In Hungary, Slovakia, Slovenia and Poland the figure is over 15 per cent.

70

Chiara Saraceno

5 A study on lone parents in the EU-15, which was based on ECHP data (Lehmann and Wirtz 2004), found that Ireland was the country with the highest incidence of lone mothers in the youngest age bracket (16–24): 11 per cent compared to the average of 2 per cent. 6 In order to identify these patterns, an agglomerative hierarchical cluster analysis was used. With this procedure, homogenous cases are grouped in order to form clusters. The ‘complete linkage’ method was implemented. In this method, the distance between two clusters is calculated between their two farthest points (Lance and Williams 1967a, 1967b). In the sample of young women, these patterns are not clearly identifiable. 7 The data on Greece are somewhat surprising. It is the only case in which the EQLS data differ substantially from other sources that usually put Greece in the same group as the other Mediterranean countries. 8 Given the small number of these situations, for the purpose of this analysis I have grouped together all childless children without a partner who are living in the parental household, without distinguishing between those living in nuclear and those living in extended households. 9 In Italy, for instance, children of better-off households leave the parental household later than children of low-income households. So, too, do only children in comparison with children with siblings. 10 Cross-survey differences in findings on this issue may depend on different factors: age composition of the sample, wording of the question, and so forth. The issue of comparability in this field has been addressed in a review of European studies by Jacobs et al. (2005). 11 For the purpose of this analysis of household status, ‘lone parent’ includes all adults living with at least one child and no partner, irrespective of the child’s age.

References Alber, J. and Kohler, U. (2004) Health and Care in an Enlarged Europe, European Foundation for the Improvement of Living and Working Conditions, Luxembourg, Office for Official Publications of the European Communities. Anderson, B. (2000) Doing the Dirty Work? The Global Politics of Domestic Labour, London, New York: Zed Books. Attias-Donfut, C., Ogg, J. and Wolff, F.-C. (2005) ‘European patterns of intergenerational and time transfers’, European Journal on Aging, 2, 3: 161-173. Barbagli, M. and Kertzer, D. (eds) (2003) ‘Introduction’, pp. xi–xliv, in M. Barbagli, and D. Kertzer (eds), Family Life in the Twentieth Century, New Haven: Yale University Press. Barbagli, M., Castiglioni, M. and Della Zuanna, G. (2003) Fare famiglia in Italia, Bologna: il Mulino. Bielenski, H., Bosch, G. and Wagner, A. (2002) Working Time Preferences in Sixteen European Countries, Luxembourg: Office for Official Publications of the European Communities. Billari, F. and Wilson, C. (2001) Convergence towards Diversity? Cohort Dynamics in the Transition to Adulthood in Contemporary Western Europe, MPIDR Working Paper WP 39, December 2001, Rostock: Max Planck Institute for Demographic Research. Cavalli, A. and Galland, O. (1996) Youth in Europe, London: Pinter. Council of Europe (2001) Recent Demographic Developments in Europe 2001, Strasbourg: Council of Europe Publications. Council of Europe (2004), Recent Demographic Developments in Europe 2004, Strasbourg: Council of Europe Publishing. Crompton, R. (2006) Employment and the Family, Cambridge: Cambridge University Press. Da Roit, B. and Sabatinelli, S. (2005) ‘Il modello mediterraneo di welfare tra famiglia e mercato’, Stato e mercato, 74, 2: 267-290. De Vos, S. and Sandefur, G. (2002) ‘Elderly living arrangements in Bulgaria, the Czech Republic, Estonia, Finland and Romania’, European Journal of Population, 18: 21-38.

Patterns of family living

71

European Commission (2003) The Social Situation in the European Union, Luxembourg: Office for Official Publications of the European Communities. Eurostat (2004) How Europeans Spend Their Time: Everyday Life of Men and Women. Data 1998–2002, Luxembourg: Office for Official Publications of the European Communities. Fahey, T. and Spéder, Z. (2004) Fertility and Family Issues in an Enlarged Europe, Report to the European Foundation for the Improvement of Living and Working Conditions, Luxembourg: Office for Official Publications of the European Communities. Franco, A. and Winqvist, K. (2002) ‘Women and men reconciling work and family life’, Statistics in Focus: Population and Social Conditions, Theme 3-9/2002. Hajnal, J. (1982) ‘Two kinds of preindustrial Household Formation System’, Population and Development Review, 8, 3: 449–494. Halman, L. (2001) The European Value Study: A Third Wave. Sourcebook, EVS/WORK, Tilburg: Tilburg University. Hantrais, L. (2004) Family Policy Matters, Bristol: Policy Press. Iacovou, M. (2000) Living Arrangements of Elderly Europeans, Working Paper, Institute for Social and Economic Research, Colchester: University of Essex. Jacobs, T., Lodewijckx, E., Craeynest, K., De Koker, B. and Vanbrabant, A. (2005) ‘Mesurer l’aide informelle: synthèse des pratiques européennes et nouvelle proposition’, retraite et société, 46: 60–89. Kohli, M., Künemund, H. and Vogel, C. (2005) ‘Intergenerational transfers in Europe: a comparative overview’, paper presented at the ESA Conference, September 2005, Torun. Lance, G.N. and Williams, W.T. (1967a) ‘A general theory of classificatory sorting strategies I: hierarchical systems’, Computer Journal, 9, 4: 373-380. Lance, G.N. and Williams, W.T. (1967b) ‘A general theory of classificatory sorting strategies II: clustering systems’, Computer Journal, 10, 3: 271-277. Lehmann, P. and Wirtz, C. (2004) ‘Household formation in the EU: lone parents’, Statistics in Focus, Population and Social Conditions, Theme 3, 5/2004. Le Play, F. (1875) L’Organisation de la famille: selon le vrai modèle signalé par l’histoire de toute les races et de tous les temps, by M.F. Le Play, 2nd edn, revd and corrected, Tours: Alfred Mame, libraires-éditeurs, Dentu, Libraire. Lewis, J. (2003) Should We Worry about Family Change?, Toronto: University of Toronto Press. Ogg, J. and Renaut, S. (2005) ‘Le soutien familial intergénérationnel dans l’Europe élargie’, retraite et société, 46: 30–59. Pfau-Effinger, B. (1998) ‘Arbeitsmarkt- und Familiendynamik in Europa. Theoretische Grundlagen der vergleichenden Analyse’, pp. 177–194, in B. Geissler, F. Maier and B. PfauEffinger (eds), Frauen ArbeitsMarkt. Der Beitrag der Frauenforschung zur sozioökonomischen Theorieentwicklung, Berlin: edition sigma. Pfau Effinger, B. (2004) Development of Culture, Welfare States and Women’s Employment in Europe, Aldershot, Burlington: Ashgate. Pinnelli, A., Hoffman-Nowotny, H.J. and B. Fux (2001) Fertility and New Types of Household Formation in Europe, Population Studies, 35, Strasbourg: Council of Europe Publishing. Plantenga, J. and Siegel, M. (2004) ‘Childcare in a changing world’, Position paper, part I, European Childcare strategies, conference ‘Childcare in a changing world’, Groningen, 23–24 October. Online: Available http: (accessed 11 December 2006). Reher, D. (1998) ‘Family ties in Western Europe: persistent contrasts’, Population and Development Review, 24: 203–234. Sabbadini, L.L. and Romano, C. (2006) ‘Principali trasformazioni dell’uso del tempo in Italia’, paper presented at the conference ‘Andare a tempo’, Turin, 20–21, January.

72

Chiara Saraceno

Saraceno, C. (2005) ‘Family work systems in Europe’, pp. 57–84, in J. Alber and W. Merkel (eds), Europas Osterweiterung: Das Ende der Vertiefung? WZB Jahrbuch 2005, Berlin: edition sigma. Saraceno, C. and Olagnero, M. (2004) ‘Households and families’, pp. 33–44, in European Foundation for the Improvement of Living and Working Conditions, The Quality of Life in Europe, Luxembourg: Office for Official Publications of the European Communities. Saraceno, C., Olagnero, M. and Torrioni, P. (2005) First European Quality of Life Survey: Families, Work and Social Networks, European Foundation for the Improvement of Living and Working Conditions, Luxembourg: Office for Official Publication of the European Communities. Schizzerotto, A. and Lucchini, M. (2004) ‘Transitions to Adulthood’, pp. 46–68, in R. Berthoud and M. Iacovou (eds), Social Europe, Cheltenham: Edward Elgar. Therborn, G. (2004) Between Sex and Power, London: Routledge. Tomka, B. (2002) ‘Demographic Diversity and Convergence in Europe, 1918–1990: The Hungarian case’, Demographic Research, 6, 2. Online. Available Http: (accessed 11 December 2006). United Nations Economic Commission for Europe, Family and Fertility Surveys (2002) Economic Survey of Europe 2002: No. 1, New York and Geneva: United Nations.

3

Is there a generational cleavage in Europe? Age-specific perceptions of elderly care and of the pension system Wolfgang Keck and Agnes Blome

Introduction1 All societies have to solve two basic problems concerning old age: how to provide incomes for elderly people who no longer draw earnings from work, and how to provide care for older people who are disabled and no longer self-sufficient. During the post-war period of prosperity European societies could for many years rely on a rather well-established division of labour between the generations. The population at working age financed fairly generous pay-as-you-go pension schemes, as a combination of full employment, high birthrates, and sizeable productivity gains helped to ease the burden of financing and allowed the raising of pensions to levels which considerably alleviated the problem of poverty in old age.2 Care needs were traditionally covered within the extended family system where a fairly large generation of adult and usually female care workers provided for a relatively small generation of elderly parents. A number of social transformations have put the viability of these arrangements increasingly into question. Shrinking fertility and extended unemployment lead to stagnant or even decreasing numbers of economically active contributors to social security systems despite growing female labour force participation, whilst the growing share of the elderly population and the prolonged life expectancy combine to boost the number of people drawing social benefits. Hence social security systems are faced with growing imbalances between the numbers of contributors and beneficiaries. The demographic changes are particularly incisive. Once the net reproduction rate shrinks below factor 1, each generation of contribution payers in pension schemes will – other things being equal – be smaller than the generation of pension recipients (Breyer 1990). Under these conditions a steady level of pensions can only be bought at the price of successive increases in contribution rates. Alternatively, the contribution rate may remain constant only at the price that each successive generation of pensioners will draw less generous pensions. Various models have been proposed to distribute the burden more fairly between the groups of contributors and

74

Wolfgang Keck and Agnes Blome

beneficiaries,3 but the problem remains that under conditions of shrinking populations existing combinations of pensions and contributions will not be sustainable. The problem is even exacerbated as shrinking levels of fertility combine with increasing life-expectancy. Pensions which were originally designed to cover merely a few years after retirement from work now have to be paid for close to two decades.4 The demographic changes also affect the caring relationship. Dependency upon care increases steeply at very advanced stages in the lifecycle, i.e. beyond ages 75, and especially beyond age 85.5 The proportion of these ‘very old’ age cohorts grows even faster than the proportion of the population aged 65+.6 Hence there will be a growing demand for care. This growing demand meets with a shrinking supply of carers in private households, because growing female labour force participation – and in the long run decreasing fertility – lead to a shrinking supply of adult daughters who traditionally used to shoulder the care work in private households.7 If the growing care needs of people at very advanced ages are to be met, the shrinking caretaker potential in private households calls for a public provision of care services, but this would further increase the tax squeeze for the population at working age. The new macro-constellation has put traditional patterns of generational relations under strain, and there has been growing concern that under the new conditions European societies might face growing generational tensions if not an outright generational conflict (see Arber and Attias-Donfut 2000 for a good overview of respective debates). From a social policy perspective, several authors have suggested considering welfare states as stratifying systems which divide the population into groups with different obligations and entitlements (Lepsius 1979; Alber 1984; Esping-Andersen 1990), and Peter Baldwin (1990) proposed the term ‘risk-categories’ to distinguish between groups for which the welfare state provides different cost/benefit ratios. From this perspective, those who finance public transfers and services and those who benefit from them form two distinct risk categories with discrepant interests. The more welfare states are targeted on needy persons, the less overlap there is between the payers and recipients of public benefits, and the more clashes of interest between these two groups may be expected. In the case of generations, however, each generation appears on both sides of the structural divide during the life-course, paying contributions and rendering care services in an earlier phase of the life-cycle, but receiving benefits and care at a later stage. In this sense, the positions in various riskcategories are not fixed but fluid. Nevertheless, different generations will probably end up with different generation-specific cost–benefit ratios as currently younger age cohorts have to expect lower benefits relative to their contributions than the current generation of pensioners (Auerbach et al. 1998, European Commission 1999). However, the fact that each person will hold positions on either side of the structural divide should considerably lower the potential for generational conflicts. The potential for generational conflicts is further lowered by the fact that family ties cut across the structural cleavage. In addition to knitting emotional bonds between parents and children, families also serve as systems of mutual support where considerable resources are transferred between the generations.8 Hence the amount of tensions between the generations is structurally under-determined and must be empirically researched rather than theoretically deducted. The purpose of this chapter is to investigate the degree to which there is an agespecific polarisation in the perception of public benefit systems in European societies. The basic interest is to describe how much polarisation between age groups there is

Is there a generational cleavage?

75

in the assessment of elderly care and of pension systems, and to analyse to what extent such age-specific discrepancies reflect country-specific welfare state arrangements.

1. Hypotheses and data The analyses depart from a number of hypotheses concerning the age-specific distribution of perceptions. The basic assumption is that younger and older age-cohorts form two distinct risk-categories with different positions concerning the distribution of the costs and benefits of public schemes: younger persons are assumed to answer questions regarding old-age risks from the perspective of the financers of social security provisions, older persons are expected to represent the views of the beneficiaries. With respect to the provision of care, one would expect that younger persons advocate shifting the burden of care from families – i.e. from themselves – to the state or to another third party. In particular, individuals in the medium age group who have elderly parents and therefore have to deal actually or at least potentially with the care issue should opt for public care responsibilities. In contrast, one would expect that older people tend to be more in favour of family solidarity. With respect to financing responsibilities, the basic assumption is that each group will have a preference for shifting the burden of financing to others or at least to public authorities. Since women are the predominant providers of care, they should advocate state schemes – which would help to unburden them – more frequently than men. The position in a risk-category on either side of the care relationship might be mediated by two intervening macro-conditions, i.e. welfare state arrangements and the strength of cultural norms regarding family care (Milar and Warman 1996). We expect age-specific preferences regarding care responsibilities to come more to the fore the less developed public services are – because under these conditions the age-specific division into givers and recipients of care is felt most incisively on the personal level. However, cultural norms stressing family solidarity may cushion the effect of interests tied to a specific risk-category, because the emphasis on family obligations has a dual effect. On the one side, countries following the subsidiarity principle and holding families responsible for the delivery of care, impede the growth of public provisions and thus burden families with care obligations. On the other side, to the extent that such policy is rooted in religious values, they also provide a moral underpinning of family responsibilities which strengthen the normative commitment to family care. This would imply that countries with a strong normative emphasis on family responsibilities will be characterised by lower degrees of age-specific polarisation in the assessment of care responsibilities due to the diffusion of respective norms. With respect to the assessment of pension schemes, older people can be expected to have a positive assessment of current pension arrangements because they have a more favourable balance of benefits and contributions than the younger generation. On the other hand older people might also develop a more negative image of pension schemes because, given that their standard of living crucially depends on pensions they are most painfully affected by cutbacks. In contrast younger generations are less immediately affected by reforms and might hence perceive the pension system in a less negative vein. On the macro-level we would expect countries with more universal, less fragmented, and more sustainable pension schemes to produce smaller degrees of age-specific polarisation in the assessment of public pensions. Finally, we assume that

76

Wolfgang Keck and Agnes Blome

negative assessments of pension arrangements provide a basis for the perception of a generational cleavage. This implies that age-specific differences in the perception of pension schemes should translate into corresponding differences in the perception of generational cleavages, and, secondly, that countries with more negative assessments of their pension schemes should prove to have higher potentials for generational tensions. The data we use to test these hypotheses stem from two different data sources: a compiled version of several Eurobarometer surveys9 which we refer to here as the harmonised candidate countries Eurobarometer 1998–2002 and the European quality of-life survey (for survey and sample details see Kohler, this volume, Ch. 17). In the Eurobarometer respondents were asked what kind of care arrangement they would prefer and who should pay for care provisions. In addition, the data provide information whether people actually provide care, and if so to whom and where. This allows us to compare professed preferences with actual behaviour. The European Quality of Life Survey (EQLS) contained two questions where respondents were asked to evaluate their national pension scheme. The first one inquired whether people tend to trust the state pension system. The second one asked to rate the quality of the state pension system on a ten-point scale from low to high quality. Moreover, the EQLS inquired to what extent individuals think that there are tensions between young and old people in society. This allows us to analyse the impact which assessments of the social security system have on perceived generational conflict.10 Our analyses are based on the distinction of three age groups. The youngest age group ranges from 15 years (18 for the EQLS) up to 34 years. This group represents a broad transition phase from education to the labour market, from the parental home to the formation of an own household as well as towards family formation. The second age group ranges from 35 to 59 years. This age category includes people exposed to the most intensive demands required both by family formation processes and by professional obligations. The older among them are likely to have elderly parents. In the post-communist countries, however, this group has also been affected most strongly by the transformation after the fall of the iron curtain, finding it difficult to keep or get jobs in the labour market. Thus, the living conditions for this age group might differ considerably between old and new member states. Including people aged 60 and over, the third group consists mostly of retirees, because age 60 is more or less the average retirement age in most European countries. The analysis proceeds as follows. First, we will analyse to what extent preferences regarding elderly care differ by age and gender. Here we first look at the preferred care arrangements (section 3.1) and then at policy choices regarding the financing of care (section 3.2). In a second step we look at age-specific similarities and differences in the assessment of national pension schemes (section 4). Finally we examine to what extent age-specific differences in the perception of pension schemes translate into the perception of generational cleavages.

2. Age-specific care preferences As pointed out above, the preferences which people profess with respect to care are presumably a function of their values or normative commitments as well as of their interests. A rational choice perspective would suggest that people express preferences which are above all shaped by their self-interests. Since persons at different stages in

Is there a generational cleavage?

77

the life course are likely to be in different positions as care givers or care receivers we expect age-specific preferences to reflect these different positions. Finding themselves frequently in the position of care recipients, members of the older generation should have a predominant interest in getting high quality care at a low price which is typically rendered by family members.11 Younger generations, on the other side, who are usually in the position of care providers, should have an interest in being relieved from caring obligations and in distributing the cost of care widely among all generations so that the cost does not exclusively impinge upon them. This logic should apply above all to those in the middle age brackets who already have elderly parents so that we expect this group to be most in favour of public care solutions which promise to unburden them. Younger people may similarly be expected to prefer public care solutions which are predominantly financed by the elderly themselves. Apart from different interests reflecting the position in care arrangements, care preferences are presumably also shaped by the interplay between cultural norms regarding family life and by the prevailing welfare state arrangements. Cultural norms may stipulate the family – i.e. women – as ‘natural’ providers for care and support, and an absence of public care services may constrain family members to fulfil otherwise orphaned care obligations. With respect to the impact of cultural norms, we expect cross-national differences in care preferences especially between countries with an emphasis on family obligations and an absence of state-provided services on the one hand, and countries in which the state takes over family responsibilities on the other. Here we assume a gradient between Southern and Northern European countries. We would expect generational differences in care preferences to be most pronounced in countries which neither have well-developed public care services nor a strong cultural emphasis on family obligations. This group should include continental welfare states as well as new member states. The following sections analyse to what extent care preferences differ between age groups and whether cultural norms and welfare state arrangements have an impact on professed attitudes. 2.1 Preferred care arrangements The CC-Eurobarometer asked respondents what arrangement they would prefer in case their elderly parents could not manage to live on their own. Five options were offered: (1) the respondent or one of his siblings should invite the parent to live with them; (2) one of the children should move in with the parent; (3) the parent and children should move closer together; (4) the parent should move into a residential care facility; (5) the parent should stay at home assisted by domestic services. We classified these answers according to whether the respondents preferred a family solution. A preference for a family solution was assumed if respondents advocated any of the first three possibilities. The answers to such questions reveal attitudes or general preferences rather than disclosing what respondents actually do in case a frail elderly relative needs care. The harmonised CC-Eurobarometer dataset, does, however, provide an opportunity to analyse the relationship between professed care preferences and actual caring behaviour, since questions regarding both aspects were asked in the same survey. The result is that people actively involved in care giving to elderly people have positive rather than negative thoughts about extended family responsibilities for care (Alber and Kohler 2004: 71). Since the respective questions were asked in different

78

Wolfgang Keck and Agnes Blome

% preferring family care for own parents

100 90

GR

80

ES RO

PT

70 IT

60

IE

50

LU

40

PL LV

HU

SI

EE

BG

LT

DE

SK AT

UK

TR

R2 = 0.6429

CZ

FR CY MT

BE

30 20

DK

NL

FI

SE

10 0 0

2

4 6 8 10 % giving care to a person over 60 inside household

12

14

Figure 3.1 Share of individuals who care for elderly persons within the household and preference of family members as care givers for own parents Source: Harmonised CC Eurobarometer data set 1998–2002

Eurobarometer surveys in the EU-15, such a micro-level analysis is limited to the new member states. A macro-level comparison which is possible for all nations shows that countries with higher proportions of active care givers within private households also have higher proportions of respondents who advocate family solutions for care (Figure 3.1). Hence on each level there is evidence that people’s professed care preferences are not merely a lip service to a general ideal but tend to reflect the actual care behaviour.12 This implies that the professed preferences are a reliable indicator of the kind of solution for care problems which Europeans from different societies consider appropriate. Figure 3.2 shows how widespread support for family care for the elderly is in each European country as well as in the various country group aggregates. National means are given for the three age groups. Comparing the country group aggregates first, the pro-family attitude is most prevalent in the candidate countries13 and least often reported in the old EU member states. In the NMS-10, which are in between those two groups, the division between the former socialist countries and Cyprus and Malta stands out. More than 50 per cent of the respondents in the post-communist transformation countries, but only around 35–40 per cent in the two Southern islands opt for the family solution. The differences between the various country aggregates are rather small compared to the large variation within the EU-15. Here, support for the family solution ranges from 16 per cent in Sweden to 89 per cent in Greece. The contrast between the Nordic states and the Netherlands with low fractions of family support on the one side and the Southern European countries with very widespread advocacy of family obligations is particularly striking and much more conspicuous than differences between old and new EU member states.

Is there a generational cleavage?

79

Sweden (1) Denmark (1) Netherlands (1) Finland France Belgium Luxembourg UK

Austria (1, 2) Germany (1, 2) Ireland Italy

Portugal Spain Greece (1) Cyprus (1) Malta Slovenia Czech Rep. Estonia Slovakia Lithuania Hungary Latvia Poland Bulgaria Turkey Romania (2) EU-15 NMS CC-3 EU-25 EU-28 0

10

20

30

40

50

60

70

80

90

100

% opt for family solution Age:

15–34 years

35–59 years

60 years and older

Figure 3.2 Share of individuals in different age groups who opt for family support as the preferred care solution by country Source: Harmonised CC-EB 1999–2002 (CC-EB 2002.1, Q20 EB 50.1 1999. Q36): ‘Let’s suppose you had an elderly father or mother who lived alone. What do you think would be best if this parent could no longer manage to live on his/her own?’ Notes: 1 Youngest age group differ significantly from the oldest age group (p